Takashi Asahara Yakult Central Institute for Microbiological Research 1796 Yaho, Kunitachi, Tokyo 186–8650, Japan (e-mail: email@example.com).
Aims: The anti-infectious activity of Bifidobacteria in combination with transgalactosylated oligosaccharides (TOS) against Salmonella enterica serovar Typhimurium LT-2 in an opportunistic antibiotic-induced murine infection model in mice was examined.
Methods and Results:B. breve (strain Yakult) with natural resistance to streptomycin sulphate (SM, MIC: > 4 mg ml–1), when given daily at a dose of 108 cfu/mouse orally under SM treatment was constantly excreted at 1010 cfu g–1 faeces so long as SM was administered, even at 2 weeks after discontinuing administration of B. breve. Explosive intestinal growth and subsequent extra-intestinal translocation of orally infected LT-2 under SM treatment were inhibited by B. breve colonization, and this anti-infectious activity was strengthened by synbiotic administration of TOS with B. breve. Comparison of anti-Salmonella activity among several Bifidobacterium strains with natural resistance to SM revealed that strains such as B. bifidum ATCC 15696 and B. catenulatum ATCC 27539T conferred no activity, even when they reached high population levels similar those of effective strains such as strain Yakult and B. pseudocatenulatum DSM 20439. Both the increase in the concentration of organic acids and the lowered pH in the intestine due to bifidobacterial colonization correlated with the anti-infectious activity. Moreover, the crude cecal extract of B. breve-colonized mice exerted growth-inhibitory activity against LT-2 in vitro, whereas that of the ineffective B. bifidum-colonized cecum showed much lower activity.
Conclusions: Intestinal colonization by bifidobacteria given exogenously together with TOS during antibiotic treatment prevents the antibiotic-induced disruption of colonization resistance to oral infection with S. enterica serovar Typhimurium, and the metabolic activity needed to produce organic acids and lower the intestinal pH is important in the anti-infectious activity of synbiotics against enteric infection with Salmonella.
Significance and Impact of the Study: These results indicate that certain bifidobacteria together with prebiotics may be used for the prophylaxis against opportunistic intestinal infections with antibiotic-resistant pathogens.
Human typhoid fever is a severe systemic infection caused by the bacterium Salmonella typhi and remains endemic in many developing countries, and the emergence of multidrug-resistant Salm. typhi in recent years has been of major concern (Rowe et al. 1997; Chitnis et al. 1999; Mermin et al. 1999; Threlfall et al. 1999; Bhutta et al. 2000; Connerton et al. 2000). Salm. typhi does not cause systemic disease in mammalian hosts other than humans. However, a murine model of human typhoid is provided by infection of susceptible mice with Salm. enterica serovar Typhimurium, which causes a systemic disease. The treatment of mice with streptomycin has been shown to decrease the infective dose of salmonellae by 1000- to 100 000-fold compared to that in untreated controls (Meynell 1955; Bohnhoff et al. 1954). This is because antimicrobial agents disrupt the ecological balance of bowel microflora. Conversely, in germ-free mice and mice that were treated with antibiotics, an abnormally high degree of susceptibility to colonization by aerobic microorganisms was strongly reduced by the administration of murine or even human anaerobic flora (Freter and Abrams 1972; Van der Waaij and Berghuis-de Vries 1974; Hazenberg et al. 1981). These data indicate that normal flora provides protection against colonization by exogenous micro-organisms (for review, see Vollaard and Clasener 1994).
Probiotics are viable cell preparations or foods containing viable bacterial cultures or components of bacterial cells that have beneficial effects on the health of the host (Lee et al. 1999). Many of these probiotics are lactic acid bacteria, and anaerobic bifidobacteria have been reported to be useful in the treatment of disturbed intestinal microflora and diarrhoeal diseases (for review, see Lee et al. 1999). There have been reports that feeding of probiotic bacteria prevents Gram-negative bacterial infections in experimental animals (Nader de Macías et al. 1992; Silva et al. 1999; Shu et al. 2000). We have found that certain strains of bifidobacteria with natural resistance to streptomycin sulphate colonize at high population levels in murine intestines during antibiotic treatment. The main purpose of the present experiments was therefore to test the hypothesis that intestinal colonization by bifidobacteria given exogenously during antibiotic treatment prevents the antibiotic-induced disruption of colonization resistance to oral infection with Salm. enterica serovar Typhimurium and reduces extra-intestinal translocation of this invasive pathogen into the systemic circulation.
MATERIALS AND METHODS
Specific-pathogen-free 7-week-old male BALB/c mice were purchased from Japan SLC (Shizuoka, Japan). Groups of six or eight mice were housed in polypropylene cages (CLEA Japan, Tokyo, Japan) with sterilized bedding under controlled lighting (12 h light, 12 h dark), temperature (24°C), and relative humidity (55%). The mice were maintained on an MF diet (Oriental Yeast Co. Ltd, Tokyo, Japan) and sterilized water (126°C for 30 min), which contained Cl2 at a final concentration of 1·5 p.p.m. (μg ml–1) ad libitum. For treatment of mice with streptomycin sulphate (SM, Sigma Chemical Co., St Louis, MO, USA), SM was dissolved in the water at a concentration of 2 mg ml–1 and given ad libitum. Water bottles were exchanged for freshly prepared bottles every 3 days. All experimental procedures were carried out according to the standards set forth in the Guide for the Care and Use of Laboratory Animals (National Institutes of Health 1985).
Bifidobacterium breve strain Yakult, B. bifidum ATCC 15696, B. catenulatum ATCC 27539T and B. pseudocatenulatum DSM 20439 were used after selection of the strains by confirmation of growth in PY broth (Holdeman et al. 1977), including SM at a dose of 4 mg ml–1. Each strain was cultivated in MRS broth (Difco Laboratories, Detroit, MI, USA) at 37°C for 24 h, and washed with saline twice and then suspended in saline at a concentration of 1–5 × 109 cfu ml–1. Colonization by bifidobacteria was established by daily administration of the bacteria to mice under treatment with SM in drinking water. Periodic examination of stool colony counts of B. breve was carried out for a subset of mice. Briefly, fresh stool specimens (1–2 pellets) were weighed and placed in an Eppendorf tube containing 1 ml of sterilized anaerobic buffer solution (KH2PO4, 0·0225% w/v; K2HPO4, 0·0225% w/v; NaCl, 0·045% w/v; (NH4)2SO4, 0·0225% w/v; CaCl2, 0·00225% w/v; MgSO4, 0·00225% w/v; Na2CO3, 0·3% w/v; L-cysteine hydrochloride, 0·05% w/v; resazurin, 0·0001% w/v) and homogenized with a PELLET PESTLE Mixer (Kontes Co., Vineland, NJ, USA). TOS agar (Sonoike et al. 1986) supplemented with 6·25 mg ml–1 SM and 1 μg ml–1 carbenicillin disodium salt (Sigma, T-CBPC agar) was used for quantification of B. breve strain Yakult, and TOS agar supplemented with 60 μg ml–1 SM was used for selective isolation of the other strains of Bifidobacterium. Medium was cultured anaerobically in an atmosphere of 7% H2 and 5% CO2 in N2 at 37°C for 72 h and the colonies on the plates were counted.
All strains of bifidobacteria were identified by PCR assay by using corresponding species-specific primers for 16S ribosomal RNA (Matsuki et al. 1999).
6′ Transgalactosylated oligosaccharides (TOS, purity > 99%) were prepared as described previously (Matsumoto et al. 1989). The TOS preparation was dissolved in distilled water (10–250 mg ml–1) and a 0·2 ml portion (2–50 mg per mouse) was administered daily to mice. The TOS hydrolysis in vitro was determined as follows: briefly, each bifidobacterium strain or Salm. enterica serovar Typhimurium strain LT-2 at a concentration of 104 cfu ml–1 was incubated in PY broth with TOS at a concentration of 1 mg ml–1 at 37°C for 24 h. The culture was centrifuged at 8000 r.p.m. for 10 min, and the supernatant was semipurified by ion-exchange chromatography (DIAION PA316: PK218=1 : 2, Mitsubishi Chemical Co., Tokyo, Japan) after pretreatment with a solid phase column (Sep-Pak Light C-18 Cartridge, Waters Co., Milford, MA, USA). Samples were then filtered through a 0·45-μm membrane and analysed by HPLC (ICA-3030, TOA Electronics Ltd, Tokyo, Japan, Kimura et al. 1995) equipped with two columns (Shodex SUGAR KS-802, 8 mm I.D. × 300 mm and Shodex SUGAR KS-G, 6 mm I.D. × 50 mm, Showa Denko Co. Ltd, Tokyo, Japan). The metabolic activity was expressed as the value calculated by the following equation: TOS metabolizing activity (%)=(1–the amount of TOS after incubation with bifidobacteria or Salm. enterica serovar Typhimurium strain LT-2/the amount of TOS after incubation in medium) × 100.
Salm. enterica serovar Typhimurium infection
A murine Salm. enterica serovar Typhimurium strain LT-2 gastrointestinal infection model was developed based on the methods of Bohnhoff et al. 1954. In order to assess extra-intestinal translocation of intestinal bacteria, spleen and mesenteric lymph nodes were removed aseptically from the mice and homogenized in 5 ml of sterile saline solution with a Teflon grinder. The numbers of viable Salm. enterica serovar Typhimurium strain LT-2 were determined by their growth on heart infusion agar (Difco) supplemented with 60 μg ml–1 SM, 0·36% (w/v) Na2S2O3·5H2O (Iwai Chemical Co., Tokyo, Japan), 0·1% (w/v) C6H5Na3O7·2H2O (Wako Chemical Co., Osaka, Japan) and 0·1% (w/v) ammonium ferric citrate (Wako, HIT agar) at 37°C for 24 h.
Mice were dissected on day 7 after challenge infection with Salm. enterica serovar Typhimurium strain LT-2. The caecum, mesenteric lymph nodes and spleen were divided longitudinally and fixed overnight in 10% neutral buffered formalin. Paraffin-embedded sections stained with haematoxylin and eosin or Gram-stain were examined light microscopically by a pathologist blinded to the infecting organism.
Examination of caecal bacterial flora
To obtain caecal contents, eight ether-anaesthetized mice per group per period were killed by cervical dislocation. The caecal contents were removed, placed in grinding tubes containing 1 ml of sterilized anaerobic buffer solution, and homogenized with a Teflon grinder. After serial dilution of the caecal suspensions with anaerobic buffer solution, 50-μl portions of the diluents were spread onto the following culture media. Heart infusion agar, supplemented with 0·2 mg ml–1 neomycin (Sigma), 0·01% (w/v) brilliant green, 0·1% (w/v) sodium taurocholate, 0·03% (w/v) L-cysteine hydrochloride, and 5% (w/v) defibrinated horse blood (modified NBGT agar), was used for selective isolation of the Bacteroidaceae. CPLX agar was used for selective isolation of Bifidobacterium (Yuki et al. 1999). LBS agar (Becton Dickinson and Company, Cockeysville, MD, USA), supplemented with 0·8% (w/v) Lab Lemco powder (Oxoid Ltd, Basingtoke, UK), 0·1% (w/v) sodium acetate-trihydrate and 0·37% (w/v) acetate, was used for selective isolation of Lactobacillus. COBA agar was used for selective isolation of Enterococcus (Petts 1984). DHL agar (Nissui Pharmaceutical Co. Ltd, Tokyo, Japan) was used for selective isolation of the Enterobacteriaceae. Staphylococcus medium no. 110 agar (Nissui) was used for selective isolation of Staphylococcus and Bacillus. T-CBPC agar was used for quantification of B. breve strain Yakult. VL-G roll tube agar (Azuma and Suto 1970) supplemented with 0·2% (w/v) cellobiose and 0·2% (w/v) maltose (modified VL-G roll tube agar) was used for determination of total anaerobe counts. Modified NBGT agar, CPLX agar, LBS agar and T-CBPC agar media were cultured anaerobically in an atmosphere of 7% H2 and 5% CO2 in N2 at 37°C for 72 h. After incubation, the colonies on the plates were counted and Gram-stained. Species and biotypes of the bacteria were identified with API systems (bioMerieux S.A., Montalieu-Vercieu, France): rapid ID 32 A for the Bacteroidaceae and Lactobacillaceae, API 20 STREP for the Enterococcaceae, API 20 E for the Enterobacteriaceae and API 20 STAPH for the Staphylococcaceae. The lower limit of bacterial detection with this procedure was 100 cfu g–1 faeces.
Scanning for anaerobic fujiform bacteria was carried out by microscopic bacterial counts (Voravuthikunchai and Lee 1987). For quantification, 10-μl portions of the diluents were put into a 10-well immunofluorescence slide (Flow Laboratories, Inc., McLean, VA, USA), fixed and Gram-stained. Fujifom bacteria were counted with the aid of an ocular grid containing 100 squares (calibrated with a stage micrometer), a 100 × objective, and a 10 × ocular. Acceptable slides considered for analysis met the following two criteria. (i) The bacteria appeared to be evenly distributed, and (ii) the number of bacteria per 100 grid squares was between 20 and 300. Counts were made in 10 fields chosen randomly.
Results are expressed as means ± standard deviation (SD) numbers of cfu per 1 g of caecal contents.
Detection of organic acids in caecal contents
The caecal contents were homogenized in 1 ml of distilled water, and the homogenate was centrifuged at 15 000 r.p.m. at 4°C for 10 min. A mixture of 0·9 ml of the resulting supernatant and 0·1 ml of 1·5 mol l–1 perchloric acid in a glass tube was mixed well and allowed to stand at 4°C for 12 h. The suspension was then passed through a filter with a pore size of 0·45 μm (Millipore Japan Ltd, Tokyo, Japan). The sample was analysed for organic acids by HPLC as described previously (Kikuchi and Yajima 1992). The HPLC was performed with a TOA system (ICA-3030 TOA Electronics Co. Ltd, Tokyo, Japan) equipped with two columns (Ion Pack KC-811 Shodex, 8 mm I.D. × 300 mm, Showa Denko). The concentrations of organic acids were calculated using external standards.
In vitro growth inhibitory activity
The caecal content (300 mg) was homogenized in 1 ml of distilled water, and the homogenate was centrifuged at 10 000 r.p.m. at 4°C for 10 min. The supernatant was filtered through a 0·22-μm filter to which sterile distilled water was added to archive a 30% (w/v) solution. Salm. enterica serovar Typhimurium LT-2 was suspended in the solution at a concentration of 104 cfu ml–1 and incubated at 37°C for 0, 1, 3 or 6 h. Viable counts in the triplicate cultures were determined using selective agar, as described above, and the results are expressed as the percentage growth: % growth=(the mean viable Salm. enterica serovar Typhimurium LT-2 count in the caecal extract of normal or Bifidobacteria-colonized mice/the mean viable LT-2 count in the caecal extract of SM-treated control mice) × 100.
The second set of experiments was conducted to assess the effect of acetic acid and pH on in vitroSalm. enterica serovar Typhimurium LT-2 growth. Briefly, conditioned BHI broth was prepared by the addition of acetic acid (AA) followed by pH adjustment: AA; 60 mmol l–1, pH 6·4: normal caecum, AA; 40 mmol l–1, pH 7·0:SM-treated control, AA; 45 mmol l–1, pH 6·75:SM-treated and B. breve-colonized and AA; 29 mmol l–1, pH 6·8:SM-treated and B. bifidum-colonized. LT-2 was then suspended in the solution at a concentration of 104 cfu ml–1 and incubated at 37°C for 0, 1, 3 and 6 h. Viable counts in the triplicate cultures were determined using selective agar, as described above, and the results were expressed as the percent growth: % growth=(the mean viable Salm. enterica serovar Typhimurium LT-2 count in the conditioned medium of the normal or Bifidobacteria-colonized group/the mean viable LT-2 count in the conditioned medium of the SM-treated control) × 100.
The average number of bacteria was analysed by the Dunnet t-test to determine significant differences between the treatment and control groups. A significant difference was defined as P < 0·05.
Stable colonization of exogenous Bifidobacterium breve strain Yakult in SM-treated mice
Continuous treatment of mice with streptomycin sulphate (SM) in drinking water (2 mg ml–1) resulted in selective intestinal flora decontamination: the indigenous species such as Bifidobacteriaceae, Lactobacillaceae and Enterobacteriaceae decreased to undetectable levels, whereas there were no significant differences in the colonization levels of the other bacterial genuses tested (Table 1). B. breve, when given daily to mice under treatment with SM in drinking water, grew aggressively in the intestine, and the increases in the treated group (more than 2 mg ml–1) were maintained for 2 weeks even after stopping administration of bifidobacteria (Fig. 1). In neither the B. breve-administered group nor the group given B. breve in combination with 6′ transgalactosylated oligosaccharides (TOS) were there any significant differences in bacterial flora species levels compared to those of SM-treated controls except for detection of the B. breve being given daily (Table 1).
Table 1. Effect of Bifidobacterium breve strain Yakult colonization on indigenous caecal microflora under Streptomycin treatment in mice
Inhibition of opportunistic intestinal infection with Salm. enterica serovar Typhimurium LT-2 by administering B. breve to SM-treated mice
As shown in Fig. 2, faecal excretion of Salm. enterica serovar Typhimurium strain LT-2 was increased dramatically after a single inoculation of the pathogen at a small inoculum of 102 cfu in the SM-treated control group, and the increased excretion level of more than 108 cfu per g faeces was maintained until day 7 after the inoculation. Daily administration of B. breve strain Yakult (108 cfu/mouse/day) for 14 days starting on day –7 markedly decreased the faecal Salm. enterica serovar Typhimurium strain LT-2 excretion levels, and B. breve in combination with TOS more markedly decreased these faecal excretion levels. TOS administration alone did not significantly change the viable counts of Salm. enterica serovar Typhimurium strain LT-2 excreted in faeces. Mice were dissected on day 7 after the Salm. enterica serovar Typhimurium strain LT-2 inoculation, and extra-intestinal translocation of the bacteria was examined (Table 2). In the saline-treated control group, extra-intestinal translocation was observed in all animals. Administration of B. breve (with TOS) completely blocked systemic bacterial translocation. Examination of the TOS dose revealed that combining B. breve and TOS at doses ranging from 2 to 50 mg per mouse per day conferred anti-infectious activity while TOS treatment by itself even at the highest dose (50 mg mouse–1 day–1) did not affect Salm. enterica serovar Typhimurium strain LT-2 infection.
Table 2. Inhibition of extra-intestinal translocation of Salmonella enterica serovar Typhimurium LT-2 by Bifidobacterium breve strain Yakult in combination with 6′ transgalactosylated oligosaccharides
Histopathological analysis of the infection control group showed clearly the inflammatory responses indicated by neutrophil infiltration and dissociation of the mucosal layer from the caecal epithelium layer (Fig. 3, panel a). Neither a notable inflammatory response nor mucosal damage was observed in the B. breve-treated group (Fig. 3, panel d). Numerous granulomatous lesions in the spleen (arrows, Fig. 3, panel b) and mesenteric lymph nodes (arrows, Fig. 3, panel c) were observed only in the control group while there was no significant histopathological damage in the organs of the B. breve-treated group (panels e, f). Neither SM treatment nor B. breve administration under SM treatment by itself caused histopathological changes in the organs examined (data not shown).
Comparison of anti-infectious activities among several strains of bifidobacteria with natural resistance to SM
Preliminary experiments revealed that four of 45 bifidobacteria strains tested for their growth ability in liquid medium, in the presence of SM at 4 mg ml–1, showed positive responses (data not shown). The strains were then assessed for their anti-infectious activities against Salm. enterica serovar Typhimurium strain LT-2 in vivo. Although the strains colonized the intestine at similar high population levels (Table 3), there were marked differences in anti-infectious activity in combination with TOS among the strains (Table 4). Two out of four strains tested showed potent anti-infectious activity but B. bifidum ATCC 15696 and B. catenulatum ATCC 27539T had no anti-infectious activity. There were no correlations between the anti-infectious activity (Table 4) and TOS-metabolic activity in vitro (Table 3) among the strains.
Table 3. Comparison of colonization level and 6′ transgalactosylated oligosaccharides availability among Bifidobacterium strain
Table 4. Comparison of anti-infectious activity among Streptomycin-resistant Bifidobacterium strains in combination with 6′ transgalactosylated oligosaccharides
The organic acid concentration was decreased and pH increased in the caecum by SM treatment. Organic acid and pH were unchanged on day 7 after Salm. enterica serovar Typhimurium strain LT-2 infection in the control group. Significantly lower pH and higher concentrations of both total organic acids and acetic acid in the two groups, i.e. B. breve-and B. pseudocatenulatum-treated groups, were detected on day 7 after the infection (Fig. 4) compared with those of the corresponding controls. On the other hand, there were no significant recoveries of such markers in the groups treated with ineffective strains such as B. bifidum and B. catenulatum.
Growth-inhibitory activity of faecal extract on Salm. enterica serovar Typhimurium strain LT-2 growth in vitro
Growth of Salm. enterica serovar Typhimurium strain LT-2 in the water-extracted caecal content (30% w/v) obtained from the normal mice was significantly inhibited compared with that in the caecal content extract obtained from the SM-treated control group (5Fig. 5a). The extract of the caecum colonized by B. breve under SM treatment showed marked growth-inhibitory activity against Salm. enterica serovar Typhimurium strain LT-2, whereas no significant activity was observed in the extract of the B. bifidum-colonized cecum (5Fig. 5a).
Growth-inhibitory activity of acetic acid against Salm. enterica serovar Typhimurium strain LT-2 in vitro was then tested in order to assess whether the lowered pH and increased acetate concentration of the caecal content were responsible for the growth-inhibitory activity of B. breve. The results presented in 5Fig. 5b show clearly that acetate at a concentration of 45 mmol l–1 at pH 6·75, which is equivalent to the condition of the B. breve-colonized caecum, inhibits the in vitro growth of Salm. enterica serovar Typhimurium strain LT-2 whereas that of the ineffective B. bifidum-colonized caecum showed no such effect.
The term colonization resistance was introduced by van der Waaij et al. in 1971 to indicate resistance to colonization by exogenous potentially pathogenic micro-organisms (PPMOs, Van der Waaij et al. 1971). Vollaard and Clasener concluded that the flora providing colonization resistance to exogenous micro-organisms are identical to the flora limiting the concentration of indigenous PPMOs (Vollaard and Clasener 1994). In previous reports, we demonstrated that injection of a lethal dose of 5-fluorouracil (400 mg kg–1) into mice induced an extraordinary increase in the levels of indigenous Escherichia coli in the intestine and systemic translocation of these bacteria to the liver and that bifidobacteria is the only species that is markedly decreased after treatment with this chemotherapeutic agent (Nomoto et al. 1991; Nomoto et al. 1998). Moreover, daily administration of fermented milk containing probiotic Bifidobacterium breve strain Yakult prevented both the drug-induced intestinal outgrowth and extra-intestinal translocation of indigenous E. coli (Asahara et al. 2001), and analysis of the concentration of short chain fatty acids (SCFAs) in the intestinal contents suggested that the administered species compensated for decreased production of SCFAs by the disrupted indigenous microflora. Herein, we have demonstrated antibiotic-induced intestinal overgrowth and extra-intestinal translocation of Salm. enterica serovar Typhimurium to be markedly inhibited by precolonization of the intestine by specific strains such as probiotic B. breve, and that pH-lowering and acid-producing effects appear to be important for this anti-infectious activity. A high colonization level of bifidobacteria by itself is not enough for protection from Salmonella infection, because B. bifidum ATCC 15696, which achieved the highest colonization levels among the strains tested both at the initiation of and during progressive phases of infection, neither exerted anti-infectious activity nor improved intestinal environmental factors such as pH and organic acid concentrations (Tables 3, 4, Fig. 4). Taken together, these observations suggested that compensation for the chemotherapy-induced disruption of colonization resistance by probiotic bifidobacteria may be effective in the prevention of intestinal infection by PPMOs and that not only the population level but also the metabolic activity of the colonizer is important for this.
6′ Transgalactosylated oligosaccharides (TOS, Matsumoto et al. 1989) are oligosaccharides produced by transgalactosylation of lactose using a β-galactosidase (β-Gal). Van der Meer et al. have reported that lactic acid in yogurt, intestinal lactic acid fermentation of non-digestible saccharides such as lactulose and supplemental dietary calcium significantly improve the intestinal resistance of rats to Salm. enteritidis infection (Bovee-Oudenhoven et al. 1997; Van der Meer and Bovee-Oudenhoven 1998). Combining TOS with B. breve excludes intestinal Salm.enterica serovar Typhimurium LT-2 more effectively than B. breve alone, while TOS by itself has no anti-infectious activity (Fig. 2). It has been shown that TOS is fermented not only by bifidobacteria but also by Lactobacillus, Bacteroides and Clostridium species (Matsumoto et al. 1989). The observation that TOS at a high dose (50 mg mouse–1 day–1) by itself was not effective against infection (Table 2) suggests that the intestinal inhabitants resulting from SM treatment do not metabolize TOS to produce adequate amounts of organic acids and a low enough pH to inhibit experimental infection, and that administration of prebiotics is effective when corresponding lactic acid bacteria are also present in the intestines. To our knowledge, this is the first evidence that anti-infectious activity against enteric pathogens is exerted by a synbiotic combination of probiotics and prebiotics under conditions in which indigenous flora have been selectively eliminated by antibiotics.
It has been reported that organic acids such as acetic acid, lactic acid and citric acid possess higher bactericidal activity than the non-organic acids such as hydrochloric acid, and that the bactericidal activity of the organic acids depends mainly on their undissociated form (Eklund 1983; Brocklehurst and Lund 1990). Undissociated organic acids can permeate a cell membrane by diffusion and release protons in the cell. The influx of protons is thought to induce cytoplasm acidification and dissipate the membrane proton potential (ΔpH) (Eklund 1983; Brocklehurst and Lund 1990; Cramer and Prestegard 1997). This leads to disruption of the proton motive force and inhibits substrate transport mechanisms, energy-yielding processes and macromolecule synthesis (Cherrington et al. 1990; Diez- Gonzalez and Russell 1997). In addition, anion accumulation is assumed to exert bacterial toxicity (Russell 1991). The results of in vitro studies (Fig. 5) suggested that the higher concentration of undissociated acetic acid plays a pivotal role in the anti-infectious activity of B. breve and B. pseudocatenulatum.
Silva et al. reported that daily treatment of conventional and gnotobiotic mice with B. bifidum milk protected the mice from a lethal infection with Salm. enteritidis subsp. typhimurium as demonstrated by survival and histopathological data (Silva et al. 1999). They concluded that the protective effect was not due to reduction of the intestinal populations of pathogenic bacteria because the treatment did not reduce the colonization levels of pathogenic bacteria, and suggested that other factors, such as immunomodulation, are required to explain the protective effect of bifidus milk against Salm. enteritidis subsp. typhimurium. On the other hand, Shu et al. have reported that feeding mice B. lactis had a protective effect, as shown by an increase in the survival rate, higher post-challenge food intake and weight gain and reduced extra-intestinal translocation of Salm. typhimurium (Shu et al. 2000). They also speculated that the anti-infectious mechanism of B. lactis involved enhancement of various parameters of immune function that are relevant to the immunological control of salmonellosis. Because much has been reported concerning host immune responses to Salm. enterica serovar Typhimurium (for review, see Jones and Falkow 1996; Schaible et al. 1999), typical intracellular bacteria, studies are now in progress to examine whether persistent colonization of the intestine by exogenously given bifidobacteria enhances host immune responses (both humoral and cellular) against Salmonella.
We would like to thank Kazumi Uchida for contributing the histopathological analysis. We also thank Keisuke Matsumoto and Masakazu Ikeda for assisting preparation of TOS and TOS hydrolysis analysis.