Protective effect of Lactobacillus casei strain Shirota against lethal infection with multi-drug resistant Salmonella enterica serovar Typhimurium DT104 in mice

Authors


Takashi Asahara, Yakult Central Institute for Microbiological Research, 1796 Yaho, Kunitachi, Tokyo 186-8650, Japan. E-mail: takashi-asahara@yakult.co.jp

Abstract

Aims:  The anti-infectious activity of lactobacilli against multi-drug resistant Salmonella enterica serovar Typhimurium DT104 (DT104) was examined in a murine model of an opportunistic antibiotic-induced infection.

Methods and Results:  Explosive intestinal growth and subsequent lethal extra-intestinal translocation after oral infection with DT104 during fosfomycin (FOM) administration was significantly inhibited by continuous oral administration of Lactobacillus casei strain Shirota (LcS), which is naturally resistant to FOM, at a dose of 108 colony-forming units per mouse daily to mice. Comparison of the anti-Salmonella activity of several Lactobacillus type strains with natural resistance to FOM revealed that Lactobacillus brevis ATCC 14869T, Lactobacillus plantarum ATCC 14917T, Lactobacillus reuteri JCM 1112T, Lactobacillus rhamnosus ATCC 7469T and Lactobacillus salivarius ATCC 11741T conferred no activity even when they obtained the high population levels almost similar to those of the effective strains such as LcS, Lact. casei ATCC 334T and Lactobacillus zeae ATCC 15820T. The increase in concentration of organic acids and maintenance of the lower pH in the intestine because of Lactobacillus colonization were correlated with the anti-infectious activity. Moreover, heat-killed LcS was not protective against the infection, suggesting that the metabolic activity of lactobacilli is important for the anti-infectious activity.

Conclusion:  These results suggest that certain lactobacilli in combination with antibiotics may be useful for prophylaxis against opportunistic intestinal infections by multi-drug resistant pathogens, such as DT104.

Significance and Impact of the Study:  Antibiotics such as FOM disrupt the metabolic activity of the intestinal microbiota that produce organic acids, and that only probiotic strains that are metabolically active in vivo should be selected to prevent intestinal infection when used clinically in combination with certain antibiotics.

Introduction

Salmonella enterica serovar Typhimurium definitive type 104 (DT104) is an increasingly common multiple-antibiotic-resistant strain that has rapidly emerged as a world health problem (Threlfall et al. 1996; Martin et al. 2004; Doorduyn et al. 2008), and most DT104 strains are characterized by chromosomal resistance to ampicillin (A), chloramphenicol (C), streptomycin (S), sulfonamides (Su) and tetracycline (T), and they are referred to as the ACSSuT-type (Threlfall et al. 1994). DT104 infection is difficult to prevent or not tr/eatable only by the antibiotics (chloramphenicol, new quinolones and fosfomycin).

Probiotics have been defined as ‘live microorganisms which when administered in adequate amounts confer a health benefit on the host’ (Reid 2005). Many probiotics contain lactic acid bacteria, and anaerobic bifidobacteria have been reported to be useful in the treatment of disturbed intestinal microbiota and diarrheal diseases (Hoppe et al. 2009). In addition, there have been reports that feeding probiotic bacteria prevents Gram-negative bacterial infections in experimental animals (Silva et al. 1999; Shu et al. 2000; Asahara et al. 2001a,b; Gill et al. 2001; Asahara et al. 2004; Zoumpopoulou et al. 2008). Szabóet al. (2009) reported that oral administration of Enterococcus faecium NCIMB 10415 enhanced specific antibody titres against DT104 in serum of weaning piglets, but found no clear protective effects against DT104 infection.

For probiotics to be effective in the prevention of DT104 infection in combination with an antibiotic, the probiotics must be resistant to the antibiotic. We have selected certain strains of lactobacilli that have natural resistance to fosfomycin disodium salt (FOM), to determine whether combined use of FOM and the selected strains would be effective in preventing salmonellosis in mice.

Materials and methods

Animals

Specific-pathogen-free 6-week-old male BALB/c mice were purchased from Charles River Japan, Inc (Kanagawa, 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 given free access to an MF diet (Oriental Yeast, Tokyo, Japan) and sterilized water (126°C for 30 min) that contained Cl2 at a final concentration of 1·5 ppm (μg ml−1). Fosfomycin disodium salt (FOM; Sigma-Aldrich Corporation, St. Louis, MO, USA) was given to mice (14 per group) at a concentration of 5 mg ml−1 in their drinking water throughout the experimental period. Water bottles were replaced with 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).

Lactobacilli

The Lactobacillus strains used were Lactobacillus brevis ATCC 14869T, Lact. casei strain Shirota (LcS), Lact. casei ATCC 334T, Lactobacillus plantarum strain ATCC 14917T, Lactobacillus reuteri strain JCM 1112T, Lactobacillus rhamnosus ATCC 7469T, Lactobacillus salivarius ATCC 11741T and Lactobacillus zeae ATCC 15820T. They were cultured in MRS broth (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) at 37°C for 24 h, washed with distilled water twice and then suspended in distilled water at a concentration of 1–3 × 109 colony-forming units (CFU) ml−1. Colonization by the Lactobacillus strains was achieved by daily administration of the bacteria to mice being treated with FOM in their drinking water. Periodic faecal colony counts of lactobacilli were made in a subset of mice. Fresh stool specimens (1–2 pellets) were placed in an Eppendorf tube containing 0·5 ml of sterilized saline solution and homogenized with a Pellet Pestle Mixer (Kimble/Kontes, Vineland, NJ, USA). LLV agar (Yuki et al. 1999) was used to quantitate the inoculated Lactobacillus strains aerobically. All strains of lactobacilli were identified by polymerase chain reaction assay by using corresponding species-specific primers for 16S ribosomal RNA genes (Watanabe et al. 1997; Song et al. 2000). The bacteria recovered from mouse faeces were identified with the administered strains using PCR-based random amplified polymorphic DNA (RAPD). PCR-based RAPD fingerprinting was carried out by the method of Akopyanz et al. (1992) using two primers (Yuki et al. 2000). LcS was identified using ELISA with strain-specific monoclonal antibodies (Yuki et al. 1999).

A heat-killed bacterial suspension was prepared by suspending the harvested LcS cells in distilled water at a concentration of 1–3 × 109 CFU ml−1, then heating them at 85°C for 6 min and cooling on ice.

Salmonella enterica serovar Typhimurium infection

Salm. Typhimurium DT104 strain T980023 (DT104) isolated from a faecal culture of a patient with clinical symptoms was kindly provided by Dr. Haruo Watanabe, National Institute of Infectious Diseases, Tokyo, Japan. For the challenge experiments, DT104 was cultured overnight in brain–heart infusion broth (Becton, Dickinson and Company). After washing with saline solution by centrifugation, the bacteria were resuspended in saline solution and adjusted to approximately 1–3 × 1010 CFU ml−1, and 100 μl of the suspension was administered orally through a gastric sonde (Fuchigami kikai, Kyoto, Japan) on day 6 after starting FOM treatment. Mice (n = 8) were observed for survival for 21 days after DT104 challenge. To obtain caecal contents, six ether-anaesthetized mice per group were killed by cervical dislocation on day 7 after challenge infection with DT104. The caecal contents were removed, placed in grinding tubes containing 1 ml of sterilized anaerobic buffer solution [0·0225% (w/v) KH2PO4 (Wako Pure Chemical, Osaka, Japan), 0·0225% (w/v) K2HPO4 (Wako), 0·045% (w/v) NaCl (Wako), 0·0225% (w/v) (NH4)2SO4 (Wako), 0·00225% (w/v) CaCl2 (Wako), 0·00225% (w/v) MgSO4 (Wako), 0·3% (w/v) Na2CO3 (Wako), 0·05% (w/v) l-cysteine hydrochloride (Wako), 0·0001% (w/v) resazurin (Sigma)] and homogenized with a Teflon grinder. After serial dilution of the caecal suspensions with anaerobic buffer solution, 50 μl portions of the diluted samples were inoculated onto Heart infusion agar (Becton, Dickinson and Company) supplemented with 60 μg ml−1 streptomycin sulfate (Sigma), 8 μg ml−1 cefsulodin sodium (Sigma), 0·36% (w/v) sodium thiosulfate pentahydrate (Iwai Chemical, Tokyo, Japan), 0·1% (w/v) trisodium citrate dihydrate (Wako) and 0·1% (w/v) ammonium ferric citrate (Wako) (HIT agar) at 37°C for 24 h. To assess extra-intestinal translocation of intestinal bacteria, the spleen and mesenteric lymph nodes were removed from the mice aseptically and homogenized in 5 ml of sterile saline solution with a Teflon grinder. Viable DT104 were counted by growing them on HIT agar at 37°C for 24 h.

Histopathology

Mice were dissected on day 7 after challenge infection with DT104. The caecum and the spleen were divided longitudinally and fixed overnight in 10% neutral buffered formalin. Paraffin-embedded sections stained with haematoxylin were examined under a light microscope by a pathologist blinded to the infecting organism.

Fluorescence in situ hybridization (FISH) analysis

The specimens for FISH were taken from the caecum of mice on day 7 after challenge infection with DT104. The specimens were fixed in 4% (w/v) paraformaldehyde without shaking and embedded into paraffin blocks by standard techniques. Four-micrometer thick sections were placed on Micro Slide Glass (MAS-GP type A, Matsunami Glass, Osaka), baked at 37°C for 16–20 h and stored at 4°C. Sections were hybridized with oligonucleotide probes as described previously (Snaidr et al. 1997). The oligonucleotide 23S rRNA sequence Sal-1 [positions 341–360 (sequence, 5′-ACAGCACATGCGCTTTTGTG-3′)] was used for specific detection of Salmonella spp (Fang et al. 2003). The probe Eub338 (Amann et al. 1996) was used as the positive control because it is complementary to the 16S rRNA of all bacteria. Oligonucleotide probes synthesized were labeled with either Cy3 (Sal-1) or Cy5 (Eub338) fluorochrome at the 5′ end. The observation and acquisition of fluorescent images were performed with a Leica Q550FW system (Leica, Wetzlar, Germany, Takada et al. 2004). Hybridized bacterial cells were observed with either a Y3 filter (excitation 535/50 nm, emission 610/75 nm) for Cy3- or a Y5 filter (excitation 620/60 nm, emission 700/75 nm) for Cy5-labelled oligonucleotide probes. Samples were counterstained with 4′, 6′-diamidino-2-phenylindole (DAPI) to visualize DNA. DAPI-stained DNA was observed with a A4 filter (excitation 360/40 nm, emission 470/40 nm).

Detection of organic acids in caecal contents

The caecal contents were homogenized in 1 ml of distilled water and the homogenate was centrifuged at 12 000 g 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, Tokyo). The sample was analysed for organic acids by HPLC as previously described (Kikuchi and Yajima 1992). The HPLC was performed with a Waters system (Waters 432 Conductivity Detector; Waters Corporation, Milford, MA, USA) equipped with two columns (Shodex Rspack KC-811; Showa Denko, Tokyo, Japan). The concentrations of organic acids were calculated by using external standards.

In vitro growth-inhibitory activity

A set of experiments was conducted to assess the effect of acetic acid and/or lactic acid and pH on in vitro DT104 growth. Briefly, conditioned BHI broth was prepared by addition of acetic acid (A.A.) and/or lactic acid (L.A.) followed by pH adjustment: FOM-treated control, A.A. 25 mmol l−1, L.A. 1·2 mmol l−1, pH 7·2; FOM-treated and LcS-colonized, A.A. 58 mmol l−1, L.A. 7·0 mmol l−1, pH 6·5; FOM-treated and Lact. salivarius-colonized, A.A. 30 mmol l−1, L.A. 3·0 mmol l−1, pH 7·15. DT104 was then suspended in the solution at a concentration of 104 CFU ml−1 and incubated 37°C for 0, 1, 3 and 6 h. Viable counts in the triplicate cultures were made using selective agar, as described above, and the results were expressed as the means and SD of value from triplicate cultures.

Statistical analysis

Average numbers of bacteria were analysed by the Dunnet t-test to identify significant differences between the treatment group and control group. Differences in survival ratios were determined using Fisher’s exact probability test followed by correction with the Bonferroni inequality equation. A significant difference was defined as < 0·05.

Results

Anti-infectious activity of Lactobacillus casei strain Shirota against Salmonella enterica serovar Typhimurium DT104 infection in FOM-treated mice

As shown in Fig. 1a, after a single inoculation of the pathogen in an inoculum of 109 CFU, the number of DT104 in the caecal contents of the infected control group was as low as 103 CFU g−1 caecal contents on day 3 after infection, which, however, increased thereafter to 107 CFU g−1 caecal contents by day 7. Daily administration of LcS (108 CFU per mouse per day) markedly decreased the level of caecal excretion of DT104 in this infection model, whereas heat-killed (HK)-LcS had no significant effect on the viable counts of DT104. Extra-intestinal translocation was observed in all animals tested in the infected control group and HK-LcS-treated group (Fig. 1b,c). By contrast, administration of living LcS inhibited the systemic bacterial translocation completely. The invasion of DT104 in the caecal epithelium was clearly shown as the FISH images of DT104 proliferating in epithelial cells and DT104-infected epithelial cells being detached into the intestinal lumen were obtained (Fig. 2a). Moreover, while marked infection was observed in both the control and group administered HK-LcS, no DT104 invasion in the caecal epithelium was noted in the group administered living LcS (Fig. 2). A dramatic decrease in body weight and subsequent death were observed in all of the mice (n = 8) in both the infected control group and HK-LcS-treated group within 11 days after DT104 infection (Fig. 3). Significant prevention of body weight loss and subsequent death was observed in the living-LcS-treated group compared with the infected control group and HK-LcS-treated group (body weight, < 0·01; Percent survival, < 0·05). Histopathological analysis of the infection control group showed clear damage to intestinal epithelial cells, especially around the Peyer’s patches and lymphoid follicles, and infiltration of cryptic abscesses and the lamina propria by inflammatory cells (Fig. 4a,d). Inflammatory cells, such as macrophages and neutrophils, had surrounded the necrotic spleen cells and formed diffuse areas of nodular necrosis. On the other hand, no marked inflammatory response or mucosal damage was observed in the living-LcS-treated group (Fig. 4b,e). The histopathological damage in the HK-LcS-treated group was similar to that in the infected control group (Fig. 4c,f). Neither the FOM treatment nor LcS administration during FOM treatment by itself caused histopathological changes in the organs examined (data not shown).

Figure 1.

 Inhibitory activity of Lactobacillus casei strain Shirota (LcS) against Salmonella enterica serovar Typhimurium DT104 (DT104) infection in fosfomycin (FOM)-treated mice. FOM was given to mice at a concentration of 5 mg ml−1 in their drinking water from day −6 until day 7. LcS (1–3 × 108 CFU per mouse), a heat-killed (HK) preparation, or saline in an inoculum of 0·1 ml per mouse was administered to the mice daily from day −5 until day 6. Mice (24 mice per group) were infected with DT104 at an inoculum of 109 CFU on day 6 after starting FOM treatment, and six mice in each group were dissected on days 1, 3, 5 and 7 after infection. Viable counts of DT104 and LcS were performed as described in the text. Symbols: •, No. of DT104 in the DT104-infected control mice; □, No. of DT104 in the live-LcS-treated mice; △, No. of DT104 in the HK-LcS-treated mice; bsl00083, No. of LcS in the live-LcS-treated mice. The results are expressed as the mean Log10 CFU and SD per g caecal contents (a) or entire spleen (b) or MLN (c).

Figure 2.

 Observation of DT104 invading the caecal epithelium by fluorescence in situ hybridization (FISH). FOM was given to mice at a concentration of 5 mg ml−1 in their drinking water from day −6 until day 7. LcS (1–3 × 108 CFU per mouse), HK-LcS or saline at an inoculum of 0·1 ml per mouse was administered to mice once daily from day −5 until day 6. Mice were orally infected with DT104 (1 × 109 CFU) on day 0 and they were dissected on day 7 after infection for FISH of the caecum. (a) Infected controls; (b) live-LcS-treated mice; (c) HK-LcS-treated mice. DT104 was shown in pink (Sal-1-Cy3 probe + Eub338 probe-Cy5).

Figure 3.

 Anti-infectious activity of LcS against DT104 infection in FOM-treated mice. FOM was given to mice at a concentration of 5 mg ml−1 in their drinking water from day −6 until day 21. LcS, 1–3 × 108 CFU per mouse per day, in an inoculum of 0·1 ml per mouse was administered to the mice once daily from day −5 until day 21. Mice (eight mice per group) were infected orally with DT104 (1 × 109 CFU) on day 0. (a) The survival of DT104-infected control mice (•), LcS-treated mice (□) and HK-LcS-treated mice (△) was monitored for 21 days after the challenge infection. (b) All the mice in each group were weighed daily until day 7. Symbols: •, infection control; □, live-LcS; △, HK-LcS. The results are expressed as the means and SDs. Significant difference was observed between the LcS-treated group and the DT104-infected control group (**< 0·01).

Figure 4.

 Histopathological analysis of infected organs. FOM was given to mice at a concentration of 5 mg ml−1 in their drinking water from day −6 until day 7. LcS (1–3 × 108 CFU per mouse), HK-LcS, or saline at an inoculum of 0·1 ml per mouse was administered to mice once daily from day −5 until day 6. Mice were orally infected with DT104 (1 × 109 CFU) on day 0 and they were dissected on day 7 after infection for histopathological analysis of the caecum [(a), (b) and (c)] and spleen [(d), (e) and (f)]. (a) and (d), infected controls; (b) and (e), live-LcS-treated mice; (c) and (f), HK-LcS-treated mice. Arrows indicate numerous granulomatous lesions in the spleen [(d) and (f)]. Magnification; ×10.

Comparison of anti-DT104 activity among different Lactobacillus strains

In the next series of experiments, the anti-DT104 activity of selected Lactobacillus strains belonging to different species because of their strong natural resistance to FOM and their similar growth and lactate production in the presence of the antibiotic in vitro was examined. As shown in Table 1 (Exp. 1), only the probiotic LcS exerted potent anti-infectious activity against DT104, and none of the other type strains of Lactobacillus species tested exerted any anti-infectious activity. In Exp. 2, an anti-infectious activity of Lact. casei ATCC 334T and Lact. zeae ATCC 15820T as well as LcS was shown. Analysis of the concentration of several organic acid molecules in the caecal contents revealed that the acetate concentration in the FOM-treated control group was quite low as less than half the level in the normal mice, reflecting the decrease in total organic acids concentration (Table 2). The acetate concentrations in the LcS, Lact. casei ATCC 334T and Lact. zeae ATCC 15820T-treated groups were maintained at almost normal, and the lactate concentration in the groups increased by more than sevenfold as those of the FOM-control group on day 7 after the DT104 infection. Moreover, pH of the caecal contents in the FOM-control group increased markedly after finishing FOM treatment, while pH was maintained at the normal level in the LcS, Lact. casei ATCC 334T and Lact. zeae ATCC 15820T -treated groups even after FOM treatment. By contrast, no significant anti-infectious activity was shown by Lact. brevis ATCC14869T, Lact. plantarum strain ATCC14917T, Lact. reuteri strain JCM1112T, Lact. rhamnosus ATCC7469T, or Lact. salivarius ATCC11741T (Tables 1 and 2).

Table 1.   Comparison of anti-infectious activity among fosfomycin (FOM)-resistant Lactobacillus strains
Treatment†No. of deaths/total no. of miceNumber of DT104 in organ‡Number of given lactobacilli in caecum‡
SpleenMLNCaecum
MeanSD MeanSD MeanSD MeanSD 
  1. NT, not tested.

  2. *< 0·05.

  3. †Saline (Control) or Lactobacillus strains at a dose of 1–3 × 108 CFU per mouse were orally administered daily to mice from day 1 after starting FOM treatment until the mice died (observation for survival) or to day 6 after infection with DT104 (determination of viable counts of DT104 in organs).

  4. ‡The results are expressed as the mean Log10 CFU and SD per entire organ (spleen and MLN) or g caecal contents.

  5. §Numbers of organs in which DT104 were detected/number of organs tested.

Exp. 1
 Saline (control)8/87·00·8(6/6)§5·20·7(6/6)9·00·4(6/6)   
 Lactobacillus casei strain Shirota3/83·70·6*(3/6)3·80·5*(4/6)3·91·0*(5/6)8·10·6(6/6)
 Lactobacillus plantarum ATCC14917T8/87·10·5(5/6)5·00·4(6/6)9·00·5(6/6)8·50·6(6/6)
 Lactobacillus reuteri JCM1112T8/87·40·5(6/6)5·50·4(6/6)9·20·3(6/6)8·10·4(6/6)
 Lactobacillus rhamnosus ATCC7469T8/87·01·0(6/6)5·00·5(6/6)9·20·3(6/6)8·10·7(6/6)
 Lactobacillus salivarius ATCC11741T8/87·10·5(6/6)5·20·3(6/6)8·51·1(6/6)8·90·5(6/6)
Exp. 2
 Saline (control)8/86·70·5(6/6) NT 8·50·3(6/6)   
 Lact. casei strain Shirota3/83·50·4*(3/6) NT 4·40·8*(4/6)8·00·2(6/6)
 Lact. casei ATCC334T3/83·9* (2/6) NT 4·50·5*(3/6)8·10·1(6/6)
 Lact. plantarum ATCC14917T8/86·70·5(6/6) NT 8·40·6(6/6)8·00·2(6/6)
 Lact. reuteri JCM1112T8/86·60·4(6/6) NT 8·30·5(6/6)8·10·2(6/6)
 Lact. rhamnosus ATCC7469T8/86·60·3(6/6) NT 8·30·5(6/6)8·10·1(6/6)
 Lact. salivarius ATCC11741T8/86·60·3(6/6) NT 8·20·5(6/6)8·10·2(6/6)
 Lactobacillus brevis ATCC14869T8/86·50·4(6/6) NT 8·30·3(6/6)8·10·1(6/6)
 Lactobacillus zeae ATCC15820T3/83·50·4*(3/6) NT 4·80·2*(4/6)8·10·3(6/6)
Table 2.   Changes in intestinal pH and organic acid concentrations after DT104 infection of FOM-treated mice
Treatment†pH‡,§Total organic acids‡,§Acetic acid‡,§Lactic acid‡,§
Day 0Day 7Day 0Day 7Day 0Day 7Day 0Day 7
MeanSDMeanSDMeanSDMeanSDMeanSDMeanSDMeanSDMeanSD
  1. *P < 0·05, **< 0·01.

  2. †Mice were treated as described in Table 1.

  3. ‡pH and organic acid concentrations were determined as described in the Materials and methods.

  4. §The results are expressed as the mean and SD per g caecal contents.

Exp. 1
 Untreated normal6·40·16·40·1891190106196081·01·0‡0·90·8
 Saline (control)7·20·17·20·13643692552451·20·8*1·30·8
 Lactobacillus casei strain Shirota6·50·2**6·60·2**7415 **7913**5812**5712**7·02·2**12·53·4**
 Lactobacillus plantarum ATCC14917T7·00·07·20·147637929727133·10·82·21·2
 Lactobacillus reuteri JCM1112T7·30·27·30·2381342112982570·50·52·82·4
 Lactobacillus rhamnosus ATCC7469T7·10·27·20·2458451629525101·70·55·02·0
 Lactobacillus salivariusATCC1174117·20·17·30·252740123062363·00·95·82·9
Exp. 2
 Untreated normal6·50·16·40·195149710641160101·00·91·10·8
 Saline (control)7·30·17·20·23953782782591·31·01·40·9
 Lact. casei strain Shirota6·70·2**6·60·2**7912**8013**5814**5513**6·22·0*10·93·4**
 Lact. caseiATCC334T6·60·1**6·50·2**8311**8712**6011**5511**8·81·4**12·83·2**
 Lact. plantarum ATCC14917T7·40·27·30·2427401131102571·20·41·80·4
 Lact. reuteri JCM1112T7·10·27·20·248945928725101·80·52·62·0
 Lact. rhamnosusATCC7469T7·20·17·00·347743103082662·00·92·81·9
 Lact. salivariusATCC11741T7·30·27·30·144103783252871·90·93·81·9
 Lactobacillus brevis ATCC14869T7·00·27·10·251746123463363·41·94·52·2
 Lactobacillus zeae ATCC15820T6·60·1**6·70·1**8212**7913**5610**55 7**6·91·9*11·83·9**

Growth-inhibitory activity of acetic acid against DT104 in vitro

Figure 5a clearly shows that acetate at a concentration of 58 mmol l−1 at pH 6·5, which is equivalent to the conditions in the caecum colonized by living LcS, inhibits the in vitro growth of DT104, whereas the combination of acetate at 30 mmol l−1 and pH7·15 corresponding to that in the caecum colonized by ineffective Lact. salivarius had no such effect. Although the increased concentration of lactic acid (7·0 mmol l−1) itself did not result in inhibition of the growth of DT104 (Fig. 5b), the combination of acetate and lactic acid augmented the inhibitory activity against the growth of DT104 (Fig. 5c). Neither such conditions equivalent to those of the Lact. salivarius-colonized caecum exhibited any growth-inhibitory activity against DT104 in vitro.

Figure 5.

 Importance of the intestinal undissociated acetic acid concentration for growth-inhibition of DT104 by lactobacilli. pH, acetic acid (A.A.) concentration and lactic acid (L.A.) concentration were adjusted in the growth medium so that the conditions were the same as in the FOM-treated control caecum (•: pH 7·2, A.A. 25 mmol l−1, L.A. 1·2 mmol l−1), FOM-treated and LcS -colonized caecum (□: pH 6·5, A.A. 58 mmol l−1, L.A. 7·0 mmol l−1) and Lactobacillus salivarius ATCC11741T-colonized caecum (△: pH 7·15, A.A. 30 mmol l−1, L.A. 3·0 mmol l−1). Then, DT104 at a final concentration of 104 CFU ml−1 in each medium was inoculated into the media supplemented with acetate alone (a), lactate alone (b), or a combination of acetate and lactate (c), and after culture at 37°C for 0, 1, 3 or 6 h. Results (Viable DT104 counts) are expressed as the means and SDs of triplicate cultures.

Discussion

Several experimental animal models of infection with antibiotic-resistant pathogens have been reported: infection of mice with vancomycin-resistant Enterococcus during treatment with vancomycin (Donskey et al. 1999) and Pseudomonas aeruginosa infection during treatment with ampicillin (Matsumoto et al. 1999). In both models, pathogens highly resistant to the antibiotic administered were used to induce reproducible opportunistic infections resulting from colonization at high population levels in the intestines during treatment of mice with the antibiotic. In this study, we used a clinical isolate of DT104 that has been confirmed to be resistant to ampicillin, chloramphenicol, streptomycin and tetracycline, but to be sensitive to FOM, and only 105 CFU g−1 were detected in the caecum on day 1 after a challenge infection with the strain at a dose of 109 CFU under feeding with FOM in drinking water. Increases in the population levels of DT104 in the intestines up to 107 CFU g−1 caecal contents by day 7 after infection seemed to be reflected in the increases in the levels of DT104 in the organs such as spleen and MLN, and their reflux to the intestines (Fig. 1). Invasion of DT104 in the caecal epithelium was clearly shown as the images of DT104 proliferating in both the epithelial layer and the DT104-infected epithelial cells being detached in the intestinal lumen were obtained (Fig. 2a). These results might suggest that DT104 cells reflux from the organs into the intestines after extra-intestinal multiplication in the organs but not that pathogens had acquired resistance to FOM in the intestines. We confirmed that the DT104 cells recovered from the caecum on day 7 after infection during FOM treatment was as sensitive as those before the challenge infection (data not shown), suggesting the above hypothesis, and the point of this infection model, therefore, is that DT104 can induce a lethal infection even under the treatment of the host with antibiotics to which DT104 is sensitive.

Daily administration of LcS completely inhibited intestinal colonization by DT104 throughout the experimental period (Figs 1 and 3). The increased concentrations of organic acids, such as acetate and lactate, appeared to be responsible for the anti-infectious activity of LcS (Fig. 4). The undissociated forms of such organic acids have been reported to inhibit the multiplication of acid-sensitive pathogens (Eklund 1983; Brocklehurst and Lund 1990). The calculated undissociated acetic acid (U.A.A.) concentration of LcS-treated group in the intestine should inhibit the growth of DT104 cells, as suggested by the preliminary data in vitro (MIC of U.A.A. for DT104: 3 mmol l−1). Moreover, the heat-killed LcS was not protective against the infection at all (Figs 1–4), suggesting that active metabolism of live LcS in the intestines is required for effective protection. pH of the caecal contents increased markedly as a result of FOM treatment in the control group, while it was maintained at the normal level in the LcS-treated group. It should be noted that such effect was not shown by Lact. brevis, Lact. plantarum, Lact. reuteri, Lact. rhamnosus, or Lact. salivarius-treated irrespective of their high population levels in the intestines (Tables 1 and 2). As we were unable to detect any differences in lactate production activity during in vitro culture in the presence of FOM among the Lactobacillus strains (data not shown), their metabolic activity may be affected by some environmental factors in vivo. In a previous study, we demonstrated marked differences in anti-salmonella activity among different bifidobacterial strains belonging to different species (Asahara et al. 2001b). As LcS itself produces little acetate, the improvement in acetate concentration in the LcS-treated group seems to be attributed to maintenance of the metabolic activity of the indigenous microbiota including certain anaerobes such as Clostridium and Bacteroidaceae to produce acetate.

Recently, the luxS/AI-2 quorum sensing system has been reported to be involved in the Salmonella pathogenicity (Vendeville et al. 2005). Moreover, some species of lactobacilli have also been reported to have the luxS/AI-2 quorum sensing system (Buck et al. 2009; Moslehi-Jenabian et al. 2009). LcS has been confirmed to have the luxS/AI-2 quorum sensing system (data not shown). Studies are in progress to clarify whether the autoinducer produced by LcS acts on the quorum sensing systems of DT104 and might reduce its pathogenicity.

In conclusion, antibiotics such as FOM disrupt the metabolic activity of the intestinal microbiota that produce organic acids, and that only probiotic strains that are metabolically active in vivo should be selected to prevent intestinal infection when used clinically in combination with certain antibiotics.

Acknowledgements

We would like to thank Dr Kazumi Uchida and Yuriko Nagata for contributing the histopathological analyses.

Ancillary