OUP user menu

In vitro investigations of the effect of probiotics and prebiotics on selected human intestinal pathogens

Laura J. Fooks , Glenn R. Gibson
DOI: http://dx.doi.org/10.1111/j.1574-6941.2002.tb00907.x 67-75 First published online: 1 January 2002

Abstract

This study investigated the effects of selected probiotic microorganisms, in combination with prebiotics, on certain human intestinal food-borne pathogens. Probiotics grown with different carbohydrate sources were observed to inhibit growth of Escherichia coli, Campylobacter jejuni and Salmonella enteritidis, with the extent of inhibition varying according to the carbohydrate source provided. Prebiotics identified as being preferentially utilised by the probiotics tested were oligofructose (FOS), inulin, xylo-oligosaccharide (XOS), and mixtures of inulin:FOS (80:20 w/w) and FOS:XOS (50:50 w/w). Two of the probiotics, Lactobacillus plantarum and Bifidobacterium bifidum were selected for further co-culture experiments. Each was combined with the selected prebiotics, and was observed to inhibit pathogen growth strongly. Results suggested that acetate and lactate were directly conferring antagonistic action, which was not necessarily related to a lowering of culture pH.

Keywords
  • Oligosaccharides
  • Lactobacillus plantarum
  • Bifidobacterium bifidum
  • Escherichia coli
  • Campylobacter jejuni
  • Salmonella enteritidis

1 Introduction

The large bowel harbours a nutritionally and physiologically diverse range of bacteria, promoting normal intestinal function, and offering the host protection against infections [1]. Disruption of the colonic flora, due to pathogens [2,3], dietary antigens [1] or other harmful substances [46] can, however, lead to intestinal dysfunction. Lactobacillus and Bifidobacterium, which have a long and safe history in the manufacture of dairy products, are traditionally included in probiotic products to help protect against such effects [7]. Both are thought to prevent the adherence, establishment, replication and/or virulence of specific enteropathogens [8]. A number of mechanisms have been proposed: decreasing the luminal pH via the production of volatile short chain fatty acids (SCFA); rendering specific nutrients unavailable to pathogens; and/or producing specific inhibitory compounds such as bacteriocins [9]. In the normal intestinal flora these mechanisms are essential to the component bacterial populations, as a means of gaining advantage over competing bacteria.

An ability to compete for limiting nutrients is perhaps the most important factor that determines the composition of the gut flora, with species that are unable to compete being effectively eliminated from the system. One of the primary objectives of this research was to identify a synbiotic, that is, a combination of a probiotic and prebiotic [10], which could ultimately be used for fermentation experiments and be an effective antimicrobial agent against common enteropathogens. The range of carbohydrates tested included non-prebiotic controls such as lactitol, starch and dextran, and prebiotics, short and long chain fructo-oligosaccharides, xylo-oligosaccharide (XOS) and lactulose.

In this study, a range of probiotics, were tested using disc/spot assays for their ability to inhibit the growth of some common enteropathogens: Campylobacter jejuni, Escherichia coli and Salmonella enteritidis. Each probiotic strain tested was combined with a range of prebiotic sources in an attempt to move towards identifying synbiotic combinations. Subsequently, two of the probiotic strains, Lactobacillus plantarum and Bifidobacterium bifidum, combined with preferred candidate prebiotics, were examined in co-culture for their ability to inhibit growth of the enteropathogens. The mechanism underlying any antimicrobial activity was addressed by monitoring changes of pH in the culture medium and levels of SCFA at specific time intervals, when samples were removed for enumeration of probiotic and pathogen strains.

2 Materials and methods

2.1 Growth substrates

Oligofructose (FOS) and inulin extracted from chicory root were supplied by Orafti Raffinerie (Tienen, Belgium). Inulin was 94% pure with fructose (∼1%), glucose (∼1%) and sucrose (∼5%) as impurities. FOS was 98% pure with fructose (0.9%), glucose (<0.1%) and sucrose (0.9%) as impurities. XOS 35P was supplied by Suntory Ltd. (Tokyo, Japan). The XOS was obtained by enzymatically treating xylan derived from corncobs. The final product contained 35% w/w oligosaccharides [11]. Growth media and supplements were obtained from Oxoid Ltd., Basingstoke, Hampshire, UK, unless otherwise stated. All other carbohydrates and chemicals were obtained from Sigma Chemicals Ltd., Poole, Dorset, UK. Carbohydrates used were FOS, inulin, inulin:FOS mixture (80:20, w/w), XOS, FOS:XOS mixture (50:50, w/w), lactulose, lactitol, starch and dextran.

2.2 Bacterial strains

Probiotic strains tested were L. plantarum 0407 and Lactobacillus pentosus 905 (St. Ivel European Food, Swindon, UK), Lactobacillus reuteri SD2112 (Biogaia), Lactobacillus acidophilus La5 and B. bifidum Bb12 (Chr. Hansen, Denmark). Enteropathogens used were E. coli NCIMB 9517 (National Collections of Industrial and Marine Bacteria Ltd.), C. jejuni ATCC 11351 (American Type Culture Collection) and S. enteritidis var. danysz NCTC 4444 (National Collection of Type Cultures).

2.3 Bacterial growth conditions

Probiotic strains were grown in basal medium (30 ml) containing, in g l−1, peptone, 2.0; yeast extract, 2.0; NaCl, 0.1; K2HPO4, 0.04; KH2PO4, 0.04; CaCl2.6H20, 0.01; MgSO4.7H20, 0.01; NaHCO3, 2.0; Tween 80, 2 ml; hemin (dissolved in three drops 1 M NaOH), 0.05; vitamin K1, 10 μl; cysteine–HCl, 0.5; bile salts, 0.5 with various carbohydrate sources (1% w/v) anaerobically in an anaerobic cabinet (Don Whitley, Yorkshire, UK; 10:10:80, H2:CO2:N2). E. coli and S. enteritidis were grown aerobically in Mueller–Hinton broth. C. jejuni was grown in Brucella broth (Difco Laboratories, Detroit, MI, USA) supplemented with Campylobacter growth supplement (Oxoid) under microaerobic conditions, achieved using a variable atmosphere incubator (VAIN; Don Whitley, Yorkshire, UK; 4:4:10:82, H2:O2:CO2:N2).

2.4 Disc assay technique

Tryptone soya agar (TSA; 2% w/v; Oxoid Ltd., Basingstoke, UK) (15 ml), prepared and autoclaved according to manufacturer's instructions, was poured into sterile Petri dishes and allowed to set. Overnight cultures (109 cfu ml−1) of each probiotic were centrifuged at 14 000×g for 10 min at 4°C. Sterile filter paper discs (Whatman No. 2, 5 mm) were soaked (10 μl) in one of the culture fractions and laid upon the agar surface. Three fractions were tested. Probiotic cells were prepared by removing supernatant and washing the cells twice with phosphate buffer (pH 7.4; 1 M) before re-suspension in 200 μl phosphate buffer. Supernatant consisted of cell-free extract following centrifugation at 14 000×g for 10 min. The pH of the supernatant from each probiotic culture was recorded. Absence of cells was confirmed by serially diluting and plating the extract; no growth was observed. Neutralised supernatant was prepared by neutralising to pH 7 using 1 M NaOH.

Controls used were basal medium, and acetate and lactate solutions (10 mmol l−1, pH 4.5). Four discs were placed on each plate, one test and three controls. Tryptic soy soft agar (10 ml; 0.75%) inoculated with ∼108 cells (0.5 ml) of the indicator organism (E. coli, S. enteritidis) in the late exponential phase was then poured over the surface. The plates were incubated anaerobically at 37°C for up to 48 h. The extent of inhibition was assessed after incubation by measuring the diameter of the clear zone surrounding each disc. Plates were prepared in triplicate, i.e. three per fraction, for each carbohydrate. The experimental set-up was repeated in triplicate.

2.5 Overlay assay technique

The disc assay technique was modified for testing with C. jejuni since reliable results were not obtained with the above technique, as its growth was restricted to the base of the inoculated agar layer. Consequently, a disc laying on the top agar surface was ineffective and 10 μl of culture fraction was therefore spotted onto 2% (w/v) TSA, allowed to evaporate and overlain with 0.75% (w/v) agar inoculated with 108 cells of C. jejuni.

2.6 Growth curves – probiotic strains

MRS broth (10 ml) was inoculated with one cryogenic bead of each probiotic strain and incubated overnight under anaerobic conditions (oxygen-free nitrogen) at 37°C. Basal medium (50 ml) contained in a 50-ml batch culture fermenter was pre-reduced overnight and 0.5 g (1% w/v) of the selected carbohydrate added. 1 ml of each overnight culture was then inoculated, in duplicate, into 50 ml carbohydrate-containing basal medium and cultures were incubated under the same conditions. The optical density at 650 nm of each culture was determined at hourly intervals for up to 24 h. This was repeated four times and mean values plotted. Subsequently, specific growth rates of the probiotics in each carbohydrate medium were calculated. Exponential phase growth was derived from growth plots of the probiotics (not shown).

2.7 Growth curves – pathogenic strains: carbohydrate effect

Mueller–Hinton broth (10 ml) was inoculated with one cryogenic bead of either E. coli or S. enteritidis and incubated overnight under aerobic conditions at 37°C. Brucella broth (50 ml) supplemented with Campylobacter growth supplement was inoculated with one cryogenic bead of C. jejuni and incubated overnight under microaerobic conditions at 37°C. Basal medium (50 ml), contained in a 50-ml batch culture fermenter, was pre-reduced overnight and 0.5 g (1% w/v) of the selected carbohydrate added. 1 ml of each overnight culture (109–1010 cells ml−1) was then inoculated, in triplicate, into 50 ml carbohydrate-containing basal medium and cultures were incubated under the same conditions for each organism. 1 ml aliquots were removed at hourly intervals and used to prepare a dilution series, which was plated onto a suitable agar medium (Oxoid Manual); MacConkey No. 3 (E. coli), Brilliant green (S. enteritidis) or Campylobacter blood-free specific (CBFS)+CCDA supplement (C. jejuni). Plates were incubated for up to 48 h, when colonies were enumerated for each time point.

2.8 Growth curves – pathogenic strains: pH effect

To assess survival of each pathogenic strain at different initial culture pH values, overnight cultures were prepared as previously described. Mueller–Hinton broth (50 ml) and Brucella supplemented broth (50 ml) were prepared, and the pH adjusted using 4 M HCl to give a range of initial pH values from 1 to 7. These were incubated overnight under aerobic (E. coli and S. enteritidis) or microaerobic (C. jejuni) atmospheric conditions. 1 ml of each overnight culture was then inoculated, in triplicate, into 50 ml of pH-adjusted medium (Mueller–Hinton for E. coli and S. enteritidis and Brucella for C. jejuni) and these were incubated under the same conditions for each organism. 1 ml aliquots were removed at hourly intervals and used to prepare a dilution series, which was plated onto a suitable agar medium; MacConkey No. 3 (E. coli), Brilliant green (S. enteritidis) or CBFS+CCDA supplement (C. jejuni). Plates were incubated for up to 48 h, when colonies were enumerated.

2.9 Co-culture experiments

On the basis of disc/spot assay results, two of the probiotic strains, L. plantarum 0407 and B. bifidum Bb12, were selected for further study because of their greater ability to inhibit pathogenic organisms. Overnight cultures (109 cells ml−1) of each probiotic strain and each enteropathogenic strain were prepared. MRS broth (10 ml) was inoculated with 0.01 g of lyophilised powder of each probiotic strain. The cultures were incubated overnight under anaerobic conditions (10:10:80, H2:CO2:N2) at 37°C. Mueller–Hinton broth (10 ml) was inoculated with one cryogenic bead of either E. coli or S. enteritidis, and incubated overnight under aerobic conditions at 37°C. Brucella broth (50 ml) supplemented with Campylobacter growth supplement (Oxoid) was inoculated with one cryogenic bead of C. jejuni, and incubated overnight under microaerobic conditions at 37°C. Basal medium (50 ml) was pre-reduced overnight and 0.5 g (1% w/v) of the selected carbohydrate added. Carbohydrates used were: starch, FOS P95, XOS 35P, mixtures of inulin:FOS (80:20 w/w) and FOS:XOS (50:50 w/w). 1 ml of each overnight culture (∼108–109 cells) was then inoculated, in triplicate, into 50 ml carbohydrate-containing basal medium and cultures were incubated under anaerobic conditions. Agar plates (Oxoid; Beerens, 1990) for the enumeration of each organism were prepared as appropriate. Growth media used were: Rogosa (L. plantarum), Beerens (B. bifidum), MacConkey No. 3 (E. coli), CBFS+CCDA supplement (C. jejuni) and Brilliant green (S. enteritidis).

Samples were removed at 0, 3, 6, 9 and 24 h. 1 ml was used to prepare a dilution series, which was then plated, in triplicate, onto the appropriate agar. For example, if L. plantarum 0407 and E. coli 9517 were co-cultured, the sample was plated onto Rogosa and MacConkey agar. Plates were incubated for up to 48 h, under appropriate atmospheric conditions (L. plantarum and B. bifidum: anaerobically, E. coli and S. enteritidis: aerobically, C. jejuni: microaerobically) and colonies enumerated. The pH of the culture medium was monitored throughout the fermentation period. Each experiment was repeated in triplicate. A further 1 ml sample was also removed for the analysis of SCFAs (lactate and acetate), according to the method described by Parham and Gibson [12].

2.10 Statistical analyses

Data were statistically analysed using pairwise Student's t-test to evaluate significance of inhibition observed (disc/spot assays) or changes in bacterial numbers (co-culture experiments).

3 Results

3.1 Plate assays

Basal medium gave no inhibition of the enteropathogens tested. Acetate and lactate solutions gave inhibition zones of 10 mm (±1.8 mm) and 7 mm (±0.6 mm) respectively. Use of lactulose, lactitol, starch and dextran as carbohydrate sources was generally ineffective in inducing any inhibitory effects from the probiotic strains.

L. plantarum 0407 and L. pentosus 905, combined with FOS, inulin, XOS, and mixtures of inulin:FOS and FOS:XOS, were effective in inhibiting growth of E. coli and S. enteritidis (Table 1). The cell supernatant (cell-free extract) consistently conferred a significantly greater inhibitory effect than either the cells (P<0.01) or neutralised supernatant (P<0.05) fractions. FOS and FOS:XOS were the only carbohydrate sources where inhibition of C. jejuni was observed with L. plantarum, whilst with L. pentosus, the most effective carbohydrate sources were FOS, inulin, XOS and mixtures thereof. Inhibition was significant for FOS (P<0.01) and FOS:XOS (P<0.05). The improved antimicrobial efficacy of the probiotic with these carbohydrate sources was repeated with L. acidophilus La5 and L. reuteri SD2112 although the magnitude of inhibition was less than that of the other two lactobacilli. In contrast, B. bifidum Bb12 was most effective, combinations with FOS, inulin, XOS and related mixtures imparting high inhibition levels against all three enteropathogens. Again, the cell-free fraction of the Bb12 culture was most effective at inhibiting pathogen growth, which was significant compared to the other cell fractions (P<0.05).

View this table:
1

Plate assay inhibition of E. coli, C. jejuni and S. enteritidis by selected probiotic microorganisms grown in broth culture with a range of carbohydrate sources

L. plantarum 0407 L. pentosus 905 L. acidophilus La5 L. reuteri SD2112 B. bifidum Bb12
CellsS/NN. S/NCellsS/NN. S/NCellsS/NN. S/NCellsS/NN. S/NCellsS/NN. S/N
E. coliFOS2.0 (1.4)8.0* (1.4)2.5 (0.7)3.5 (2.1)5.0 (0.0)4.0 (1.4)4.5 (0.7)0.5 (0.5)0.5 (0.0)16.0* (1.4)7.5 (0.7)
Inulin4.5 (2.1)7.5* (3.5)5.0 (1.4)12.0* (0.0)2.5 (0.7)4.5 (0.7)4.0 (0.0)2.5 (0.7)0.5 (0.5)4.0 (0.0)4.0 (0.0)5.5 (0.7)1.5 (0.7)
XOS5.5 (0.7)9.5 (0.7)0.5 (0.7)9.5* (0.7)2.0 (1.4)5.5(0.7)0.5 (0.5)2.5 (0.7)6.0 (1.4)11.5* (0.7)
FOS:XOS2.5 (0.7)5.5 (0.7)3.5 (3.5)0.5 (0.5)5.5* (2.1)2.0 (0.0)1.0 (0.0)1.5 (0.7)2.0 (0.0)0.5 (0.5)2.5 (2.1)12.5* (3.5)5.5 (0.7)
C. jejuniFOS5.0 (0.0)10.5 (0.7)8.5 (0.7)8.0* (0.0)2.5 (0.7)3.5 (0.7)4.5 (2.1)1.5 (2.1)5.5 (0.7)5.0 (0.0)8.5 (0.7)13.0* (1.4)9.5 (0.7)
FOS:XOS4.5 (0.7)9.0 (0.0)7.0 (0.0)3.5 (0.7)7.0 (0.0)4.5 (0.7)5.5 (0.7)5.5 (0.7)15.5* (0.7)12.5 (0.7)
S. enteritidisFOS4.0 (0.0)6.5 (0.7)2.5 (0.7)2.5 (0.7)3.5 (0.7)3.0 (1.4)3.5 (0.7)6.0 (1.4)3.0 (0.0)1.0 (0.7)2.5 (0.7)0.5 (0.0)6.0 (0.0)6.0 (1.4)2.5 (1.4)
Inulin6.0 (0.0)1.0 (0.0)3.0 (1.4)1.5 (0.7)4.5 (0.7)4.0 (1.4)1.5 (1.4)0.5 (0.0)6.0 (1.4)1.5 (0.7)1.5 (0.7)5.5 (0.7)2.5 (0.7)
Inulin:FOS3.0 (0.0)6.0 (1.4)2.5 (0.7)3.0 (0.0)2.5 (0.7)5.5 (0.7)6.5 (2.1)4.0 (2.8)4.5 (0.7)2.0 (0.0)
XOS6.5* (3.5)3.0 (1.4)7.0 (2.8)3.0 (0.0)2.5 (0.7)2.0 (1.4)4.0 (0.0)1.5 (0.7)2.5 (0.7)7.5 (0.7)4.5 (0.7)
FOS:XOS4.0 (0.0)9.0* (2.8)1.0 (0.0)4.5 (0.7)4.0 (1.4)3.0 (1.4)0.5 (0.3)3.5 (0.7)2.0 (1.4)6.0 (1.4)4.0 (1.4)2.0 (0.0)10.0* (4.2)1.0 (0.0)
  • Cultures were fractioned into cells, supernatant (S/N) and neutralised (pH 7.0) supernatant (N. S/N). Results are diameter of the inhibition zone (mm) surrounding the disc. Results are corrected for the disc size (5 mm) for E. coli and S. enteritidis. Values in parentheses are S.E.M. of nine determinations. Asterisks denote significantly greater inhibitory effect of supernatant compared to cells or neutralised supernatant.

The pH of the cell-free extract was recorded for each probiotic strain grown with the various carbohydrates (Table 2). pH was approximately 7.0 prior to inoculation. In general, the largest decrease in pH was observed when FOS and the mixture of FOS:XOS (50:50) was used as the carbohydrate source, with the variation being strain-dependent.

View this table:
2

pH values of supernatant fraction of the overnight cultures of probiotics grown with various carbohydrate sources

Probiotic strainFOSInIn:FOSXOSFOS:XOSLactuloseLactitolStarchDextran
L. plantarum 04074.29 (±0.05)4.52 (±0.04)5.00 (±0.03)4.92 (±0.02)4.78 (±0.03)5.98 (±0.04)6.55 (±0.05)6.56 (±0.04)6.73 (±0.07)
L. pentosus 9054.35 (±0.08)4.77 (±0.05)5.02 (±0.03)4.83 (±0.09)4.72 (±0.05)6.14 (±0.05)6.39 (±0.04)6.54 (±0.06)6.84 (±0.03)
L. acidophilus La54.31 (±0.06)5.10 (±0.02)5.38 (±0.04)5.15 (±0.03)4.63 (±0.07)6.27 (±0.04)6.09 (±0.05)6.33 (±0.08)6.41 (±0.05)
L. reuteri SD21124.54 (±0.03)5.26 (±0.04)5.24 (±0.08)5.85 (±0.06)4.89 (±0.05)6.23 (±0.07)6.86 (±0.02)6.78 (±0.04)6.67 (±0.04)
B. bifidum Bb124.27 (±0.02)4.49 (±0.06)5.72 (±0.06)4.47 (±0.03)4.17 (±0.03)6.08 (±0.03)6.11 (±0.04)6.05 (±0.01)6.18 (±0.03)
  • Values in parentheses are average±S.E.M. of triplicate determinations. In: Inulin; In:FOS, mixture 80:20 w/w; FOS:XOS, mixture 50:50 w/w.

3.2 Probiotic growth curves

Highest growth rates of the lactobacilli were obtained in media that contained glucose as the carbon and energy source. For the bifidobacterial strain, the mixture of FOS:XOS gave optimal growth. L. plantarum 0407 and B. bifidum Bb12 generally grew faster than the other strains, regardless of the carbohydrate source used. In general, the oligosaccharides FOS, XOS or their mixtures were fermented preferably to all other carbohydrate sources (Fig. 1).

1

Growth rates (h−1) of probiotic and enteropathogenic bacteria utilising various carbohydrate sources. In, Inulin; In:FOS, mixture 80:20 w/w; FOS:XOS, mixture 50:50 w/w

3.3 Enteropathogen growth curves

Each of the three enteropathogens was tested for its ability to grow with various carbohydrate sources (Fig. 1). E. coli grew well on glucose. S. enteritidis also utilised glucose effectively, although growth rate was approximately half that observed for E. coli. C. jejuni failed to grow on any of the carbohydrates. The effect of pH on growth of the pathogens was also investigated. Initial medium pH of 3 or below completely inhibited growth of E. coli (Fig. 2a) and of C. jejuni (Fig. 2b) and S. enteritidis (Fig. 2c) at initial pH values of 4 and below.

2

Survival of (a) E. coli, (b) C. jejuni and (c) S. enteritidis at different initial culture pH values. Samples were removed at hourly intervals and enumerated using on selected agars. Data are averaged (±S.E.M.) from triplicate determinations.

3.4 Co-culture investigation

Both L. plantarum 0407 (Table 3) and B. bifidum Bb12 (Table 4) survived in all culture conditions tested such that, generally, their numbers after 24 h fermentation were maintained. B. bifidum numbers were enhanced by 1-log value in the presence of FOS and up to 2 logs when a FOS:XOS (50:50) mixture was used as the carbohydrate source. An increase in probiotic numbers after 24 h fermentation was observed when both L. plantarum and B. bifidum were grown with C. jejuni, again regardless of the carbohydrate source used.

View this table:
3

Inhibition of enteropathogens by probiotic L. plantarum 0407 in co-culture experiments

Carbohydrate sourceChange in numbers, pH or concentration (0-24 h)
StarchFOSInulin:FOSFOS:XOSXOS
E. coliProbiotic numbers+1.62×109 (±1.33×109)+2.14×109 (±2.00×109)+1.35×1010 (±2.83×109)+1.40×108 (±2.02×107)+9.40×108 (±2.11×109)
Pathogen numbers+4.50×109 (±4.33×108)−8.25×108** (±2.83×107)+3.27×109 (±1.31×109)−2.37×107 (±6.66×107)−2.25×108 (±1.98×107)
Culture pH+0.03 (±0.04)−1.15 (±0.06)−1.02 (±0.03)−1.00 (±0.24)−0.94 (±0.22)
[Acetate] mM+2.32 (±0.33)+2.24 (±0.18)+2.35 (±0.47)+1.23 (±0.04)+1.20 (±0.09)
[Lactate] mM+16.64 (±2.12)+25.38 (±0.55)+9.77 (±2.02)+23.06 (±0.51)+17.62 (±0.75)
C. jejuniProbiotic numbers+9.98×1012 (±0.00)+1.00×1013 (±0.00)+1.00×1013 (±0.00)+1.00×1013 (±0.00)+1.00×1013 (±0.00)
Pathogen numbers−1.29×107 (±1.78×105)−2.09×109*** (±0.00)−1.74×109 (±0.00)−1.30×109 (±0.00)−1.89×109 (±0.00)
Culture pH−0.41 (±.021)−2.20 (±0.06)−1.99 (±0.14)−2.33 (±0.25)−2.13 (±0.18)
[Acetate] mM+2.73 (±0.33)+2.26 (±0.13)+2.27 (±0.11)+1.87 (±0.30)+1.92 (±0.05)
[Lactate] mM+17.39 (±2.86)+20.17 (±0.41)+11.37 (±1.12)+21.27 (±0.89)+20.68 (±1.29)
S. enteritidisProbiotic numbers+1.21×1010 (±7.36×109)+2.15×1010 (±1.70×1010)+2.20×1010 (±1.48×1010)+1.76×108 (±5.65×107)+6.10×109 (±9.19×108)
Pathogen numbers+7.42×1010 (±8.08×109)−7.25×1010*** (±0.00)−8.08×108 (±2.09×102)−2.68×107 (±1.04×107)−7.15×106 (±2.26×106)
Culture pH−0.35 (±0.04)+1.79 (±0.04)−1.55 (±0.04)−1.87 (±0.08)−1.43 (±0.03)
[Acetate] mM+2.58 (±0.43)+0.94 (±0.10)+1.52 (±0.22)+2.04 (±0.16)+1.81 (±0.05)
[Lactate] mM+11.55 (±5.36)+20.89 (±0.70)+11.06 (±1.62)+20.74 (±0.38)+13.46 (±0.33)
  • Data presented are changes in parameter between 0 and 24 h inoculation. Probiotic and enteropathogen numbers were enumerated at various time points after inoculation. Results are cfu ml−1 (±S.E.M.) averaged from six determinations. For culture pH and acetate and lactate concentrations (mM), 1-ml samples were removed at specified intervals and parameters determined. Results are average (±S.E.M.) of triplicate determinations. + and − denote an increase or decrease respectively of the described parameter. A significant decrease of pathogen numbers from baseline (0 h fermentation) is denoted; **P<0.01, ***P<0.001.

View this table:
4

Inhibition of enteropathogens by probiotic B. bifidum Bb12 in co-culture experiments

Change in numbers, pH or concentration (0–24 h)
Carbohydrate source
StarchFOSInulin:FOSFOS:XOSXOS
E. coliProbiotic numbers+1.69×107 (±5.99×107)+1.89×109 (±1.20×109)+5.27×107 (±3.09×107)+1.26×109 (±2.18×108)+2.02×108 (±6.78×107)
Pathogen numbers−2.56×108 (±2.61×108)−1.92×107 (±6.46×106)−4.80×107 (±4.89×107)−1.25×106 (±6.13×105)−6.68×107 (±4.08×107)
Culture pH+0.16 (±0.13)−1.46 (±0.09)−1.02 (±0.52)−0.77 (±0.12)−0.97 (±0.21)
[Acetate] mM+14.02 (±0.57)+28.39 (±1.32)+22.70 (±1.07)+14.63 (±0.47)+11.15 (±1.18)
[Lactate] mM+9.05 (±1.34)+12.64 (±0.72)+14.52 (±0.53)+29.11 (±2.38)+18.29 (±1.40)
C. jejuniProbiotic numbers+1.00×1013 (±1.11×105)+1.00×1013 (±0.00)+1.00×1013 (±0.00)+1.00×1013 (±0.00)+1.00×1013 (±0.00)
Pathogen numbers−1.26×108 (±4.92×107)−2.29×108** (±1.24×109)−2.38×108** (±0.00)−2.77×108*** (±0.00)−2.07×108 (±5.03×108)
Culture pH−0.49 (±0.73)−2.20 (±0.18)−2.06 (±0.19)−2.43 (±0.26)−1.97 (±0.51)
[Acetate] mM+10.82 (±1.85)+25.82 (±0.97)+21.71 (±2.05)+26.19 (±1.04)+15.19 (±2.49)
[Lactate] mM+10.93 (±0.92)+12.80 (±1.92)+14.23 (±2.00)+16.50 (±1.28)+11.01 (±1.27)
S. enteritidisProbiotic numbers+3.68×108 (±0.00)+1.42×109 (±3.91×108)+8.84×107 (±5.95×107)+8.75×108 (±2.07×108)+1.94×107 (±9.00×106)
Pathogen numbers−1.66×108 (±1.11×105)−1.36×109 (±0.00)−3.19×107 (±1.99×107)−1.70×109*** (±0.00)−3.05×108 (±2.10×108)
Culture pH+0.09 (±0.16)−1.59 (±0.08)−1.38 (±0.64)−1.37 (±0.13)−0.96 (±0.12)
[Acetate] mM+10.65 (±1.25)+21.55 (±1.97)+22.98 (±1.23)+26.26 (±3.03)+12.81 (±3.98)
[Lactate] mM+11.40 (±1.50)+13.08 (±0.25)+17.12 (±2.25)+17.45 (±3.20)+11.49 (±1.62)
  • Data presented are changes in parameter between 0 and 24 h inoculation. Probiotic and enteropathogen numbers were enumerated at various time points after inoculation. Results are cfu ml−1 (±S.E.M.) averaged from six determinations. For culture pH and acetate and lactate concentrations (mM), 1-ml samples were removed at specified intervals and parameters determined. Results are average (±S.E.M.) of triplicate determinations. + and − denote an increase or decrease respectively of the described parameter. A significant decrease of pathogen numbers from baseline (0 h fermentation) is denoted; **P<0.01, ***P<0.001.

L. plantarum combined with FOS was the most effective at inhibiting pathogen growth (Table 3). A significant, 6-log (P<0.01) decrease in E. coli numbers was observed when FOS was used, whilst after the same time period, C. jejuni and S. enteritidis were undetectable (P<0.001). B. bifidum combined with FOS:XOS proved an effective synbiotic combination (Table 4). C. jejuni and S. enteritidis were decreased to below detectable levels (P<0.001), whilst a 2-log decrease in E. coli numbers was observed.

Changes in pH during the 24-h fermentation were observed but were not significant for either L. plantarum 0407 (Table 3) or B. bifidum Bb12 (Table 4). No definitive correlation could be made between the observed decrease in pH and the extent of inhibition. Culture pH decrease was generally observed after fermentation for 3 h, yet a lowering of pathogen numbers was not observed until fermentation for 9–24 h. Furthermore, when B. bifidum Bb12 was co-cultured with E. coli in the presence of XOS, a pH decrease was observed, even though there was no evidence of antimicrobial activity.

With L. plantarum, the final concentration of lactate varied depending on the carbohydrate source provided, ranging from 12–17 mM with starch and inulin:FOS, to 20–25 mM with FOS and FOS:XOS (Table 3). For B. bifidum, acetate concentrations were consistently higher than levels of lactate by the end of the fermentation period (Table 4). Both FOS and FOS:XOS gave rise to higher levels of acetate, reaching approximately 20–30 mM 24 h after inoculation.

4 Discussion

All of the probiotic strains tested using plate assays offered some inhibition of each of the pathogenic strains, E. coli, S. enteritidis and C. jejuni. The extent of inhibition was dependent on the probiotic strain, such that L. plantarum 0407 and B. bifidum Bb12 tended to inhibit pathogen growth to a greater extent than that observed for the other strains included, particularly with E. coli. These results, in this respect, compound the findings of a number of investigations, which used similar pure culture in vitro methodologies to establish the antimicrobial potential of lactobacilli and bifidobacteria. Jacobsen et al. [13] used the spot assay to examine the capabilities of 47 strains of Lactobacillus spp. to inhibit a range of pathogenic organisms, including E. coli and Salmonella typhimurium. Twenty eight of the strains tested inhibited E. coli growth, whilst 12 of the 47 strains tested inhibited growth of S. typhimurium. A plethora of information surrounds the proposed probiotic activity of L. reuteri[14], primarily ascribed to production of an antimicrobial protein called reuterin [15,16], whose synthesis requires glycerol in the growth medium [17]. This was omitted from media used in the current investigation, since it is not known how much glycerol would be present in the gut, to become available for utilisation by L. reuteri, in the production of reuterin [14]. Six B. bifidum strains, tested using well diffusion assays, inhibited growth of E. coli and to a lesser extent, Salmonella typhosa[18,19]. Gibson and Wang [20] reported the strain-dependent ability of eight different species of bifidobacteria, including B. bifidum, to inhibit the growth of a range of pathogenic bacteria, including E. coli, a Salmonella spp., and a Campylobacter spp.

The antimicrobial potential exhibited by each of the probiotics used here appeared to depend on the carbohydrate source used. FOS, inulin, XOS, and mixtures of FOS:XOS (50:50 w/w) and inulin:FOS (80:20 w/w) all caused greater inhibition than lactulose, lactitol, starch and dextran, perhaps suggesting a structure-to-function relationship in terms of the prebiotic used. The type of bond linking the component monomers, in view of specific cleavage enzymes being required for fermentation of the carbohydrate, may effect fermentation rate, and thereby determine the speed at which potential inhibitory metabolic end products are released. Chain length of the carbohydrate is also likely to be a contributory factor, since long chain oligosaccharides, with multiple branching, require more enzymatic hydrolysis by the organisms before its complete fermentation.

The major metabolic end products of lactobacilli and bifidobacterial fermentations are acetate and lactate [21], leading to a decrease in the culture pH. In both plate assay and co-culture experiments, the pH of the cell-free extract was lower when FOS, inulin and XOS, and mixtures thereof were provided as the carbohydrate sources. The extent of inhibition with these prebiotic carbohydrates was often correspondingly increased, suggesting that a possible mechanism of antimicrobial action may be attributable to the low culture pH. However, since some results contradicted this finding, it cannot be definitively stated that low pH acts as the principal inhibitory parameter. The enhanced antimicrobial effect of FOS and FOS:XOS may be attributed to the action of the synbiotic since numbers of the pathogen decreased in co-cultures, but not in mono-cultures of the pathogen. Confirmatory evidence of the inability of E. coli, C. jejuni and S. enteritidis to withstand an acidic environment was obtained (Fig. 2) where these bacteria failed to proliferate at the pH environment (pH≤5.0) normally induced by bifidobacterial and lactobacilli fermentation. Similar findings have been observed for B. bifidum 1452 [19], an L. plantarum strain [22] and a bifidobacterial strain B. longum[23], where a decrease in pH during incubation time increased the antagonistic properties against E. coli, Staphylococcus aureus, Klebsiella pneumoniae, and a number of clostridia and bacteroides species, including Clostridium perfringens and Bacteroides fragilis.

The toxicity of fermentation acids at a low pH has been traditionally explained by the transmembrane flux of undissociated acids, dissociation of the acids in the more alkaline cytoplasm, and metabolic uncoupling [24]. However, Russell and Diez-Gonzalez [25] suggested that the mechanistic explanation lies with the pH gradient-mediated anion accumulation within the bacterial cytoplasm. Fermentation acid dissociation in the more alkaline interior causes an accumulation of the anionic species, and this accumulation is dependent on a pH gradient across the membrane. Where culture pH of the cell-free extract decreased to pH 4.17, when short chain carbohydrates had been utilised, this would theoretically induce a considerable pH gradient across the pathogenic organism membrane, thus allowing accumulation of the fermentation acids in the cytoplasm. Furthermore, since the cell-free extract promoted the greatest bactericidal effect, it can be assumed that the fermentation acids accumulated in this fraction.

In conclusion, this study has shown that lactobacilli and bifidobacteria species can inhibit some important pathogenic species. This antagonism was influenced by the carbohydrate provided in vitro. The inhibitory mechanism underlying the effect has been initially addressed, and a strong case has been presented for the production of SCFA as the underlying mechanism of inhibition of enteropathogens.

Acknowledgements

L.J.F. was a recipient of a PhD studentship from St. Ivel European Food, Wootton Bassett, UK.

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
View Abstract