Relative predominance of each of five probiotic strains was investigated in the ileum of weaned pigs, compared with that in feces, when administered in combination at c. 5 × 109 CFU day−1 for 28 days. Probiotic was excreted at 106–109 CFU g−1 feces, while ileal survival ranged from 102 to 106 CFU g−1 digesta. In contrast to the feces, where Lactobacillus murinus DPC6002 predominated, the bacteriocin-producing Lactobacillus salivarus DPC6005 dominated over coadministered strains both in the ileum digesta and in mucosa. Probiotic administration did not alter counts of culturable fecal Lactobacillus or Enterobacteriaceae but higher ileal Enterobacteriaceae were observed in the ileal digesta of probiotic-fed pigs (P<0.05). We observed decreased CD25 induction on T cells and monocytes (P<0.01) and decreased CTLA-4 induction (P<0.05) by the mitogen phytohemagglutinin on CD4 T cells from the probiotic group. Probiotic treatment also increased the proportion of CD4+ CD8+ T cells within the peripheral T-cell population and increased ileal IL-8 mRNA expression (P<0.05). In conclusion, superior ileal survival of L. salivarius compared with the other coadministered probiotics may be due to a competitive advantage conferred by its bacteriocin. The findings also suggest that the five-strain combination may function as a probiotic, at least in part, via immunomodulation.
Traditionally, subtherapeutic doses of antibiotics have been used as a management tool to help maintain the health of pigs at weaning (Dritz et al., 2002; Gaskins et al., 2002). However, the emergence of antibiotic-resistant pathogens coupled with the European ban on the growth promotional use of antibiotics has created a need for development of alternatives to antibiotics (Anadon, 2006). Probiotics (live microorganisms, which when administered in adequate amounts confer a health benefit on the host; Pineiro & Stanton, 2007) are one such alternative. Cultures most commonly used in swine are lactic acid bacteria (mainly lactobacilli and enterococci) and also Bacillus and yeasts (Simon et al., 2003). Improvements in growth performance, decreased incidence of diarrhea and reduced mortality have previously been reported for probiotics, when used as antibiotic alternatives in pigs (Simon et al., 2003; Lalles et al., 2007, for review). Probiotics may also reduce the carriage of enteric pathogens, such as Salmonella, thereby offering a means of improving preharvest food safety (Callaway et al., 2004).
While the precise mechanistic basis of the beneficial effects of probiotics is unknown and will most likely vary depending on the strain and species used, a number of mechanisms have been suggested. Modulation of the host immune response is one, and immunomodulatory effects have been demonstrated in humans and laboratory animals (Corthésy et al., 2007); however, data on the effects of probiotics on porcine immune function are limited. Another proposed mechanism is the prevention of intestinal pathogen colonization through competitive exclusion and/or the synthesis of inhibitory compounds, such as organic acids, hydrogen peroxide or bacteriocins (Simon et al., 2003). In fact, the production of antimicrobials, in particular bacteriocins, is often cited as a desirable trait for candidate probiotic strains (O'Connor et al., 2005). A number of studies have demonstrated the efficacy of bacteriocin-producing cultures, both in preventing enteric infections and in reducing pathogen shedding in livestock (Diez-Gonzalez, 2007; for review). However, more compelling evidence for the role of bacteriocins in mediating probiotic effects is provided by a recent study in which Corr et al. (2007) demonstrated that the antilisterial effect of a Lactobacillus salivarus probiotic in mice was attributed to the Abp 118 bacteriocin produced by the strain. Bacteriocin production may also be considered a probiotic trait in terms of aiding survival of orally administered cultures within the gastrointestinal tract (GIT), because it potentially gives the producing organisms a competitive advantage over the resident microbial communities (O'Connor et al., 2005). However, the interaction of bacteriocin-producing cultures with coadministered probiotics and/or with commensal gut microbiota has not been a focus of research to date.
A five-strain Lactobacillus/Pediococcus probiotic developed in our laboratory (Casey et al., 2004; Gardiner et al., 2004) has proven effective in reducing Salmonella shedding in pigs (Casey et al., 2007). Each of the five probiotic strains has been characterized extensively in vitro (Casey et al., 2004) and one of the components, a L. salivarius strain, produces a bacteriocin, salivaricin P (characterized by Barrett et al., 2007), to which three of the other component strains (two Lactobacillus murinus and a Pediococcus pentosaceus) are sensitive. The objective of the present study was to evaluate the relative performance of each of the five probiotic strains in the ileum of weaned pigs, compared with that in feces, when administered in combination. The ileum is the main site of Salmonella invasion in the porcine GIT, and as such is important when considering the effects of anti-Salmonella probiotics. The ability of the probiotic to impact on some immune parameters was also investigated. The overall rationale, therefore, was to provide more insight into the functionality of this five-strain combination as a probiotic for pigs and to investigate whether bacteriocin production can offer a competitive advantage to probiotics in vivo.
Materials and methods
Bacterial strains and culture conditions
Lactobacillus murinus DPC6002 and DPC6003, Lactobacillus pentosus DPC6004, L. salivarius DPC6005, and P. pentosaceus DPC6006 were previously isolated from pig cecal contents and characterized extensively with respect to their probiotic properties (Casey et al., 2004, 2007; Gardiner et al., 2004). These strains are rifampin-resistant (Rifr) variants to facilitate enumeration in the porcine GIT. They were routinely cultured at 37 °C in de Man Rogosa and Sharpe broth (MRS; Difco Laboratories, Detroit, MI) in anaerobic jars with CO2-generating kits (Anaerocult A; Merck, Darmstadt, Germany). A probiotic skim milk suspension containing these five strains was prepared for administration to pigs as described by Casey et al. (2007), with the following modifications: after incubation, centrifugation and washing, each of the five strains was resuspended in 200 mL sterile 10% (w/v) reconstituted skim milk plus 0.5% (w/v) yeast extract (RSMYE) and combined in equal proportions. The probiotic suspension was then stored at 4 °C until it was fed to the pigs (within 1 h of preparation). Throughout the pig-feeding trial, bacterial numbers in this probiotic suspension were checked at regular intervals by plating on MRS agar, as described previously (Gardiner et al., 2004).
The pig-feeding trial complied with European Union Council Directive 91/630/EEC (outlines minimum standards for the protection of pigs) and European Union Council Directive 98/58/EC (concerns the protection of animals kept for farming purposes) and was approved by, and a license obtained from, the Irish Department of Health and Children. A total of 24 crossbred (Large White × Landrace) pigs (12 male and 12 female) were weaned at c. 26 days of age. For the first 7 days postweaning pigs were fed a commercial starter diet (16.25 MJ digestible energy kg−1 and 16.0 g kg−1 lysine) containing 155 mg kg−1 copper and 2500 mg kg−1 zinc to ease the weaning transition and avoid disease outbreak. This was followed by a 7-day acclimatization period, during which 100 mL of sterile RSMYE was fed daily to each pig in addition to a basal diet, which was formulated to contain 14.5 MJ kg−1 digestible energy and 12.9 g kg−1 total lysine. This basal diet did not contain medication but contained 175 mg kg−1 copper and 120 mg kg−1 zinc. At 14 days postweaning (day 0) pigs were blocked by sex, weight (mean body weight=11.9 kg) and ancestry. One pig from each block (eight pigs in total) was harvested at day 0 for determination of baseline immunological parameters. The remaining 16 pigs were assigned to one of two treatment groups (control or probiotic, each containing eight pigs) for the 28-day treatment period (day 0–28). Pigs were individually penned and each treatment group was housed in separate but identical rooms to prevent cross-contamination with the probiotic. During the treatment period, in addition to the basal diet, each pig in the probiotic group received 100 mL probiotic skim milk containing c. 5 × 107 CFU mL−1 daily (prepared as outlined above; providing a total dose of c. 5 × 109 CFU probiotic day−1), while pigs in the control group received 100 mL sterile RSMYE daily as well as basal diet. At all times during the trial, pigs had ad libitum access to fresh water and the relevant diet.
Fecal, intestinal and blood sampling
On day 0, the eight untreated pigs were slaughtered by captive bolt stunning followed by exsanguination and the entire GIT was removed. Ileal mucosal scrapings were taken 15 cm before the ileo–cecal junction, immediately snap frozen in liquid nitrogen and stored at −80 °C for baseline cytokine gene expression analysis. Fecal samples were collected in sterile containers following rectal stimulation from each of the remaining 16 pigs (n=8 per treatment) before (day 0) and during (days 4, 13 and 27) probiotic/control treatment where day 0 was the first day of administration. Fecal samples were stored at 4 °C until microbiological analysis was performed (on the day of collection, as outlined below). On day 28, these 16 pigs were harvested, as outlined above, and the entire GIT was removed. Immediately postharvest, ileal mucosal scrapings were sampled and stored as above. In addition, 30 cm before the ileo–cecal junction both ileal digesta and tissue were sampled for microbiological analysis. These samples were placed on ice for transport to the laboratory where they were stored at 4 °C until analysis (on the day of collection, as outlined below). Whole blood samples were taken from the anterior vena cava of each pig and collected in heparinized blood collection tubes (BD Vacutainer Systems, Franklin Lakes, NJ) on days 0, 14 and 28. Samples were stored at room temperature and peripheral blood mononuclear cells (PBMCs) were isolated and assayed within 30 h, as outlined below.
Microbiological analysis of pig fecal and ileal samples
Total Lactobacillus and Enterobacteriaceae were enumerated in ileal digesta and fecal samples as described by Gardiner et al. (2004). A total probiotic count was obtained from both fecal samples and ileal digesta by spread-plating on LBS–RIF agar, i.e. Lactobacillus selective agar (LBS; Becton Dickinson, Cockeysville, MD) containing 100 μg rifampin mL−1 as a selective agent and 50 U nystatin (Sigma-Aldrich, St Louis, MO) mL−1 to inhibit yeasts and molds. These plates were incubated anaerobically at 37 °C for 2 days. Ileal tissue samples were rinsed gently in maximum recovery diluent (MRD; Bectin Dickinson, Franklin Lakes, NJ) to remove digesta and were further washed by immersing in MRD and shaking vigorously for 5 min. Tissue samples were then homogenized in fresh MRD as 10-fold dilutions using a stomacher (Seward, London, UK). The resulting homogenate was further diluted 10-fold in MRD and appropriate dilutions were spread-plated on LBS–RIF agar to enumerate adherent probiotic.
Representative colonies (20 in total or as many as were present on plates with low bacterial counts) were randomly selected from LBS–RIF plates from each probiotic-fed animal on day 13 (fecal samples) and day 28 (ileal digesta and tissue). Colonies were cultured in MRS broth, and fingerprinted by randomly amplified polymorphic DNA PCR (RAPD PCR) using the R2 primer, as described previously (Gardiner et al., 2004), to determine the relative predominance of each of the administered probiotic strains.
Determination of cytokine mRNA expression by quantitative real-time PCR (qPCR) analysis
Ileal mucosal scrapings (50–100 mg) were placed in 1 mL volumes of Tripure Isolation Reagent (Roche Diagnostics, Penzberg, Germany) and homogenized using a tissue homogenizer (T25 Ultra Turrax, IKA, Staufen, Germany). Total RNA was isolated according to the manufacturer's recommendations (Roche Diagnostics) and suspended in 50 μL of nuclease-free water (Chemicon, Temecula, CA). Total RNA was quantified using a spectrophotometer (Cary 100 Bio; Varian, Palo Alto, CA) at OD260 nm and purity was assessed by determining the OD260 nm : OD280 nm ratio. The purity and integrity of the total RNA was verified by visualization of the 18S and 28S bands with glyoxol dye (Ambion, Austin, TX) following electrophoresis through a 1.5% agarose gel. One microgram of total RNA was used for reverse transcription reactions (total volume 20 μL) using the QuantiTect Reverse Transcription Kit (Qiagen, Valencia, CA). The resultant single-stranded cDNA was then used for evaluation of gene expression of proinflammatory [IL-8, IL-1β, tumor necrosis factor α (TNFα), IL-6] and anti-inflammatory (IL-10) cytokines. Quantitative analysis of these genes was performed in a LightCycler instrument (Roche Diagnostics) using a dilution series of external plasmid DNA standards (Pfaffl, 2001). These plasmid standards were created by cloning gene-specific cDNA PCR products using a TOPO TA Cloning kit (Invitrogen, Life Technologies, Carlsbad, CA). One microliter of each cDNA sample was used per 10 μL LightCycler reaction. The LightCycler FastStart DNA Master SYBR Green I kit (Roche Diagnostics) was used for quantification according to the manufacturer's instructions using 3 mM MgCl2 and 0.5 μM of each primer. The PCR primers were either designed for this study or previously published, as outlined in Table 1. The real-time PCR program began with initial denaturation at 95 °C for 10 min, followed by 50 cycles of quantification consisting of 10 s denaturation at 95 °C, annealing at primer-determined temperature (Table 1) for 4 s followed by 7 s elongation at 72 °C. The specificity of each of the PCR primer sets was determined before their use and analysed following every qPCR reaction by melting curve analysis. This was performed on each product by heating from 5 °C above the annealing temperature for 30 s to 95 °C in the continuous fluorescence acquisition mode. A specific amplicon is defined by a single peak at a discrete melting temperature. For each set of primers used, the single amplicon generated on melting curve analysis was sequenced to ensure that the correct gene segment was amplified. Subsequently all results were reviewed using melting curve analysis, to ensure that single peaks were obtained for all reactions and that the temperatures at which specific amplicons generated their specific peak remained constant.
Table 1. Primers used for quantitative real-time PCR (qPCR) analysis
Isolation and stimulation of PBMCs, immune cell phenotyping and cytokine measurement
Density gradient centrifugation using Lymphoprep™ (Nycomed Pharma, Langebjerg, Denmark) was employed to isolate PBMCs from heparinized blood. PBMCs were resuspended in complete medium [IMDM (Gibco, Invitrogen, Paisley, UK) containing 20% (v/v) heat-inactivated fetal calf serum, 100 U mL−1 penicillin and 100 μg mL−1 streptomycin]. Stimulation of PBMCs was performed with PBS, 25 ng mL−1 lipopolysaccharide (Sigma) or 2 μg mL−1 phytohemagglutinin (Sigma) for 18 h at 37 °C in a 5% (v/v) CO2 humidified atmosphere. Following stimulation, the cell culture supernatant was collected and stored at −80 °C. Concentrations of IL-8, IL-1β, and IL-10 were subsequently determined in these supernatants using porcine-specific cytokine ELISA kits (R&D Systems, Minneapolis, MN) in accordance with the manufacturer's instructions. In addition, PBMCs were analysed by flow cytometry using a BD FACSCalibur™. Antibodies used were antiporcine CD3 fluorescein isothiocyanate (FITC), antiporcine CD3, antiporcine CD4 FITC, antiporcine CD8 phycoerythrin (PE), antiporcine CTLA-4 PE/Cy5, antiporcine CD25, antiporcine CD14 (VMRD Inc., Pullman, WA), antimouse IgE PE/Cy5, and corresponding isotype-matched control antibodies (all antibodies obtained from BD Biosciences, Devon, UK, unless otherwise stated). Antibodies were used according to the manufacturers' instructions.
All data were analysed as a complete randomized block design using the GLM procedure of SAS (2000) (SAS Inst. Inc., Cary, NC). For all response criteria, a single pig was the experimental unit. Treatment effect was tested against residual error term with initial bodyweight used as a blocking factor. To examine changes in fecal excretion over time, the model included treatment, replicate, and time and treatment by time interaction in a factorial analysis. Microbiology data were log-transformed before analysis to ensure data points were normally distributed.
Fecal excretion of administered probiotic strains
By day 4, all of the probiotic-fed pigs excreted between 2.1 × 106 and 1.5 × 109 CFU probiotic g−1 of feces and continued to shed between 5.5 × 104 and 4.6 × 107 CFU g−1 for the 28-day treatment period (Fig. 1a). In fact, mean total fecal excretion of the administered five-strain probiotic during the entire 28-day treatment period was 1.5 × 107 CFU g−1. When examined at each time-point, mean fecal probiotic excretion was highest in the probiotic-treated animals on day 4 of administration (6.4 × 107 CFU g−1 feces) and this declined progressively over the 28-day trial period to reach 1.9 × 107 and 2.9 × 106 CFU g−1 feces on days 13 and 27, respectively. Indeed, the mean day 27 fecal probiotic count was significantly lower (P<0.05) than mean counts obtained on day 4 or day 13 (Fig. 1a). This is also reflected by the finding that probiotic counts constituted on average 15.3% (0.9–34.4%) of the total fecal Lactobacillus population at day 13 but only 2.4% (0.04–11.8%) on day 27.
RAPD PCR fingerprinting on colonies selected from fecal samples taken on day 13 from probiotic-fed pigs confirmed that the majority of the rifampin-resistant isolates excreted were one or other of the administered probiotic strains (Fig. 1b and c). However, the strain composition varied between animals. In seven of the eight pigs sampled, L. murinus DPC6002 was the predominant probiotic strain recovered (Fig. 1b). Among the representative probiotic colonies analysed, L. pentosus DPC6004 was identified in six of the eight animals, while L. murinus DPC6003 was detected in three pigs and L. salivarius DPC6005 was found in only one animal. Pediococcus pentosaceus DPC6006 was not identified within the predominant feces-derived probiotic colonies analysed from any of the eight pigs sampled on day 13 (Fig. 1b).
Survival of administered probiotic strains in the ileum
The presence of the administered probiotic cultures was examined in the ileal digesta sampled on day 28 from all eight pigs fed probiotic. Total probiotic counts ranged from 6 × 102 to 8.6 × 106 CFU g−1 digesta and averaged at 1.3 × 105 CFU g−1 (Fig. 2a). This constituted on average 8% of the total ileal Lactobacillus population and ranged from 0.002% to 34%, depending on the animal. Lactobacillus salivarius DPC6005 was detected in the ileal digesta of all eight pigs and was the predominant probiotic isolate found in five of these pigs (Fig. 2b). Lactobacillus murinus DPC6002 was detected in six pigs, and was the dominant probiotic strain detected in three pigs, while L. murinus DPC6003 and L. pentosus DPC6004 were detected as part of the predominant ileal probiotic population less frequently (in four and three of the pigs, respectively; Fig. 2b). Pediococcus pentosaceus DPC6006, which was not identified in the feces of any of the probiotic-fed pigs on day 13, was found as part of the predominant ileal probiotic population in the ileal digesta of two pigs on day 28.
Ileal adherence of administered probiotic strains
Following rigorous washing, the ileal tissue was homogenized and a total count of adherent probiotic was obtained for each animal on day 28. Probiotic bacteria were detectable in the ileal mucosa of all except one of the pigs, with counts ranging from 1 × 102 to 9.7 × 103 CFU g−1 tissue (mean 2.2 × 102 CFU g−1) (data not shown). On fingerprinting representative probiotic isolates recovered from the ileal tissue homogenates, L. salivarus DPC6005 was found as the predominant probiotic strain bound to the ileal mucosa in five of the eight probiotic-fed pigs (Fig. 3). Lactobacillus murinus DPC6002 and L. pentosus DPC6004 were recovered as part of the predominant probiotic population adhering to the ileum in four and three of the eight pigs, respectively. On the other hand, L. murinus DPC6003 and P. pentosaceus DPC6006 were found in the ileal homogenate of only one of the pigs (Fig. 3).
Effects on cultivable fecal and ileal Lactobacillus and Enterobacteriaceae populations
There were no differences in cultivable fecal Lactobacillus counts between control and probiotic-fed pigs on days 0, 13, or 27 and counts did not change significantly during the 28-day treatment period (P>0.05) (Fig. 1a). Ileal counts of Lactobacillus did not differ between treatments following 28 days of probiotic administration (Fig. 2a). Mean counts of total Lactobacillus in the ileum of control and probiotic pigs on day 28 (2.1 × 107 and 3.1 × 107 CFU g−1, respectively; Fig. 2a) were c. 10-fold lower than counts in feces at day 27 (6.8 × 108 and 6.2 × 108 CFU g−1, respectively; Fig. 1a). There were no differences in fecal Enterobactericaceae counts between control and probiotic-fed pigs at any time point (Table 2). However, Enterobacteriaceae counts were higher (P<0.05) in the ileum of probiotic-fed pigs on day 28 (Table 2). Mean Enterobacteriaceae counts in the ileum (2.5 × 105 and 2.8 × 106 CFU g−1) and feces (5.8 × 105 and 4.7 × 106 CFU g−1) of control and probiotic pigs, respectively, were similar (Table 2).
Table 2. Enterobacteriaceae counts (log10 CFU g−1) in feces and ileal digesta of weanling pigs fed either c. 5 × 109 CFU day−1 of a five-strain probiotic combination in skim milk (probiotic group) or control skim milk without probiotic (control group) for 28 days
Effect of probiotic administration on phenotypes of immune cell populations
We investigated whether the five-strain probiotic combination might affect systemic immune cell populations. On day 28, we observed an increase (P<0.05) in the percentage of CD4+CD8+T cells present in the peripheral T cell population of pigs fed the probiotic compared with that present in T cells from the control pigs (7.1% vs. 5.2%; Fig. 4a). PBMCs were stimulated with phytohemagglutinin, a potent mitogen and Th1 agonist, to determine whether probiotic administration had an effect on immune response to a challenge. Expression of two cell-surface antigens, CD25 and CTLA-4, was investigated on immune cell subsets by flow cytometry. On day 28, both stimulated CD4+T cells and stimulated monocytes from probiotic-treated pigs expressed lower CD25 levels (P<0.01 and P<0.05, respectively) than those derived from control pigs (2.6- vs. 5.8- and 2.3- vs. 4.7-fold increase over resting cells for CD4+T cells and monocytes, respectively; Fig. 4b). On day 28, CTLA-4 expression on the stimulated CD4+T cells was less (P<0.05) in probiotic-treated pigs compared with control pigs (2.0- vs. 2.4-fold increase; data not shown).
Effect of probiotic administration on representative cytokines
There were no differences in either IL-1β or IL-10 production by stimulated PBMCs from probiotic-fed vs. control pigs at day 28 (data not shown). Following lipopolysaccharide stimulation on day 28, lower IL-8 production was observed in PBMCs from the probiotic-treated animals compared with control pigs (4219 vs. 7290 pg mL−1; P<0.05; data not shown). There were no significant differences in the basal expression of IL-10, IL-6, IL-1β or TNFα in ileal mucosal tissue from the probiotic-treated vs. control pigs on day 28 (Table 3). However, the mRNA expression of IL-8 in the ileal mucosa of probiotic-fed pigs was greater (P<0.05) than control pigs on day 28.
Table 3. Cytokine gene expression (copy number per μg of total RNA) at day 28 in the ileal mucosa of weanling pigs fed either c. 5 × 109 CFU day−1 of a five-strain probiotic combination in skim milk (probiotic group) or control skim milk without probiotic (control group) for 28 days
The five porcine-derived strains used in this study have been extensively characterized with respect to their probiotic traits and were chosen on the basis of their anti-Salmonella properties (Casey et al., 2004; Gardiner et al., 2004). Most notably, in a recent challenge study, a combination of all five strains has proven effective in preventing Salmonella infection in vivo, with reduced pathogen shedding and a lower incidence and duration of diarrhea in probiotic-treated pigs (Casey et al., 2007). Overall, we have accumulated data on survival and persistence of the five strains in the cecum and feces of pigs, which emphasizes the benefit of using the strains in combination (Gardiner et al., 2004), in agreement with others (Timmerman et al., 2004). However, information is lacking for the ileum and this is the main site of Salmonella invasion in the porcine GIT (Salyers & Whitt, 2002). The present study therefore examined the ability of each of the five coadministered probiotic strains to survive in and adhere to the ileum, with the intention of comparing ileal probiotic composition to that of the feces and using larger numbers of animals than in previous studies. One of the objectives was also to investigate whether bacteriocin production can offer a competitive advantage to probiotics in vivo. Furthermore, as modulation of host immunity is one of the most commonly proposed benefits of probiotics (Corthésy et al., 2007), some preliminary experiments were performed to provide insight into potential immunomodulatory effects.
Probiotic excretion and ileal survival
Fecal probiotic excretion rates were similar to those found previously in healthy pigs administered the five-strain probiotic as a fermentate (Gardiner et al., 2004) and for other probiotic strains (De Angelis et al., 2007). The decline in fecal probiotic counts even with continued daily administration has previously been observed in both healthy and infected pigs fed this five-strain probiotic (Gardiner et al., 2004; Casey et al., 2007) and in pigs fed an Enterococcus faecium probiotic in our laboratory (Gardiner et al., 1999). The fact that L. murinus DPC6002 constituted the greatest proportion of representative probiotic colonies recovered from the feces agrees with previous findings for fermentate-administered probiotic (Gardiner et al., 2004). However, in a previous study where these five cultures were administered as a skim milk suspension (as in the present study) but to Salmonella-challenged pigs, L. salivarus DPC6005 accounted for the majority of feces-derived probiotic isolates (Casey et al., 2007). This may be because the DPC6005 strain might have an advantage over its counterpart strains in a Salmonella challenge situation, as it has the most potent anti-Salmonella activity of all five strains in vitro and demonstrated the greatest ability to prevent Salmonella invasion of an intestinal epithelial cell line (Casey et al., 2004). The current study failed to identify P. pentosaceus DPC6006 as a predominant member of the fecal probiotic recovered, which concurs with previous findings (Casey et al., 2007). This strain was also the least frequently recovered from both the ileum digesta and mucosa. Assuming that survival during intestinal transit and colonization within the GIT are important for probiotic efficacy, the relevance of including this Pediococcus strain in the probiotic combination is therefore questionable.
Evidence that a bacteriocin-producing probiotic may have a competitive advantage in the GIT
Contrary to findings for the feces, of the strains fed, L. salivarus DPC6005 was the predominant one detected both in the ileum digesta and bound to the ileal mucosa of the majority of pigs. This supports previous observations with this probiotic and others that fecal microbial composition does not always accurately represent probiotic colonization within the GIT (Gardiner et al., 2004; Murphy et al., 1999) and serves to highlight the fact that fecal detection of probiotics may not be the most accurate method of predicting survival within the GIT. Interestingly, L. salivarius DPC6005 produces an antilisterial bacteriocin, salivaricin P, which is also highly active against lactic acid bacteria, including Lactobacillus and Enterococcus spp. (Barrett et al., 2007). Bacteriocin production may therefore have afforded the strain a competitive advantage, allowing it to successfully compete with resident gut microbial communities and establish itself to a greater extent in the ileum than the other four coadministered strains. While L. salivarus DPC6005 persisted poorly in the cecum in a previous study (Gardiner et al., 2004), its ability to survive in and colonize the ileum may provide more of an insight into its anti-Salmonella activity, considering the importance of the ileum as a site of attachment and invasion for Salmonella (Salyers & Whitt, 2002). Preliminary findings indicate that the relative abilities of the five strains to adhere to epithelial (HT-29) cells are similar (unpublished data). This provides further evidence that bacteriocin production may account for the ileal predominance of this strain over the other coadministered probiotics, three of which are bacteriocin sensitive. Interestingly, production of the bacteriocin mutacin has previously been shown to provide a selective advantage in the oral cavity, allowing an inoculated Streptococcus mutans strain to displace existing S. mutans populations and persist over time (Hillman et al., 2000). The fact that bacteriocin production can influence complex microbial ecosystems is further evidenced by the fact that Cheddar cheese microbial communities can be deliberately manipulated via the addition of a lacticin 3147-producing culture (Ryan et al., 2001). While previous studies have fed bacteriocin-producing probiotics to pigs (Mare et al., 2006; Strompfova et al., 2006), no attempt was made to elucidate the role of the bacteriocins. In order to establish the role of the salivaricin !P bacteriocin in probiotic colonization in the GIT, further pig-feeding studies using a nonbacteriocin-producing isogenic control strain are required. It would also be desirable to investigate bacteriocin production and stability in vivo, as we have recently done for lacticin 3147 (Gardiner et al., 2007).
Effect of probiotic administration on representative cultivable fecal and ileal microbial populations
Previous studies have demonstrated reductions in fecal Enterobacteriaceae in pigs as a result of probiotic Lactobacillus administration (Chang et al., 2001; De Angelis et al., 2007). Such reductions were also previously observed for our five-strain probiotic, although not significant and most likely due to the low pH and lactic acid content of the fermentate used to deliver the strains (Gardiner et al., 2004). This may explain the lack of effects on day 4 and 13 in the present study, as the strains were administered as a milk suspension. However, probiotic-fed pigs had higher ileal Enterobacteriaceae counts on day 28, even though no such increases were observed on repeating the study (unpublished data). This leads us to believe that the increases we observed are likely related to specific Enterobacteriaceae populations in this particular group of pigs. In any case, more meaningful data on pathogen reduction have been obtained from deliberate pathogen challenge studies, which demonstrated the efficacy of the five-strain probiotic in reducing Salmonella shedding (Casey et al., 2007). Furthermore, while the administration of Lactobacillus to weanling pigs has been shown to increase the overall Lactobacillus populations (Chang et al., 2001; De Angelis et al., 2007), our findings are in agreement with previous studies using the same five-strain combination (Gardiner et al., 2004) or other probiotic lactobacilli (Matijasic et al., 2006), in that no change in cultivable fecal Lactobacillus populations were observed. This is despite the fact that the salivaricin P bacteriocin produced by the L. salivarus DPC6005 component has anti-Lactobacillus activity (Barrett et al., 2007). Findings from the present study also confirm that there were no effects on cultivable ileal Lactobacillus populations (Gardiner et al., 2004). However, it should be noted that, as only a minor part of the gut microbiota are cultivable (Leser et al., 2002), perturbations of the gut microbiota could theoretically have occurred without being detected using the approaches employed.
Effects of probiotic administration on porcine immune phenotype
Probiotic bacteria are thought to confer health benefits in part through modulation of host immune responses (Corthésy et al., 2007). Expression of proinflammatory cytokines, IL-6, TNFα, IL-1β, and the anti-inflammatory cytokine IL-10 in intestinal mucosa of healthy pigs was unaffected by probiotic administration, which was similar to findings for other probiotics administered to pigs (Shirkey et al., 2006) and mice (Pavan et al., 2003). There was, however, an increase in the mucosal expression of IL-8 in the probiotic-treated animals but macroscopic examination indicated that this did not induce ileal inflammation (data not shown). It is possible that the constitutive low level of IL-8 expression serves to lower the threshold of activation of the immune system, driving activation during an infection, but having no effect on healthy animals, an effect previously reported in a mouse model (Lum et al., 2004). We investigated CD25 and CTLA-4 expression on T cell subsets following probiotic administration. CD25 is commonly used as a marker of lymphocyte activation (Piriou et al., 2003) and CTLA-4 is a T cell coreceptor that plays a key role in regulation of T cell activation (Inobe & Schwartz, 2004). Decreased CD25 induction on the surface of CD4+(helper) T cells and monocytes in probiotic-fed pigs compared with control animals may suggest that these cells are less active in response to a challenge following probiotic administration, which in turn suggests a potential immunomodulatory role of the five-strain probiotic. Previously, rats cocolonized with Lactobacillus plantarum and Escherichia coli demonstrated a marked and persistent increase in cells expressing CD25 in the laminae propria but effects on peripheral CD25 expression were not reported (Telemo et al., 1997). CTLA-4 is a coreceptor that curtails many aspects of an initial T cell immune response (Inobe & Schwartz, 2004). It is possible therefore that the decreased surface expression of CTLA-4 on CD4+(helper) T cells following stimulation, which was observed as a result of probiotic treatment, is indicative of altered regulation of the inflammatory T cell response. The increase in the CD4+CD8+double positive T cell subset observed after 28 days of probiotic administration may reflect an increase in memory T cells in response to any subclinical infections to which the animals were exposed. While it would be tempting to speculate that probiotic administration led to a reduction in inflammation in response to a challenge, more in-depth immunological analyses are required before any conclusions can be drawn regarding an immune mechanism.
In conclusion, our observations demonstrate that a probiotic L. salivarius survives in and adheres to the small intestine more successfully than other coadministered strains, which we suggest may be due to a competitive advantage conferred by its bacteriocin, salivaricin P. However, further studies with appropriate controls are required in order to determine the functional significance of this bacteriocin within the gut microbial ecosystem. In addition, ileal probiotic strain composition did not reflect that in the feces and assuming that intestinal survival and colonization are important for probiotic efficacy, our findings also support previous observations that question the relevance of including all five strains in the probiotic combination. Furthermore, while the five-strain probiotic demonstrated some potential to modulate host immunity, further mechanistic studies are needed to gain a better understanding of this mode of action and to gain further insight into the way in which these probiotics function to impact pig health, especially under disease challenge conditions.
We thank Dan O'Donovan, Jim Dowling, and the staff of the Moorepark pig unit for technical assistance and Liam O'Mahony for providing access to the flow cytometer. This work was supported by Enterprise Ireland and Science Foundation Ireland.