Colonisation ofpoultry bySalmonella Enteritidis S1400 is reduced by combined administration of Lactobacillus salivarius 59 and Enterococcus faecium PXN-33
Alun Carter1,2*, Martin Adams2, and Roberto M. La Ragione3Martin J. Woodward4
1Department of Bacteriology, Animal and Plant Health Agency, Addlestone, Surrey, KT15 3NB, UK.
2Faculty of Health and Medical Sciences, AX Building, University of Surrey, Guildford, Surrey, GU2 7XH, UK.
3Department of Pathology and Infectious Disease, School of Veterinary Medicine, Faculty of Health and Medical Sciences, Vet School Main Building, Daphne Jackson Road, University of Surrey, Guildford, GU2 7AL, UK.
4Department of Food and Nutrition, The University of Reading, Whiteknights Park, Reading, RG6 6AP, UK.
*Corresponding author.
Current address:Department of Medicine, Division of Oncology, Washington University, St. Louis, Missouri, USA.
Tel: (314) 757 0427
Email:
Funding: The studies presented here were supported by a commercial grant from Probiotics International Ltd (Protexin).
Key words: Probiotic, Salmonella Enteritidis, Lactobacillus salivarius, Enterococcus faecium, poultry, competitive exclusion
Title:Colonisation of poultry bySalmonella Enteritidis S1400 is reduced by combined administration of Lactobacillus salivarius 59 and Enterococcus faecium PXN-33
Abstract: Salmonella Enteritidis remains a significant issue within the poultry industry andone potential solution isto use probiotic bacteria to prevent Salmonella colonisation through competitive exclusion (CE). We demonstratethat combined administration of Lactobacillus salivarius 59 and Enterococcus faecium PXN33 were effective competitive excluders of Salmonella Enteritidis S1400 in poultry. Two models were developed to evaluate the efficacy of probiotic where birds received SalmonellaEnteritidis S1400 by a) oral gavage and b) sentinel bird to bird transmission. A statistically significant (p<0.001) 2 log reduction of Salmonella Enteritidis S1400 colonisation was observed in the ileum, caecum and colonat day 43 using combined administration of the two probiotic bacteria. However, no Salmonella Enteritidis S1400 colonisation reduction was observed when either probiotic wasadministered individually. In the sentinel bird model the combined probiotic administered at days 12 and 20 was more effective than one-off or double administrations at age 1 and 12 days. In vitro cell free culture supernatant studies suggest the mechanism of SalmonellaEnteritidis S1400 inhibition was due to a reduction in pH by the probiotic bacteria. Our current study provides further evidence that probiotics can significantly reduce pathogenic bacterial colonisation in poultry and that mixed preparation of probiotics provide superior performance when compared to individual bacterial preparations.
Introduction
Salmonella is a major cause of food poisoning that accounted for an estimated 88,715 confirmed cases in the EU in 2014 (Osimani et al., 2016). Public health concerns over Salmonellosis remain due to several prominent out-breaks, including a reported 250,000 cases in Minnesota in 1994 and a recent hospital outbreak of 287 cases in the UK (Hennessy et al., 1996; Inns et al., 2014). Contaminated eggs and poultry meat are a major source of food poisoning with 46.1% and 6.4% of Salmonellosis being attributed to eggs and broiler meat respectively (Osimani et al., 2016). Since the introduction of EU legislation, member states have targeted the reduction of Salmonella in poultry (O'Brien, 2013). However, there was a considerable increase (15.3%) in reported Salmonella cases in the EU between 2013 and 2014 (EFSA and ECDC, 2015) despite regular use of vaccines in the layer sector and improved barrier security in the broiler meat sector.
Growth promoting antibiotics have been used previously to increase bird weight gain and led to a passive control strategy for Salmonella species (de Oliveira et al., 2004). However, with the increasing emergence of antimicrobial resistance, withdrawal of antibiotics in animal feed came into force in 2006(European-Commission, 1998). Probiotics and prebiotics remain an appealing alternative control measure due to the potential competitive exclusion (CE) of pathogens, improved feed conversion rates and relatively low additional cost to production (Carter et al., 2009). Performance of probiotic preparations varies and there is a continuing need for product development and safety evaluation. Since 1972 several successful undefined Salmonella Entertitdis CE products have been developed including Aviguard and BROILACT (Nurmi and Rantala, 1973;Nuotio et al., 1992; Nakamura, et al., 2002). Concerns with safety and the spread of antibiotic resistance has led to the development of defined preparations such as the FM-B11 (Higgins et al., 2008; Vicente et al., 2008; Prado-Rebolledo et al., 2016). Multi-species and single strain probiotic cultures have been shown to reduce Salmonella in poultry although complex cultures are significantly more effective(La Ragione and Woodward, 2003; Timmerman et al., 2004; Chapman et al., 2011). Our work aimed to evaluate the efficacy of novel probiotic preparations to reduce SalmonellaEnteritidis S1400 colonisation in chickens.
Materials and Methods
Bacterial strains, culture and enumeration
Enterococcusfaecium PXN-33 and Lactobacillussalivarius 59 were kindly supplied by Probiotics International Ltd. Salmonella Enteritidis (S1400 Nalr) was used for in vivo model challenges and has been described previously (Clifton-Hadley et al., 2002; La Ragione and Woodward, 2003).Salmonella Braenderup H9812 wasused as the PFGE standard and was obtained from the Animal and Plant Health Agency (APHA) culture collection. Escherichia coli O111 and NM B171, E. coli O127:H6 EC2348/69 from the APHA culture collection was used as the controls for adhesion assays. Lactobacilli were grown for 48hrs micro-aerophilically using BBL® GasPaks® (Becton and Dickinson™ Oxford, U.K.) on de Man, Rogosa, Sharpeagar (MRS). Enterococci were grown micro-aerophilically on Slanetz and Bartley (SB) agar at 37oC for 16hrs. S. Enteritidis was grown for 16hrs aerobically on brilliant green agar (BGA). Broth cultures for enterococci, lactobacilli and Salmonella were cultured in Heart Infusion Broth (HIB), MRS andLuria-Bertani without glucose (LB-G),respectively with agitation for 16hrs at 37oC,unless stated otherwise in the methods. Prior to experimental dosing of birds, broth cultureswere centrifuged at 1700g for 10mins at room temperature and adjusted to the appropriate bacterial counts in 0.1M phosphate-buffered saline (PBS) (pH 7.2).
For culture of probiotic and S.Enteritidis S1400 isolates from in vivo studies circa 1g of tissue was added to9mls 0.1M PBS (pH 7.2), homogenized using a CAT S620® (SLS)tissue macerator, serially diluted, plated and enumeratedafter incubation in a 5% CO2atmosphere at 37oC for 24hrs: MRS, Slanetz and Bartley (SB) agar and BGA plates (supplemented with 15µg of nalidixic acid for selection of S1400) were used to culture lactobacilli, Enterococci and S.Enteritidis S1400,respectively.
Probiotic adherence to avian in vitro organ culture (IVOC) gut tissue
IVOC assays were performed as previously described with some modifications (Allen-Vercoe & Woodward, 1999; La Ragione et al., 2000). Day old SPF White Leghorn chicks (SPAFAS) were killedby cervical dislocation. Approximately 2cm sections of tissue from the crop, duodenum, jejunum, ileum, caeca and colon, were removed aseptically and placed in pre-warmed (42oC) sterile Ringer’s solution for immediateuse. The tissue loops were sliced down the longitudinal axis to expose the epithelial surface. Tissue loopswerewashed in sterile Ringer’s solution twice, placed innew 10ml of sterile pre-warmed Ringer’s solution, inoculated with 100µl of 5x108 cfu/ml of L. salivarius 59 or E. faecium PXN33bacteria and incubated aerobically at 42oC with shaking for 2 hrs. The tissues were subsequently rinsed in Ringer’s solution three times and homogenized and bacterial counts were determined. Assays were performed using three chicks from which two duplicate intestinal sections were aseptically removed and used in the association assays(adhesion and invasion). Experiments were repeated on two separate occasions. The bacteriological procedures were as described above.
Adherence of probiotic to human cell monolayers
Tissue culture assays were performed essentially as described previously with minor modifications (Dibb-Fuller et al., 1999).Briefly, HEp-2 and CaCo-2 cells were reconstituted in Dulbecco’s Modified Eagles Medium D5671 (DMEM) (Sigma) supplemented with foetal calf serum (10% v/v, Autogenbioclear), non-essential amino-acids (1% v/v, Sigma) and gentamicin (50g/ml, Sigma) and grown to confluency in 24 well micro-titre plates. HEp-2 and CaCo-2 mono-layers were washed twice in HBSSand inoculated with 5 x 107 CFU/mlL. salivarius 59 and E. faecium PXN-33 Mono-layers were then incubated at 37C supplemented with 5% CO2 in air for 3hrs. The supernatant was removed and the mono-layers were washed (x3) to remove non-adherent bacteria. Mono-layers were disrupted with 1% Triton X-100 (Sigma) and adherent bacterianumbers were determined by plating serial dilutions.
Scanning electron microscopy of HEp-2 cells
Mono-layers were grown on in 24 well plates and prepared for bacterial adherence as described above. Supernatant was removed from the mono-layers and fixed for 16 hours in 3% (v/v) glutaraldehyde in 0.1M PBS (pH 7.2). Samples were washed in 0.1M PBS (pH 7.2) and post fixed in 1% (w/v) osmium tetroxide, washed in PBS, dehydrated in ethanol and placed in hexamethyldisizane. Samples were subjected to critical point drying with liquid carbon dioxide. Air dried specimens were fixed to aluminium stubs with silver conductive paint, sputter coated with gold and examined using a Stereo-scan S250 MarkIII SEM at 10-20KV.
General in vivopoultry methods
Mixed sex SPF white leghorn chicks were used in all in vivo studies. All chicks were hatched and transferred to sterile Wey-isolatorsmaintained under negative pressure (source; APHA, Weybridge).A commercial antibiotic freefeed (complete mash diet-chick crumbs;Zootechnical Products) and water were sterilized and made available to the chicks ad libitum. All licensed procedures were approved by the local ethics committee and performed under the jurisdiction of project licenses70/6435 and 70/5282 at the APHA.
Probiotic feeding trial after S. Enteritidis S1400 oral challenge
The probiotic feeding trial used in this study was performed as described previously with minor modifications (Pascual et al., 1999).One hundred day old chicks were dividedrandomly into 4 groups of 25 birds housed in Wey-isolators. Probiotic and S.Enteritidis S1400 administration was performed by oral gavage in a final volume of 0.1ml PBS.Fourrandomized groups of birds were dosed by oral gavage with 1x109 cfu of probiotic in 100µl 0.1M PBS (pH 7.2) or 100µl of PBS only for control birds; Group 1) PBS by oral gavage,Group 2) E.faeciumPXN-33,Group 3) L.salivarius 59 or Group 4) a 50:50 preparation of both probiotic strains (combined group). At 2days old all the birds in the 4 groupswere dosed with 5x104 cfu of S. Enteritidis (S1400 Nalr). To confirm colonisation of the chicks by S. Enteritidis S1400 10 birds from each group selected at random were cloacally swabbed at 3, 6, and 8days of age and plated on BGA (supplemented with nalidixic acid). At 2 days of age and 3 days of age 3 birds were sacrificed to determine probiotic and S. Enteritidis S1400 colonisation, respectively.At 3, 6, 24 and 43 days of age 3 birds were killedby cervical dislocation and subjected topost-mortemexamination.Circa one gram of the ileum, caecum and colon were aseptically sampled for bacteriology. Probiotic colonisation was determined at days 24 and 43days of age by growth onMRS agar or SB agar for lactobacilli and enterococci, respectively. Recovered isolates were screened by PFGE(see below)to confirm identity as the challenge probiotic strain.
At 6, 24 and 43 days of age the ileum, caecum and colon samples wereplated onto BGA supplemented with nalidixic acidfor the recovery of S. Enteritidis S1400. Where no Salmonellawas recovered by direct plating, homogenateswereenriched inSelenite broth at 42oC, aerobically. Samples were re-plated onto selective media after 1 and 7 days. Approximately 10% of putativeSalmonella isolates recovered from all studieswere tested by O9 slide agglutination.
Probioticfeeding trial after S. Enteritidis S1400 sentinel bird challenge
The approach used followed that of Clifton-Hadley et al. (2002) with several modifications. Day-old birds were randomly assorted into six groups of birds with a total of 28 birds per group (Groups A-E) and one group of 30 birds (sentinelbird group: Group F) andhoused in Wey-isolators. Four of the fiveexperimental groups were dosedby oral gavage with 1x109 cfu of a combined 50:50 preparation of L.salivarius59 and E.faeciumPXN33 in 100µl 0.1M PBS (pH 7.2) (Table 1).The dosing regimes were as follows;Group A receivedno probiotic (control group),Group B received a dose at age 1 day, Group C at age 1 & 12 days, Group D at age 12 & 20 days and Group E at age 12 days (Table 1).
Sentinel birds (Group F) were dosed withS. Enteritidis S1400 by oral gavage and subsequently introduced to the experimental groups as follows. Birds in the sentinelgroup were dosed with 5x104 cfu of S.Enteritidis S1400 in 100µl 0.1M PBS (pH 7.2) at 1 and again at 12 days of age. At 12 days of age6 birds from the sentinelgroup were introduced into each of the remaining 5 trial groups(3 sentinel birds and 14 experimental birds per isolator)(detailed in Table 1).
At 16, 23, 30 and 43 days of age 3 birds from each isolator (which were not sentinelbirds) were euthanized and subjected to post-mortem examination. The ileum, caecum and colon were removed. Bacteriological enumeration of S. Enteritidis S1400,enterococci and lactobacilliwere performed as described above.
Pulse Field Gel Electrophoresis for probiotic strain identification
PFGE was performed as previously described with modifications (Jacobsen et al., 1999). Bacterial pellets harvested from 16hr cultures were washed once in 1ml of SE buffer (75mM NaCl and 25mM EDTA pH 7.4). Bacterial pellets were imbedded in tempered 2% Seakem Gold (Cambrex, East Rutherford, N.J.) and lysed in lysis buffer (50mM EDTA pH 8.2, 0.05% N-lauroyl sarcosine, 2mg/ml lysozyme and 3U/ml of mutanolysin) at 37oC for 16hrs. Plugs were incubated for 16hrs at 53°C in 10mM Tris, 0.5 M EDTA (pH 8.5), 1% sodium dodecyl sulphate (SDS), and 2 mg of proteinase K per ml and washed(x6)with SE buffer (Jacobsen et al., 1999). Plugs were digested with 25 units of SmaI (Promega, Southampton, United Kingdom) for 2hrs at 25°C. Pulsed-field gel electrophoresis (PFGE) was performed on a CHEF DRIII system (Bio-Rad, Hercules, Calif.) in 0.5% TBE extended-range buffer (Bio-Rad) and resolved in 0.8% SeaKem Gold. DNA from Salmonella Braenderup H9812 cleaved with XbaI was used as a size marker. Restriction fragments were resolved under the running conditions: Block 1, 200 V, initial time, 3.5 s, final time, 25 s, 12hrs; block 2, 200 V, initial time, 1 s, final time, 5 s; 8hrs; total time, 20hrs. The gels were visualised usingethidium bromide (Sigma, Aldrich).
Probiotic conditioned medium and disc diffusion assays
For conditioned media assays L. salivarius59,E. faeciumPXN33 and S. Enteritidis S1400 was grown overnight in Brain Heart Infusion Broth(BHIB). Probiotic supernatantswere collected and filter sterilized using 0.2µm filters. Two sets of cell free supernatant were used: pH adjusted to pH 7.2 and unadjusted. Duplicates of the conditioned medium were inoculated with ~1x105 cfu/ml of S. Enteritidis S1400 and incubated in 96 well plates for 24hrs. Controls included S. Enteritidis S1400 in unconditioned BHIB media and BHIBmedia only. Optical density readings were measured at 600nm using a Flurostar Optima®. Experiments were conducted on three separate occasions.
For disc diffusion assays semi-confluent lawns of S. Enteritidis S1400 were prepared as previously described (Andrews, 2006). S. Enteritidis S1400 was inoculated onto Iso-Sensitest agar (Oxoid,Cheshire, United Kingdom) to give semi-confluent lawn of growth. Six mm diameter BD blank paper discs were inoculated with 10µl, 15µl and 20µl of E. faeciumPXN-33 and L. salivarius59 overnight cultures grown in MRS broth. A bank disc and disc of containing 30µg of amoxycillin/clavulanic acid was used as a negative and positive control, respectively. Plates were incubated at 37oC for 24 hrs and the zones of inhibition were measured.The assay was performed in triplicate and a Student’s unpaired T-test used to compare zones of inhibition.
Statistical analysis
Statistical analysis of the data was evaluated using StatXact, Unistat and GraphPad Prism 4, (GraphPad Software Inc.) software. Data comparisons for ex vivo analysis were One-way ANOVA and significant differences were further analysedusing Bonferroni's Multiple Comparison Test post-test to compare the treatment groups to controls.In vivotransformedlog10colonisation data was analysed in StatXact system using a Two-way ANOVA General Linear Model (GLM) where the values of the control were compared to the probiotic treated group. The swabbing data from the in vivo studies was analysed using a Kruskal-Wallis,where significant differences were observed further analysis was conducted using Dunnett’s post-test to compare the treatment groups to the control group. P values of <0.05 were considered statistically significant.
Results
Adherenceof probiotic bacteriato poultry derived IVOC gut tissue and human cell lines
E. faecium PXN-33 adhered to crop tissuewithsignificantly higher numbers (circa 1 log higher) than the duodenum, jejunum, ileum and colon (3.90x105 cfu/g) (Figure 1A) (p<0.01). No significant difference was observed between the counts for the remaining tissues (p>0.05) (Figure 1A).L. salivarius 59 adhered to the crop tissue at significantly higher levels than the duodenum, jejunum, ileum, caecum and colon (Figure 1B) (circa 1 to 2 logs 4.26x106 cfu/g)(p<0.01). No significant differences were observed between the remaining tissues (p>0.05) (Figure 1B).
E. faecium PXN-33associated to HEp-2 cells and CaCo-2 cells at3.87x106 cfu (3.7% of input)and8.54x106 cfu (4.8% of input) with no significant difference between the cell lines (p=0.1325) (Figure 1C and D).L.salivarius 59 associated to HEp-2 cells and CaCo-2 cells at 8.26x105 cfu (1.2% of input) and5.77x105 cfu (0.7% of input), respectively. No significant difference between the cell lines was observed (p=0.1534) (Figure 1C and D).
E. faecium PXN-33 associated in greater numbers than L. salivarius 59 to both HEp-2 and CaCo-2 cells (p<0.0001 and p=0.0016, respectively); E. coli0127:H6 and E. coli 0111 was used as positive adhesion controls (Figure 1C and D).To gain some insight into the distribution of adhesion of the probiotic bacteria, electron microscopy was performed.These studies revealed that E. faecium PXN33 bound as single or short chains of bacterial cocciwhereas L. salivarius 59 formed dense clusters of bacterial rods (Figure 1E and 1F).
In vivo evaluation of the effect of probiotic bacteria on weight gain, GI integrity and mucin production in SPF White Leghorn chickens
At age 5, 15 and 22 days the average weight of the birds in the control group and probiotic group were 53.3g (SD± 3.6g), 144.7g (SD± 9.3g)and 232.2g (SD± 28.9g) & 52.0g (SD± 2.8g), 139.5g (SD± 6.4g) and 186.5g (SD± 81.8g), respectively. No significant difference observed between groups(p>0.05). Birds were observed at regular intervals throughout the study; no morbidity or mortality was observed in the control and probiotic treated group.Tissues were evaluated for by Alcian blue and PAS staining for acidic and neutral mucins; no significant difference in mucin production was observed between the control and probiotic treated birds when compared to the untreated control group.(SupplementalFigure 1AtoE,respectively).
Probiotic feeding trialfollowing direct oral challenge with S. Enteritidis S1400
No significant difference were observed in S.Enteritidis S1400 colonisation in birds treated with the E. faecium PXN-33 orL. salivarius 59 as single preparations at any time point tested when compared to the control (p>0.05) (data not shown).However, at age 43 days a 2 log reduction of S. Enteritidis S1400 colonisation was observed in the ileum, caecum and colon when treated with the combined probiotic preparation;the difference between the combinedprobiotic treated group compared to the control group at age 43 days was significant (p<0.001) (Figure2A).