FEMS Microbiology Letters, 362, 2015, 1–7 doi: 10.1093/femsle/fnu034 Advance Access Publication Date: 4 December 2014 Research Letter
R E S E A R C H L E T T E R – Virology
Remarkable diversity of Salmonella bacteriophages in swine and poultry ´ Denis A. Spricigo, Carlota Bardina and Montserrat Llagostera∗ Pilar Cortes, ` ` ` Departament de Genetica i de Microbiologia, Universitat Autonoma de Barcelona, 08193 Cerdanyola del Valles (Bellaterra), Barcelona, Spain ∗ Corresponding author. Departament de Genetica ` ` i de Microbiologia. Edifici C. Campus de Bellaterra. Universitat Autonoma de Barcelona, 08193
` (Bellaterra), Barcelona, Spain. Tel: +34-93-581-26-15; Fax: +34-93-581-23-87. E-mail: [email protected] Cerdanyola del Valles One sentence summary: Evidence of a remarkable diversity of Salmonella bacteriophages in poultry and swine is provided. RAPD-PCR is suggested as a suitable method for diversity studies of bacteriophages in environment. Editor: Richard Calendar
ABSTRACT The diversity of 55 Salmonella-specific bacteriophages isolated from 191 fecal samples of poultry and swine from farms located in diverse geographic areas of Spain was determined using lysis profiling, DNA restriction and random amplification of polymorphic DNA (RAPD-PCR). Among them, lysis profiling and RAPD-PCR exhibited 100% typeability and DNA restriction 96%, with discriminatory power of 0.978 (± 0.016), 0.938 (± 0.028) and 0.982 (± 0.013), respectively. The highest concordance (0.974) was that between RAPD-PCR and lysis profiling. None of the bacteriophages isolated from poultry and swine shared any DNA restriction or RAPD-PCR patterns and only two lysis profiles were common to bacteriophages isolated from poultry and swine. The major part of the lysis and RAPD-PCR profiles from the bacteriophages isolated from poultry included only one or two bacteriophages, while those obtained from swine contained more than two bacteriophages. Overall, our results provide evidence of the remarkable diversity exhibited by bacteriophages of Salmonella in farm animals. Moreover, they also show that RAPD-PCR may also be suitable for the pre-screening of the diversity of Salmonella bacteriophages for further use in biocontrol and therapeutic strategies. Key words: Salmonella; bacteriophage diversity; lysis profile; RAPD-PCR; DNA restriction profile
INTRODUCTION In the European Union, Salmonella is the principal cause of foodborne illness (EFSA 2012). Salmonellosis in humans is often due to the ingestion of contaminated meat and animal products (poultry, swine, beef, etc.) or of fruits and vegetables contaminated by animal waste (EFSA 2012), consistent with the endemic prevalence of Salmonella enterica in intensive commercial livestock production. In 2010, as in previous years, Enteritidis and Typhimurium were the serovars most frequently isolated from humans (46.6% and 6.9%, respectively). Similar findings were reported in 2009 by the US Centers for Disease Control and Prevention. In that year, there were more than 40 000 cases of
salmonellosis in the USA; most of them caused by the serovars Enteritidis (17.5%) and Typhimurium (15.0%) (CDC 2011). While the need to control the infection of livestock with pathogenic microorganisms is clear, increasing concerns regarding the transmission of multi-drug resistant bacteria and the presence of drug residues throughout the food chain have led to searches for non-pharmacological methods. The target specificity of bacteriophages, their rapid bacterial killing and their ability to self-replicate underlie the growing interest in bacteriophage-based therapeutic applications, both in humans and in animals (reviewed in Sulakvelidze, Alavidze and Morris 2001; Atterbury 2009; Chan, Abedon and Loc-Carrillo 2013). For example, numerous studies have
Received: 11 October 2014; Accepted: 11 November 2014 C FEMS 2014. All rights reserved. For permissions, please e-mail: [email protected]
1
2
FEMS Microbiology Letters, 2015, Vol. 362, No. 1
examined the utility of bacteriophages as biocontrol agents of zoonotic bacteria, including Salmonella, in animal production and food safety, with the aim of preventing food-borne illnesses (Pao et al., 2004; Fiorentin, Vieira and Barioni 2005a,b; Toro et al., 2005; Atterbury et al., 2007; Borie et al., 2008; Sharma et al., 2009; Hooton et al., 2011; Bardina et al., 2012; Spricigo et al., 2013). Bacteriophages are known to proliferate wherever their bacterial hosts exist, a reflection of their common ecology (Goyal et al., 1987). Indeed, the global population size of bacteriophages is estimated to be of the order of 1031 (Hendrix 2003). Despite the natural abundance of bacteriophages, their diversity in association with their specific bacterial host, as for example Salmonella, is still underexplored (Wang et al., 2010; Moreno Switt et al., 2013). Moreover, bacteriophage characterization requires reliable methodologies with a high discriminatory capacity. In previous studies, we demonstrated the utility of three Salmonella-specific bacteriophages in poultry therapy and on food biocontrol (Bardina et al., 2012; Spricigo et al., 2013) which belong to a collection of 55 bacteriophages isolated from poultry and swine feces. Here, we study more accurately the diversity of this collection using methodologies based on lysis profiles, DNA restriction with different enzymes and random amplification of polymorphic DNA (RAPD-PCR). In this sense, a remarkable diversity of bacteriophages belonging to farm animals is revealed and highlighted the efficacy of RAPD-PCR as a screening method suitable for this purpose.
MATERIAL AND METHODS Bacterial strains and growth conditions Lysis profile studies were carried out on 67 non-clonal strains of S. Typhimurium (n = 49) and S. Enteritidis (n = 18), isolated from avian, swine and human sources, and previously characterized by pulsed-field gel electrophoresis typing studies of a collection of 141 Salmonella isolates (Bardina et al., 2012). All Salmonella strains were routinely grown for 18 h at 37o C either in Luria Bertani (LB) broth or on LB agar plates.
Bacteriophage isolation Salmonella-specific bacteriophages were isolated from 191 feces obtained from cloacal or rectal swabs collected between 2007 and 2009 from 88 chickens and 103 pigs from different farms located in diverse geographic areas of Spain. For the isolation of bacteriophages, S. Typhimurium strains (8880, S4426, 887, 1711/F06, 1624/F06 and S6254) and S. Enteritidis strains (7358, 9310, 951, 9449 and 9609) were used (Bardina et al., 2012). Bacteriophages were isolated either by a previously described method (Connerton et al., 2004; Muniesa et al., 2005) or by an enrichment procedure. In the former, 1 g of the sample was diluted in 10 ml of SM buffer (50 mM Tris-HCl, pH 7.5; 0.1 M NaCl; 8 mM MgSO4 ·7H2 O; 0.01% gelatin), incubated at 4o C for 24 h, and then centrifuged at 7000 × g for 10 min. In the enrichment procedure, 1 g of the sample was diluted in 10 ml of peptone water and incubated at 37o C for 18 h. One milliliter of this enrichment culture was inoculated in 10 ml of Muller-Kauffmann selective ¨ broth (Merck), incubated at 37o C for 24 h, and then centrifuged at 8000 × g (Beckman J2-21) for 10 min. The supernatants obtained from both methods were filtered through a 0.45-μm polyvinylidene fluoride membrane filter (PVDF; Millipore) and then used as described below. Bacteriophages were isolated and purified in spot tests. Briefly, 10 μl of each filtrate was spotted onto the surfaces of
Salmonella lawns by the double agar layer method. After incubation of the bacterial plates at 37◦ C, morphologically different plaques were selected and resuspended in 1 ml of sterile SM buffer. Ten-fold serial dilutions of the bacteriophage suspensions were plated by the double agar layer method and bacteriophages that produced clear plaques were selected. This procedure was repeated at least three times to obtain a single type of bacteriophage.
In vitro multiplication of bacteriophages When necessary, large-scale cultures of bacteriophages were produced by inoculating exponential cultures of Salmonella (108 colony-forming units (cfu) ml−1 ) grown in LB medium at 37◦ C at a multiplicity of infection of 1 plaque-forming unit (pfu) cfu−1 , followed by incubation at 37◦ C for 5 h. The infected cultures were then centrifuged at 8000 × g for 10 min (Beckman) and the supernatants were filtered through 0.45-μm syringe PVDFs (Millipore). The bacteriophage titer was determined by plating serial dilutions onto LB plates using the double agar layer method. The plates were incubated at 37◦ C for 24 h, after which the bacteriophage plaques were counted. When necessary, bacteriophages were concentrated by ultracentrifugation at 51 000 × g for 2 h (Beckman).
Lysis profile The lysis profile of each bacteriophage was determined by spotting 10 μl of the above-described lysates (108 pfu ml−1 ) on lawns of each of the 67 clonally unrelated strains of S. Typhimurium and S. Enteritidis. The plates were incubated at 37◦ C for 24 h and bacterial lysis was recorded.
Bacteriophage DNA restriction analysis The restriction profiles of bacteriophage DNA, obtained from lysates (1011 pfu ml−1 ) as previously described (Ausubel et al., 1987), were analyzed with several different restriction enzymes (BstEII, EcoRV, EcoRI, HindIII, PstI and XbaI) in order to avoid possible digestion resistance. All the enzymes were obtained from New England Biolabs and the restriction reactions were performed according to the manufacturer’s recommendations. DNA fragments were analyzed by electrophoresis in 0.7% agarose gels in Tris-acetate-EDTA buffer at a constant voltage of 80 V and visualized under UV light by staining with ethidium bromide. Bacteriophage lambda DNA digested with HindIII (Roche) and BstEII (New England Biolabs) was used as the molecular weight marker.
RAPD-PCR analysis RAPD-PCR was carried out according to a previously described ´ method (Gutierrez et al., 2011) in individual reactions containing 8 μM of the primers OPL5 (5 -ACGCAGGCAC-3’), P1 (5 -CCGCAGCCAA-3 ) and P2 (5 -AACGGGCAGA-3 ). Ten nanograms of purified bacteriophage DNA was PCR-amplified in reactions supplemented with 3 mM MgCl2 and 5% v/v dimethyl ´ sulfoxide under conditions previously described (Gutierrez et al., 2011). The RAPD-PCR products were separated electrophoretically on a 0.8% agarose gel. The DNA molecular weight marker consisted of a mixture of bacteriophage lambda DNA digested with BstEII (New England Biolabs) and 174 DNA (Promega). A similarity analysis was performed using the software FPQuest
´ et al. Cortes
(Bio-Rad) with a Dice coefficient. Clustering was done using the unweighted pair group method with arithmetic averages (Struelens et al., 1996). Profiles were considered different when at least one polymorphic band was identified.
Typing methods analysis The typeability of each typing method was ascertained as reported (Struelens et al., 1996). The discriminatory power (± 95% confidence interval), as defined by Struelens et al. (1996), was measured by calculating the Simpson’s index of diversity (Hunter and Gaston 1988). Moreover, the concordance among the typing methods was determined by cross-classifying all possible pairs of bacteriophages (1485 pairs) on the basis of matched or mismatched types. The resulting 2×2 table was evaluated with a chi-square statistic and the percentage of concordant cells was calculated (Robinson et al., 1998).
Statistical analysis Bacteriophages were studied by their ability to lyse Salmonella, considering both the serovar and the source of the bacterial strains. To do so, we used a bivariate analysis, applying the appropriate discrete distributions homogeneity test (chi-square test, Fisher’s exact or likelihood ratio) based on the performance of the application criteria. The bivariate analysis was performed using the software SAS v9.2 (SAS Institute Inc., Cary, NC, USA). The significance level was set to 0.05.
RESULTS During the years 2007–2009, 191 fecal samples from chicken and swine were obtained from farms located in different geographic areas of Spain. Analysis of the samples for the presence of Salmonella bacteriophages revealed that 28.8% were positive. These 55 bacteriophages, 32% (33/103) from swine and 25% (22/88) from poultry fecal samples, were further studied and characterized. Spotting tests of 67 Salmonella strains showed a high variability of the bacteriophages, distinguishing 39 different lysis profiles (Table 1 and Fig. S1, Supporting Information) defined according to the absence or presence of lysis on the tested bacterial strains. Twenty lysis profiles represented the 22 bacteriophages isolated from poultry and 21 lysis profiles were distinguished from the 33 bacteriophages isolated from swine. Only two lysis profiles (numbers 8 and 22, Table 1) were shared by bacteriophages isolated from poultry and swine. Moreover, the majority of the bacteriophages (n = 51) showed a broad lysis profile (lysing more than 10 Salmonella strains). Among them, 35 bacteriophage isolates, representing 20 lysis patterns, lysed more than 50 Salmonella strains. The most common lysis profile was that of six bacteriophages isolated from swine. Only four bacteriophages lysed fewer than 10 bacterial strains. A bivariate analysis was conducted to determine any statistically significant shared features among the bacteriophage lysis profiles and the source and serovar of the lysed strains of Salmonella. In this analysis, those phages that were isolated from samples obtained at different times but nonetheless had identical lysis profiles were included. For bacteriophages with the same lysis patterns and isolated from samples collected on the same date, only one bacteriophage was chosen as representative. Based on these criteria, the statistical comparison was conducted with 46 out of 55 isolated bacteriophages. Bivariate analysis indicated that 41.3% of them (19 out of 46) infected a higher percentage of S. Enteritidis than S. Typhimurium strains
3
(p ≤ 0.01), regardless of the source of the bacterial strain (poultry, swine or humans). The 55 bacteriophages were further characterized by DNA restriction analysis. The combined analysis of the DNA patterns obtained with the enzymes BstEII, EcoRV, EcoRI, HindIII, PstI and XbaI produced 22 different restriction profiles (Table 2). The identification of 21 of them was possible with the enzymes EcoRV and XbaI. EcoRV had a greater discriminatory capacity, distinguishing 19 patterns, while bacteriophages which showed the pattern 11 (Fig. 1A) were further distinguished into patterns 1, 4 and 5 (Fig. 1B) with XbaI enzyme. Restriction profile 22 (Table 2) corresponded to one bacteriophage whose DNA was resistant to all the enzymes tested except BstEII. Two bacteriophages could not be characterized because their DNA was resistant to digestion by all of the restriction enzymes tested. Bacteriophages isolated from poultry and swine did not share any restriction patterns. Eleven restriction patterns were obtained from the 22 bacteriophages isolated from poultry and the remaining 11 from the 33 bacteriophages isolated from swine (Table 2). These results were confirmed in duplicate experiments. RAPD-PCR was used as an additional typing method to characterize the collection of Salmonella bacteriophages. This method was previously used to type bacteriophages from dif´ ferent bacterial species but not Salmonella (Gutierrez et al., 2011). The 55 bacteriophages in our study were typed in single reactions with three different primers, OLP5, P1 and P2. Primer P1 showed the highest discriminatory capacity, identifying 34 different PCR patterns; OLP5 and P2 yielded 31 and 24 patterns, respectively. A dendrogram generated by combining the results obtained with the three primers allowed the differentiation of 40 RAPD profiles (Fig. 2). Nine band patterns, associated with bacteriophage isolates from swine, differed only in one band, as determined in two different experiments (Fig. 2). No patterns common to bacteriophages from poultry and swine were identified after RAPD-PCR. Reproducible patterns were obtained with different bacteriophages DNA in separate experiments (data not shown). Regarding the typing characteristics of the methods used in this study, the typeability of lysis profiling was 100%, that of RAPD-PCR 100% as well and that of restriction analysis 96%. The discriminatory power (D-values of the Simpson’s diversity index) was 0.978 (± 0.016), 0.938 (± 0.028) and 0.982 (± 0.013) for lysis profiling, DNA restriction and RAPD-PCR, respectively. The highest concordance regarding the discriminative ability (Robinson et al., 1998) was between RAPD-PCR and lysis profiling (0.974), followed by DNA restriction and RAPD-PCR (0.884) and lysis profiling and DNA restriction (0.868).
DISCUSSION This study used lysis profiling, DNA restriction patterns and RAPD-PCR analysis to characterize the diversity of bacteriophages infecting the serovars Typhimurium and Enteritidis in poultry and swine. The bacteriophages were isolated from 28.8% of the 191 cloacal or rectal swabs collected from poultry and swine between 2007 and 2009. A higher percentage of bacteriophages (32%) was detected in swine than in poultry (25%). This finding is in agreement with previous reports on the prevalence of Salmonella in poultry and swine farms. Thus, in Spain, during the same period of time, the prevalence of Salmonella on swine farms was >50% and on poultry farms 25.3% (EFSA 2010). These results suggest that the presence of bacteriophages infecting Salmonella in farm animals roughly parallels that of the
4
FEMS Microbiology Letters, 2015, Vol. 362, No. 1
Table 1. Lysis profiles of the bacteriophages studied.
Lysis profile number
Number of bacteriophages by source
Number of lysed Salmonella strainsa
Poultry
Swine
Typhimurium
Enteritidis
1 2 3 4 5 6 7 8b 9 10 11 12 13 14 15 16 17 18 19 20 21 22b 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
– – – – – – 1 2 – – – – 1 1 – – 1 1 1 1 1 1 2 1 1 1 1 1 – – – – – 1 1 1 1 – –
1 3 4 2 6 1 – 1 1 1 1 1 – – 1 1 – – – – – 2 – – – – – – 1 1 1 1 1 – – – – 1 1
41 39 40 40 40 40 42 40 40 46 0 0 41 27 17 7 20 24 43 15 40 47 13 11 11 16 0 46 12 11 2 11 45 41 41 36 43 8 21
16 16 15 14 16 14 17 14 15 18 4 7 15 18 16 13 14 15 18 13 17 18 14 13 14 15 2 18 12 11 6 10 18 18 17 15 18 13 16
Number of total bacteriophages
22
33
Number of total lysis profiles
20
21
–Lysis profile not detected. a Number of lysed strains, from a total of 49 S. Typhimurium and 18 S. Enteritidis strains. b Lysis profile shared for bacteriophages from poultry and swine feces.
bacterium. It should be pointed out that the overall bacteriophage isolation efficiency in this study was higher than that reported by other authors also analyzing poultry and swine (Fiorentin et al., 2004; Toro et al., 2005). The higher efficiency can perhaps be attributed to the enrichment isolation procedure, which may have influenced the specific selection of Salmonella and therefore of the specific bacteriophages infecting these bacterial strains. The phenotypic and genotypic diversity of the 55 bacteriophages isolated was considerable, as indicated by the 39
different lysis profiles and further demonstrated by RAPD-PCR (40 profiles). By contrast, DNA restriction analysis had a lower discriminatory capacity, even though six different restriction enzymes were employed. However, this method detected the same number of profiles for bacteriophages isolated from poultry (n = 11) and from swine (n = 11), as identified by lysis profiling and RAPD-PCR. Although RAPD-PCR was previously used to type 26 bacteriophages infecting gram-positive bacteria and Escherichia ´ coli (Gutierrez et al., 2011), prior to our study it had not been compared with the other methodologies commonly used to type
´ et al. Cortes
Table 2. DNA restriction profiles of the studied bacteriophages, according to the combined analysis of the patterns obtained with the enzymes BstEII, EcoRV, EcoRI, HindIII, PstI and XbaI.
Restriction profile
a
Number of bacteriophages by source
numbera
Poultry
Swine
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 NDb
1 – – 1 – 1 – – – 1 – 1 5 6 1 1 – – 3 1 – – 0
– 4 6 – 9 – 1 1 1 – 1 – – – – – 3 1 – – 3 1 2
Number of total bacteriophages
22
33
Number of total profiles
11
11
All the bacteriophages belonging to a restriction profile showed the same restriction pattern for all the enzymes used. b ND, bacteriophages whose DNA could not be digested with any of the enzymes used.
5
bacteriophages. Our results show that this typing method can be applied to other gram-negative bacteria, with results comparable to lysis profiling, the conventional approach. Moreover, an advantage of RAPD-PCR is that it is rapid, because it can be used ´ directly with bacteriophage suspensions (Gutierrez et al., 2011). The high agreement between lysis profiling and RAPD-PCR supports the use of the latter method in bacteriophage diversity studies. Among the bacteriophages isolated from poultry, it was remarkable that many of the lysis and RAPD profiles represented only one or two of them, while for bacteriophages isolated from swine, there were profiles containing more than two bacteriophages. This suggests that there is a greater presence of some bacteriophage patterns in swine than in poultry. The difference in abundance could be related to the respective animal production systems, in that closed-cycle swine production is still very common in Spain whereas poultry farms rely almost entirely on a fattening/breeding system. Although the latter is a more specialized form of production, it can promote the greater local movement of microorganisms, including Salmonella and their bacteriophages. In addition, a greater diversity of bacteriophages may allow the development of new bacteriophage, through recombination processes. This mechanism has been shown to explain the mosaic structure of bacteriophage genomes and the generation of new gene combinations in bacteriophages (Casjens and Thuman-Commike 2011). Moreover, bacteriophage diversity seems linked to evolution of host specificity because bacteriophages infecting Salmonella strains from poultry and pigs present different lysis profiles. Interestingly, the bacteriophages isolated from poultry and swine did not overlap with respect to their restriction patterns and RAPD-PCR profiles and only two lysis profiles were shared (Tables 1 and 2; Fig. 2). However, the relationship between the source of the bacteriophages (poultry and swine) and that of the Salmonella strains infected by them (poultry and swine) was not statistically significant. Thus, poultry and swine carry different bacteriophages, which nonetheless infect Salmonella strains isolated from either of these animals.
Figure 1. Nineteen DNA restriction patterns (lines 1 to 19) obtained with EcoRV (A) and five restriction patterns (lines 1 to 5) distinguished with XbaI (B) for the 55 bacteriophages studied. Lambda DNA digested with HindIII (M1) and lambda DNA digested with BstEII mixed with φX714 DNA digested with HinfI (M2) were used as molecular markers.
6
FEMS Microbiology Letters, 2015, Vol. 362, No. 1
Figure 2. Dendrogram based on the combined analysis of RAPD patterns obtained with primers OLP5, P1 and P2. The letters P and S in the name of bacteriophages indicate the poultry and swine origin of the bacteriophages, respectively.
In conclusion, poultry and swine contain a remarkable diversity of bacteriophages and most of them are able to efficiently infect a great variety of non-clonal strains of S. Typhimurium and S. Enteritidis. Our results also demonstrate a high concordance between lysis profiling and RAPD-PCR, with the latter providing a useful, quick and easy method to assess the diversity of bacteriophages in the environment. In addition, RAPD-PCR may also be suitable for pre-screening bacteriophages to be applied for other purposes. In this sense, those bacteriophages having a broader lysis spectrum and with different RAPD patterns would be the best candidates for biocontrol and therapeutic strategies against Salmonella.
SUPPLEMENTARY DATA Supplementary data is available at FEMSLE online.
ACKNOWLEDGEMENTS We are deeply indebted to the Laboratori de Sanitat Animal (DAAM, Generalitat de Catalunya, Spain), the Santa Creu i Sant Pau Hospital, and the Vall d’Hebron Hospital for providing us with the different Salmonella strains used in this study and to ´ Laboratorios Callier and Laboratorio General de Diagnostico for providing the fecal samples.
´ et al. Cortes
FUNDING This work was supported by the Instituto Nacional de Investi´ y Tecnolog´ıa Agraria y Alimentaria; Ministerio de Ciencia gacion ´ Espanol ˜ (grant number RTA2006-00065) and from e Innovacion Generalitat de Catalunya (grant number SGR2014/572). CB is the recipient of a predoctoral fellowship from the Comissionat per ´ Universia Universitats i Recerca del Departament d’Innovacio, tats i Empresa de la Generalitat de Catalunya i del Fons Social Europeu. DS is supported by a predoctoral fellowship from CAPES ˜ de Aperfeic¸oamento de Pessoal de N´ıvel Superior), (Coordenac¸ao Brazil. Conflict of interest statement. None declared.
REFERENCES Atterbury RJ. Bacteriophage biocontrol in animals and meat products. Microb Biotechnol 2009;2:601–12. Atterbury RJ, Van Bergen MAP, Ortiz F, et al. Bacteriophage therapy to reduce Salmonella colonization of broiler chickens. Appl Environ Microb 2007;73:4543–9. Ausubel FM, Brent R, Kingston RE, et al. Current Protocols in Molecular Biology, Vol. 1. New York: Greene/Wiley-Interscience, 1987. ´ P, et al. Significance of the bacBardina C, Spricigo DA, Cortes teriophage treatment schedule in reducing Salmonella colonization of poultry. Appl Environ Microb 2012;78:6600–7. ´ Borie C, Albala I, Sanchez P, et al. Bacteriophage treatment reduces Salmonella colonization of infected chickens. Avian Dis 2008;52:64–7. Casjens SR, Thuman-Commike PA. Evolution of mosaically related tailed bacteriophage genomes seen through the lens of phage P22 virion assembly. Virology 2011;411:393–415. Center for Disease Control and Prevention. Salmonella: Annual Summary (2009) Atlanta, GA: CDC, Division of Bacterial and Mycotic Diseases, 2011. http://www.cdc.gov/ncezid/dfwed/PDFs/ SalmonellaAnnualSummaryTables2009.pdf Chan BK, Abedon ST, Loc-Carrillo C. Phage cocktails and the future of phage therapy. Future Microbiol 2013;8:769–83. Connerton PL, Loc Carrillo CM, Swift C, et al. Longitudinal study of Campylobacter jejuni bacteriophages and their hosts from broiler chickens. Appl Environ Microb 2004;70:3877–83. European Food Safety Authority. The Community summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in the European Union in 2008. EFSA J 2010;8:1496. European Food Safety Authority. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2010. EFSA J 2012;10:2597. Fiorentin L, Vieira ND, Barioni W, Jr, et al. In vitro characterization and in vivo properties of salmonellae lytic bacteriophages isolated free-range layers. Braz J Poult Sci 2004;6:121–8. Fiorentin L, Vieira ND, Barioni W, Jr. Oral treatment with bacteriophages reduces the concentration of Salmonella Enteri-
7
tidis PT4 in caecal contents of broilers. Avian Pathol 2005a;34: 258–63. Fiorentin L, Vieira ND, Barioni W, Jr. Use of lytic bacteriophages to reduce Salmonella Enteritidis in experimentally contaminated chicken cuts. Braz J Poult Sci 2005b;7:255–60. Goyal SM, Gerba CP, Bitton G. Phage Ecology. New York, NY: John Wiley and Sons, 1987, 321. ´ Gutierrez D, Mart´ın-Platero AM, Rodr´ıguez A, et al. Typing of bacteriophages by randomly amplified polymorphic DNA (RAPD)-PCR to assess genetic diversity. FEMS Microbiol Lett 2011;322:90–7. Hendrix RW. Bacteriophage genomics. Curr Opin Microbiol 2003;6:506–11. Hooton SPT, Atterbury RJ, Connerton IF. Application of a bacteriophage cocktail to reduce Salmonella Typhimurium U288 contamination on pig skin. Int J Food Microbiol 2011;151: 157–63. Hunter PR, Gaston MA. Numerical index of the discriminatory ability of typing systems: an application of Simpson’s index of diversity. J Clin Microbiol 1988;26:2465–6. Moreno Switt AI, den Bakker HC, Vongkamjan K, et al. Salmonella bacteriophage diversity reflects host diversity on dairy farms. Food Microbiol 2013;36:275–85. Muniesa M, Blanch AR, Lucena F, et al. Bacteriophages may bias outcome of bacterial enrichment cultures. Appl Environ Microb 2005;71:4269–75. Pao S, Randolph SP, Westbrook EW, et al. Use of bacteriophages to control Salmonella in experimentally contaminated sprout seeds. J Food Sci 2004;69: M127–30. Robinson DA, Hollingshead SK, Musser JM, et al. The IS1167 insertion sequence is a phylogenetically informative marker among isolates of serotype 6B Streptococcus pneumoniae. J Mol Evol 1998;47:222–9. Sharma M, Patel JR, Conway WS, et al. Effectiveness of bacteriophages in reducing Escherichia coli O157:H7 on fresh-cut cantaloupes and lettuce. J Food Protect 2009;72:1481–5. ´ P, et al. Use of a bacteriophage Spricigo DA, Bardina C, Cortes cocktail to control Salmonella in food and the food industry. Int J Food Microbiol 2013;165:169–74. the Members of the European Study Group on epidemiological Markers (ESGEM) of the European Society for Clinical Microbiology and Infectious Diseases (ESCMID)Struelens MJ, . Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin Microbiol Infect 1996;2:2–11. Sulakvelidze A, Alavidze Z, Morris JG, Jr. Bacteriophage Therapy. Antimicrob Agents Ch 2001;45:649–59. Toro H, Price SB, McKee AS, et al. Use of bacteriophages in combination with competitive exclusion to reduce Salmonella from infected chickens. Avian Dis 2005;49: 118–24. Wang S, Zhao W, Raza A, et al. Prevalence of Salmonella infecting bacteriophages associated with Ontario pig farms and the holding area of a high capacity pork processing facility. J Sci Food Agr 2010;90:2318–25.
R E S E A R C H L E T T E R – Virology
Remarkable diversity of Salmonella bacteriophages in swine and poultry ´ Denis A. Spricigo, Carlota Bardina and Montserrat Llagostera∗ Pilar Cortes, ` ` ` Departament de Genetica i de Microbiologia, Universitat Autonoma de Barcelona, 08193 Cerdanyola del Valles (Bellaterra), Barcelona, Spain ∗ Corresponding author. Departament de Genetica ` ` i de Microbiologia. Edifici C. Campus de Bellaterra. Universitat Autonoma de Barcelona, 08193
` (Bellaterra), Barcelona, Spain. Tel: +34-93-581-26-15; Fax: +34-93-581-23-87. E-mail: [email protected] Cerdanyola del Valles One sentence summary: Evidence of a remarkable diversity of Salmonella bacteriophages in poultry and swine is provided. RAPD-PCR is suggested as a suitable method for diversity studies of bacteriophages in environment. Editor: Richard Calendar
ABSTRACT The diversity of 55 Salmonella-specific bacteriophages isolated from 191 fecal samples of poultry and swine from farms located in diverse geographic areas of Spain was determined using lysis profiling, DNA restriction and random amplification of polymorphic DNA (RAPD-PCR). Among them, lysis profiling and RAPD-PCR exhibited 100% typeability and DNA restriction 96%, with discriminatory power of 0.978 (± 0.016), 0.938 (± 0.028) and 0.982 (± 0.013), respectively. The highest concordance (0.974) was that between RAPD-PCR and lysis profiling. None of the bacteriophages isolated from poultry and swine shared any DNA restriction or RAPD-PCR patterns and only two lysis profiles were common to bacteriophages isolated from poultry and swine. The major part of the lysis and RAPD-PCR profiles from the bacteriophages isolated from poultry included only one or two bacteriophages, while those obtained from swine contained more than two bacteriophages. Overall, our results provide evidence of the remarkable diversity exhibited by bacteriophages of Salmonella in farm animals. Moreover, they also show that RAPD-PCR may also be suitable for the pre-screening of the diversity of Salmonella bacteriophages for further use in biocontrol and therapeutic strategies. Key words: Salmonella; bacteriophage diversity; lysis profile; RAPD-PCR; DNA restriction profile
INTRODUCTION In the European Union, Salmonella is the principal cause of foodborne illness (EFSA 2012). Salmonellosis in humans is often due to the ingestion of contaminated meat and animal products (poultry, swine, beef, etc.) or of fruits and vegetables contaminated by animal waste (EFSA 2012), consistent with the endemic prevalence of Salmonella enterica in intensive commercial livestock production. In 2010, as in previous years, Enteritidis and Typhimurium were the serovars most frequently isolated from humans (46.6% and 6.9%, respectively). Similar findings were reported in 2009 by the US Centers for Disease Control and Prevention. In that year, there were more than 40 000 cases of
salmonellosis in the USA; most of them caused by the serovars Enteritidis (17.5%) and Typhimurium (15.0%) (CDC 2011). While the need to control the infection of livestock with pathogenic microorganisms is clear, increasing concerns regarding the transmission of multi-drug resistant bacteria and the presence of drug residues throughout the food chain have led to searches for non-pharmacological methods. The target specificity of bacteriophages, their rapid bacterial killing and their ability to self-replicate underlie the growing interest in bacteriophage-based therapeutic applications, both in humans and in animals (reviewed in Sulakvelidze, Alavidze and Morris 2001; Atterbury 2009; Chan, Abedon and Loc-Carrillo 2013). For example, numerous studies have
Received: 11 October 2014; Accepted: 11 November 2014 C FEMS 2014. All rights reserved. For permissions, please e-mail: [email protected]
1
2
FEMS Microbiology Letters, 2015, Vol. 362, No. 1
examined the utility of bacteriophages as biocontrol agents of zoonotic bacteria, including Salmonella, in animal production and food safety, with the aim of preventing food-borne illnesses (Pao et al., 2004; Fiorentin, Vieira and Barioni 2005a,b; Toro et al., 2005; Atterbury et al., 2007; Borie et al., 2008; Sharma et al., 2009; Hooton et al., 2011; Bardina et al., 2012; Spricigo et al., 2013). Bacteriophages are known to proliferate wherever their bacterial hosts exist, a reflection of their common ecology (Goyal et al., 1987). Indeed, the global population size of bacteriophages is estimated to be of the order of 1031 (Hendrix 2003). Despite the natural abundance of bacteriophages, their diversity in association with their specific bacterial host, as for example Salmonella, is still underexplored (Wang et al., 2010; Moreno Switt et al., 2013). Moreover, bacteriophage characterization requires reliable methodologies with a high discriminatory capacity. In previous studies, we demonstrated the utility of three Salmonella-specific bacteriophages in poultry therapy and on food biocontrol (Bardina et al., 2012; Spricigo et al., 2013) which belong to a collection of 55 bacteriophages isolated from poultry and swine feces. Here, we study more accurately the diversity of this collection using methodologies based on lysis profiles, DNA restriction with different enzymes and random amplification of polymorphic DNA (RAPD-PCR). In this sense, a remarkable diversity of bacteriophages belonging to farm animals is revealed and highlighted the efficacy of RAPD-PCR as a screening method suitable for this purpose.
MATERIAL AND METHODS Bacterial strains and growth conditions Lysis profile studies were carried out on 67 non-clonal strains of S. Typhimurium (n = 49) and S. Enteritidis (n = 18), isolated from avian, swine and human sources, and previously characterized by pulsed-field gel electrophoresis typing studies of a collection of 141 Salmonella isolates (Bardina et al., 2012). All Salmonella strains were routinely grown for 18 h at 37o C either in Luria Bertani (LB) broth or on LB agar plates.
Bacteriophage isolation Salmonella-specific bacteriophages were isolated from 191 feces obtained from cloacal or rectal swabs collected between 2007 and 2009 from 88 chickens and 103 pigs from different farms located in diverse geographic areas of Spain. For the isolation of bacteriophages, S. Typhimurium strains (8880, S4426, 887, 1711/F06, 1624/F06 and S6254) and S. Enteritidis strains (7358, 9310, 951, 9449 and 9609) were used (Bardina et al., 2012). Bacteriophages were isolated either by a previously described method (Connerton et al., 2004; Muniesa et al., 2005) or by an enrichment procedure. In the former, 1 g of the sample was diluted in 10 ml of SM buffer (50 mM Tris-HCl, pH 7.5; 0.1 M NaCl; 8 mM MgSO4 ·7H2 O; 0.01% gelatin), incubated at 4o C for 24 h, and then centrifuged at 7000 × g for 10 min. In the enrichment procedure, 1 g of the sample was diluted in 10 ml of peptone water and incubated at 37o C for 18 h. One milliliter of this enrichment culture was inoculated in 10 ml of Muller-Kauffmann selective ¨ broth (Merck), incubated at 37o C for 24 h, and then centrifuged at 8000 × g (Beckman J2-21) for 10 min. The supernatants obtained from both methods were filtered through a 0.45-μm polyvinylidene fluoride membrane filter (PVDF; Millipore) and then used as described below. Bacteriophages were isolated and purified in spot tests. Briefly, 10 μl of each filtrate was spotted onto the surfaces of
Salmonella lawns by the double agar layer method. After incubation of the bacterial plates at 37◦ C, morphologically different plaques were selected and resuspended in 1 ml of sterile SM buffer. Ten-fold serial dilutions of the bacteriophage suspensions were plated by the double agar layer method and bacteriophages that produced clear plaques were selected. This procedure was repeated at least three times to obtain a single type of bacteriophage.
In vitro multiplication of bacteriophages When necessary, large-scale cultures of bacteriophages were produced by inoculating exponential cultures of Salmonella (108 colony-forming units (cfu) ml−1 ) grown in LB medium at 37◦ C at a multiplicity of infection of 1 plaque-forming unit (pfu) cfu−1 , followed by incubation at 37◦ C for 5 h. The infected cultures were then centrifuged at 8000 × g for 10 min (Beckman) and the supernatants were filtered through 0.45-μm syringe PVDFs (Millipore). The bacteriophage titer was determined by plating serial dilutions onto LB plates using the double agar layer method. The plates were incubated at 37◦ C for 24 h, after which the bacteriophage plaques were counted. When necessary, bacteriophages were concentrated by ultracentrifugation at 51 000 × g for 2 h (Beckman).
Lysis profile The lysis profile of each bacteriophage was determined by spotting 10 μl of the above-described lysates (108 pfu ml−1 ) on lawns of each of the 67 clonally unrelated strains of S. Typhimurium and S. Enteritidis. The plates were incubated at 37◦ C for 24 h and bacterial lysis was recorded.
Bacteriophage DNA restriction analysis The restriction profiles of bacteriophage DNA, obtained from lysates (1011 pfu ml−1 ) as previously described (Ausubel et al., 1987), were analyzed with several different restriction enzymes (BstEII, EcoRV, EcoRI, HindIII, PstI and XbaI) in order to avoid possible digestion resistance. All the enzymes were obtained from New England Biolabs and the restriction reactions were performed according to the manufacturer’s recommendations. DNA fragments were analyzed by electrophoresis in 0.7% agarose gels in Tris-acetate-EDTA buffer at a constant voltage of 80 V and visualized under UV light by staining with ethidium bromide. Bacteriophage lambda DNA digested with HindIII (Roche) and BstEII (New England Biolabs) was used as the molecular weight marker.
RAPD-PCR analysis RAPD-PCR was carried out according to a previously described ´ method (Gutierrez et al., 2011) in individual reactions containing 8 μM of the primers OPL5 (5 -ACGCAGGCAC-3’), P1 (5 -CCGCAGCCAA-3 ) and P2 (5 -AACGGGCAGA-3 ). Ten nanograms of purified bacteriophage DNA was PCR-amplified in reactions supplemented with 3 mM MgCl2 and 5% v/v dimethyl ´ sulfoxide under conditions previously described (Gutierrez et al., 2011). The RAPD-PCR products were separated electrophoretically on a 0.8% agarose gel. The DNA molecular weight marker consisted of a mixture of bacteriophage lambda DNA digested with BstEII (New England Biolabs) and 174 DNA (Promega). A similarity analysis was performed using the software FPQuest
´ et al. Cortes
(Bio-Rad) with a Dice coefficient. Clustering was done using the unweighted pair group method with arithmetic averages (Struelens et al., 1996). Profiles were considered different when at least one polymorphic band was identified.
Typing methods analysis The typeability of each typing method was ascertained as reported (Struelens et al., 1996). The discriminatory power (± 95% confidence interval), as defined by Struelens et al. (1996), was measured by calculating the Simpson’s index of diversity (Hunter and Gaston 1988). Moreover, the concordance among the typing methods was determined by cross-classifying all possible pairs of bacteriophages (1485 pairs) on the basis of matched or mismatched types. The resulting 2×2 table was evaluated with a chi-square statistic and the percentage of concordant cells was calculated (Robinson et al., 1998).
Statistical analysis Bacteriophages were studied by their ability to lyse Salmonella, considering both the serovar and the source of the bacterial strains. To do so, we used a bivariate analysis, applying the appropriate discrete distributions homogeneity test (chi-square test, Fisher’s exact or likelihood ratio) based on the performance of the application criteria. The bivariate analysis was performed using the software SAS v9.2 (SAS Institute Inc., Cary, NC, USA). The significance level was set to 0.05.
RESULTS During the years 2007–2009, 191 fecal samples from chicken and swine were obtained from farms located in different geographic areas of Spain. Analysis of the samples for the presence of Salmonella bacteriophages revealed that 28.8% were positive. These 55 bacteriophages, 32% (33/103) from swine and 25% (22/88) from poultry fecal samples, were further studied and characterized. Spotting tests of 67 Salmonella strains showed a high variability of the bacteriophages, distinguishing 39 different lysis profiles (Table 1 and Fig. S1, Supporting Information) defined according to the absence or presence of lysis on the tested bacterial strains. Twenty lysis profiles represented the 22 bacteriophages isolated from poultry and 21 lysis profiles were distinguished from the 33 bacteriophages isolated from swine. Only two lysis profiles (numbers 8 and 22, Table 1) were shared by bacteriophages isolated from poultry and swine. Moreover, the majority of the bacteriophages (n = 51) showed a broad lysis profile (lysing more than 10 Salmonella strains). Among them, 35 bacteriophage isolates, representing 20 lysis patterns, lysed more than 50 Salmonella strains. The most common lysis profile was that of six bacteriophages isolated from swine. Only four bacteriophages lysed fewer than 10 bacterial strains. A bivariate analysis was conducted to determine any statistically significant shared features among the bacteriophage lysis profiles and the source and serovar of the lysed strains of Salmonella. In this analysis, those phages that were isolated from samples obtained at different times but nonetheless had identical lysis profiles were included. For bacteriophages with the same lysis patterns and isolated from samples collected on the same date, only one bacteriophage was chosen as representative. Based on these criteria, the statistical comparison was conducted with 46 out of 55 isolated bacteriophages. Bivariate analysis indicated that 41.3% of them (19 out of 46) infected a higher percentage of S. Enteritidis than S. Typhimurium strains
3
(p ≤ 0.01), regardless of the source of the bacterial strain (poultry, swine or humans). The 55 bacteriophages were further characterized by DNA restriction analysis. The combined analysis of the DNA patterns obtained with the enzymes BstEII, EcoRV, EcoRI, HindIII, PstI and XbaI produced 22 different restriction profiles (Table 2). The identification of 21 of them was possible with the enzymes EcoRV and XbaI. EcoRV had a greater discriminatory capacity, distinguishing 19 patterns, while bacteriophages which showed the pattern 11 (Fig. 1A) were further distinguished into patterns 1, 4 and 5 (Fig. 1B) with XbaI enzyme. Restriction profile 22 (Table 2) corresponded to one bacteriophage whose DNA was resistant to all the enzymes tested except BstEII. Two bacteriophages could not be characterized because their DNA was resistant to digestion by all of the restriction enzymes tested. Bacteriophages isolated from poultry and swine did not share any restriction patterns. Eleven restriction patterns were obtained from the 22 bacteriophages isolated from poultry and the remaining 11 from the 33 bacteriophages isolated from swine (Table 2). These results were confirmed in duplicate experiments. RAPD-PCR was used as an additional typing method to characterize the collection of Salmonella bacteriophages. This method was previously used to type bacteriophages from dif´ ferent bacterial species but not Salmonella (Gutierrez et al., 2011). The 55 bacteriophages in our study were typed in single reactions with three different primers, OLP5, P1 and P2. Primer P1 showed the highest discriminatory capacity, identifying 34 different PCR patterns; OLP5 and P2 yielded 31 and 24 patterns, respectively. A dendrogram generated by combining the results obtained with the three primers allowed the differentiation of 40 RAPD profiles (Fig. 2). Nine band patterns, associated with bacteriophage isolates from swine, differed only in one band, as determined in two different experiments (Fig. 2). No patterns common to bacteriophages from poultry and swine were identified after RAPD-PCR. Reproducible patterns were obtained with different bacteriophages DNA in separate experiments (data not shown). Regarding the typing characteristics of the methods used in this study, the typeability of lysis profiling was 100%, that of RAPD-PCR 100% as well and that of restriction analysis 96%. The discriminatory power (D-values of the Simpson’s diversity index) was 0.978 (± 0.016), 0.938 (± 0.028) and 0.982 (± 0.013) for lysis profiling, DNA restriction and RAPD-PCR, respectively. The highest concordance regarding the discriminative ability (Robinson et al., 1998) was between RAPD-PCR and lysis profiling (0.974), followed by DNA restriction and RAPD-PCR (0.884) and lysis profiling and DNA restriction (0.868).
DISCUSSION This study used lysis profiling, DNA restriction patterns and RAPD-PCR analysis to characterize the diversity of bacteriophages infecting the serovars Typhimurium and Enteritidis in poultry and swine. The bacteriophages were isolated from 28.8% of the 191 cloacal or rectal swabs collected from poultry and swine between 2007 and 2009. A higher percentage of bacteriophages (32%) was detected in swine than in poultry (25%). This finding is in agreement with previous reports on the prevalence of Salmonella in poultry and swine farms. Thus, in Spain, during the same period of time, the prevalence of Salmonella on swine farms was >50% and on poultry farms 25.3% (EFSA 2010). These results suggest that the presence of bacteriophages infecting Salmonella in farm animals roughly parallels that of the
4
FEMS Microbiology Letters, 2015, Vol. 362, No. 1
Table 1. Lysis profiles of the bacteriophages studied.
Lysis profile number
Number of bacteriophages by source
Number of lysed Salmonella strainsa
Poultry
Swine
Typhimurium
Enteritidis
1 2 3 4 5 6 7 8b 9 10 11 12 13 14 15 16 17 18 19 20 21 22b 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
– – – – – – 1 2 – – – – 1 1 – – 1 1 1 1 1 1 2 1 1 1 1 1 – – – – – 1 1 1 1 – –
1 3 4 2 6 1 – 1 1 1 1 1 – – 1 1 – – – – – 2 – – – – – – 1 1 1 1 1 – – – – 1 1
41 39 40 40 40 40 42 40 40 46 0 0 41 27 17 7 20 24 43 15 40 47 13 11 11 16 0 46 12 11 2 11 45 41 41 36 43 8 21
16 16 15 14 16 14 17 14 15 18 4 7 15 18 16 13 14 15 18 13 17 18 14 13 14 15 2 18 12 11 6 10 18 18 17 15 18 13 16
Number of total bacteriophages
22
33
Number of total lysis profiles
20
21
–Lysis profile not detected. a Number of lysed strains, from a total of 49 S. Typhimurium and 18 S. Enteritidis strains. b Lysis profile shared for bacteriophages from poultry and swine feces.
bacterium. It should be pointed out that the overall bacteriophage isolation efficiency in this study was higher than that reported by other authors also analyzing poultry and swine (Fiorentin et al., 2004; Toro et al., 2005). The higher efficiency can perhaps be attributed to the enrichment isolation procedure, which may have influenced the specific selection of Salmonella and therefore of the specific bacteriophages infecting these bacterial strains. The phenotypic and genotypic diversity of the 55 bacteriophages isolated was considerable, as indicated by the 39
different lysis profiles and further demonstrated by RAPD-PCR (40 profiles). By contrast, DNA restriction analysis had a lower discriminatory capacity, even though six different restriction enzymes were employed. However, this method detected the same number of profiles for bacteriophages isolated from poultry (n = 11) and from swine (n = 11), as identified by lysis profiling and RAPD-PCR. Although RAPD-PCR was previously used to type 26 bacteriophages infecting gram-positive bacteria and Escherichia ´ coli (Gutierrez et al., 2011), prior to our study it had not been compared with the other methodologies commonly used to type
´ et al. Cortes
Table 2. DNA restriction profiles of the studied bacteriophages, according to the combined analysis of the patterns obtained with the enzymes BstEII, EcoRV, EcoRI, HindIII, PstI and XbaI.
Restriction profile
a
Number of bacteriophages by source
numbera
Poultry
Swine
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 NDb
1 – – 1 – 1 – – – 1 – 1 5 6 1 1 – – 3 1 – – 0
– 4 6 – 9 – 1 1 1 – 1 – – – – – 3 1 – – 3 1 2
Number of total bacteriophages
22
33
Number of total profiles
11
11
All the bacteriophages belonging to a restriction profile showed the same restriction pattern for all the enzymes used. b ND, bacteriophages whose DNA could not be digested with any of the enzymes used.
5
bacteriophages. Our results show that this typing method can be applied to other gram-negative bacteria, with results comparable to lysis profiling, the conventional approach. Moreover, an advantage of RAPD-PCR is that it is rapid, because it can be used ´ directly with bacteriophage suspensions (Gutierrez et al., 2011). The high agreement between lysis profiling and RAPD-PCR supports the use of the latter method in bacteriophage diversity studies. Among the bacteriophages isolated from poultry, it was remarkable that many of the lysis and RAPD profiles represented only one or two of them, while for bacteriophages isolated from swine, there were profiles containing more than two bacteriophages. This suggests that there is a greater presence of some bacteriophage patterns in swine than in poultry. The difference in abundance could be related to the respective animal production systems, in that closed-cycle swine production is still very common in Spain whereas poultry farms rely almost entirely on a fattening/breeding system. Although the latter is a more specialized form of production, it can promote the greater local movement of microorganisms, including Salmonella and their bacteriophages. In addition, a greater diversity of bacteriophages may allow the development of new bacteriophage, through recombination processes. This mechanism has been shown to explain the mosaic structure of bacteriophage genomes and the generation of new gene combinations in bacteriophages (Casjens and Thuman-Commike 2011). Moreover, bacteriophage diversity seems linked to evolution of host specificity because bacteriophages infecting Salmonella strains from poultry and pigs present different lysis profiles. Interestingly, the bacteriophages isolated from poultry and swine did not overlap with respect to their restriction patterns and RAPD-PCR profiles and only two lysis profiles were shared (Tables 1 and 2; Fig. 2). However, the relationship between the source of the bacteriophages (poultry and swine) and that of the Salmonella strains infected by them (poultry and swine) was not statistically significant. Thus, poultry and swine carry different bacteriophages, which nonetheless infect Salmonella strains isolated from either of these animals.
Figure 1. Nineteen DNA restriction patterns (lines 1 to 19) obtained with EcoRV (A) and five restriction patterns (lines 1 to 5) distinguished with XbaI (B) for the 55 bacteriophages studied. Lambda DNA digested with HindIII (M1) and lambda DNA digested with BstEII mixed with φX714 DNA digested with HinfI (M2) were used as molecular markers.
6
FEMS Microbiology Letters, 2015, Vol. 362, No. 1
Figure 2. Dendrogram based on the combined analysis of RAPD patterns obtained with primers OLP5, P1 and P2. The letters P and S in the name of bacteriophages indicate the poultry and swine origin of the bacteriophages, respectively.
In conclusion, poultry and swine contain a remarkable diversity of bacteriophages and most of them are able to efficiently infect a great variety of non-clonal strains of S. Typhimurium and S. Enteritidis. Our results also demonstrate a high concordance between lysis profiling and RAPD-PCR, with the latter providing a useful, quick and easy method to assess the diversity of bacteriophages in the environment. In addition, RAPD-PCR may also be suitable for pre-screening bacteriophages to be applied for other purposes. In this sense, those bacteriophages having a broader lysis spectrum and with different RAPD patterns would be the best candidates for biocontrol and therapeutic strategies against Salmonella.
SUPPLEMENTARY DATA Supplementary data is available at FEMSLE online.
ACKNOWLEDGEMENTS We are deeply indebted to the Laboratori de Sanitat Animal (DAAM, Generalitat de Catalunya, Spain), the Santa Creu i Sant Pau Hospital, and the Vall d’Hebron Hospital for providing us with the different Salmonella strains used in this study and to ´ Laboratorios Callier and Laboratorio General de Diagnostico for providing the fecal samples.
´ et al. Cortes
FUNDING This work was supported by the Instituto Nacional de Investi´ y Tecnolog´ıa Agraria y Alimentaria; Ministerio de Ciencia gacion ´ Espanol ˜ (grant number RTA2006-00065) and from e Innovacion Generalitat de Catalunya (grant number SGR2014/572). CB is the recipient of a predoctoral fellowship from the Comissionat per ´ Universia Universitats i Recerca del Departament d’Innovacio, tats i Empresa de la Generalitat de Catalunya i del Fons Social Europeu. DS is supported by a predoctoral fellowship from CAPES ˜ de Aperfeic¸oamento de Pessoal de N´ıvel Superior), (Coordenac¸ao Brazil. Conflict of interest statement. None declared.
REFERENCES Atterbury RJ. Bacteriophage biocontrol in animals and meat products. Microb Biotechnol 2009;2:601–12. Atterbury RJ, Van Bergen MAP, Ortiz F, et al. Bacteriophage therapy to reduce Salmonella colonization of broiler chickens. Appl Environ Microb 2007;73:4543–9. Ausubel FM, Brent R, Kingston RE, et al. Current Protocols in Molecular Biology, Vol. 1. New York: Greene/Wiley-Interscience, 1987. ´ P, et al. Significance of the bacBardina C, Spricigo DA, Cortes teriophage treatment schedule in reducing Salmonella colonization of poultry. Appl Environ Microb 2012;78:6600–7. ´ Borie C, Albala I, Sanchez P, et al. Bacteriophage treatment reduces Salmonella colonization of infected chickens. Avian Dis 2008;52:64–7. Casjens SR, Thuman-Commike PA. Evolution of mosaically related tailed bacteriophage genomes seen through the lens of phage P22 virion assembly. Virology 2011;411:393–415. Center for Disease Control and Prevention. Salmonella: Annual Summary (2009) Atlanta, GA: CDC, Division of Bacterial and Mycotic Diseases, 2011. http://www.cdc.gov/ncezid/dfwed/PDFs/ SalmonellaAnnualSummaryTables2009.pdf Chan BK, Abedon ST, Loc-Carrillo C. Phage cocktails and the future of phage therapy. Future Microbiol 2013;8:769–83. Connerton PL, Loc Carrillo CM, Swift C, et al. Longitudinal study of Campylobacter jejuni bacteriophages and their hosts from broiler chickens. Appl Environ Microb 2004;70:3877–83. European Food Safety Authority. The Community summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in the European Union in 2008. EFSA J 2010;8:1496. European Food Safety Authority. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2010. EFSA J 2012;10:2597. Fiorentin L, Vieira ND, Barioni W, Jr, et al. In vitro characterization and in vivo properties of salmonellae lytic bacteriophages isolated free-range layers. Braz J Poult Sci 2004;6:121–8. Fiorentin L, Vieira ND, Barioni W, Jr. Oral treatment with bacteriophages reduces the concentration of Salmonella Enteri-
7
tidis PT4 in caecal contents of broilers. Avian Pathol 2005a;34: 258–63. Fiorentin L, Vieira ND, Barioni W, Jr. Use of lytic bacteriophages to reduce Salmonella Enteritidis in experimentally contaminated chicken cuts. Braz J Poult Sci 2005b;7:255–60. Goyal SM, Gerba CP, Bitton G. Phage Ecology. New York, NY: John Wiley and Sons, 1987, 321. ´ Gutierrez D, Mart´ın-Platero AM, Rodr´ıguez A, et al. Typing of bacteriophages by randomly amplified polymorphic DNA (RAPD)-PCR to assess genetic diversity. FEMS Microbiol Lett 2011;322:90–7. Hendrix RW. Bacteriophage genomics. Curr Opin Microbiol 2003;6:506–11. Hooton SPT, Atterbury RJ, Connerton IF. Application of a bacteriophage cocktail to reduce Salmonella Typhimurium U288 contamination on pig skin. Int J Food Microbiol 2011;151: 157–63. Hunter PR, Gaston MA. Numerical index of the discriminatory ability of typing systems: an application of Simpson’s index of diversity. J Clin Microbiol 1988;26:2465–6. Moreno Switt AI, den Bakker HC, Vongkamjan K, et al. Salmonella bacteriophage diversity reflects host diversity on dairy farms. Food Microbiol 2013;36:275–85. Muniesa M, Blanch AR, Lucena F, et al. Bacteriophages may bias outcome of bacterial enrichment cultures. Appl Environ Microb 2005;71:4269–75. Pao S, Randolph SP, Westbrook EW, et al. Use of bacteriophages to control Salmonella in experimentally contaminated sprout seeds. J Food Sci 2004;69: M127–30. Robinson DA, Hollingshead SK, Musser JM, et al. The IS1167 insertion sequence is a phylogenetically informative marker among isolates of serotype 6B Streptococcus pneumoniae. J Mol Evol 1998;47:222–9. Sharma M, Patel JR, Conway WS, et al. Effectiveness of bacteriophages in reducing Escherichia coli O157:H7 on fresh-cut cantaloupes and lettuce. J Food Protect 2009;72:1481–5. ´ P, et al. Use of a bacteriophage Spricigo DA, Bardina C, Cortes cocktail to control Salmonella in food and the food industry. Int J Food Microbiol 2013;165:169–74. the Members of the European Study Group on epidemiological Markers (ESGEM) of the European Society for Clinical Microbiology and Infectious Diseases (ESCMID)Struelens MJ, . Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin Microbiol Infect 1996;2:2–11. Sulakvelidze A, Alavidze Z, Morris JG, Jr. Bacteriophage Therapy. Antimicrob Agents Ch 2001;45:649–59. Toro H, Price SB, McKee AS, et al. Use of bacteriophages in combination with competitive exclusion to reduce Salmonella from infected chickens. Avian Dis 2005;49: 118–24. Wang S, Zhao W, Raza A, et al. Prevalence of Salmonella infecting bacteriophages associated with Ontario pig farms and the holding area of a high capacity pork processing facility. J Sci Food Agr 2010;90:2318–25.
