Salmonella Bacteriophage Diversity According to Most Prevalent Salmonella Serovars in Layer and Broiler Poultry Farms from Eastern Spain

Simple Summary There is a lack of knowledge about the impact that phages present in the environment may have against certain Salmonella serovars. Thus, an improved understanding of Salmonella phage diversity will provide a better insight into the role of phages in Salmonella ecology and diversity. The results of this study showed that the poultry farm environment could represent a valuable source of Salmonella phages, which are more varied in broiler than in layer farms. Abstract The exploration of novel nonantibiotic interventions in the field, such as the use of bacteriophages, is necessary to avoid the presence of Salmonella. Bacteriophages are a group of viruses widely distributed in nature, strictly associated with the prokaryotic cell. Researchers have demonstrated the success of phage therapy in reducing Salmonella counts in poultry products. However, the impact that phage concentration in the environment may have against certain Salmonella serovars is not well understood. Therefore, the aim of this study was to assess Salmonella phage prevalence in commercial poultry farms in terms of the production type: layers or broilers. The most prevalent Salmonella serovars isolated in poultry production were used for phage isolation. Salmonella specific phages were isolated from 141 layer and broiler farms located in the Valencia region during 2019. Analysis of the samples revealed that 100% presented Salmonella phages, the most prevalent being against serovar S. Enteritidis (93%), followed by S. Virchow (59%), S. Typhimurium (55%), S. Infantis (52%) and S. Ohio (51%). These results indicate that poultry farms could represent an important source of Salmonella phages. Nevertheless, further studies are needed to assess the epidemiology of phages against other serovars present in other countries and their diversity from the point of view of molecular studies.


Introduction
Salmonella spp. remain one of the main bacteria involved in food-borne outbreaks and are a major public health hazard worldwide [1]. It is estimated that nontyphoidal Salmonella worldwide cause around 94 million cases of illness and 155,000 deaths per year [2]. The latest data published by the European Food Safety Authority (EFSA) reported 91,857 human cases, 43.2% of which included hospitalization [3].
There are numerous sources of human salmonellosis infection, but eggs and poultry meat are reported to be the most common sources [3]. The latest data recorded in 2019 showed that 4% of tested flocks were positive for Salmonella detection, from which 1.1% were S. Enteritidis and S. Typhimurium target serovars [3]. However, among those outside the target serovars, the most common reported was S. Infantis. Considering the production chain for meat and meat products, the highest percentages of positive samples were found for fresh broilers meat, with S. Enteritidis, S. Typhimurium, and the S.a monophasic Typhimurium variant [3] as the main serovars involved in human outbreaks. In this line, the introduction of National Salmonella Control Programmes (NSCP) to control the bacterium at the field level resulted in an important reduction in the prevalence of poultry Salmonella serovars in Europe [4]. However, total elimination of the bacterium from poultry flocks is still difficult, and new cases of salmonellosis emerge every year, resulting in economically significant losses for the poultry sector [3].
In addition, the emergence of several Salmonella serovars resistant to multiple antibiotics in poultry-derived products underscores a significant food safety and poultry production hazard [5]. For this reason, the exploration of novel nonantibiotic interventions in the field should be studied to avoid the presence of antibiotic-resistant strains [5].
Bacteriophages or phages are a group of viruses widely distributed in nature, whose life cycle is strictly associated with the prokaryotic cell [6,7]. The use of host-specific phages has been promoted as a cost-effective and adaptable approach to control zoonotic bacteria [8][9][10][11]. Moreover, phages seem to be a good alternative due to their self-perpetuating, self-limiting and specificity characteristics [12]. Researchers have demonstrated the success of phage therapy in poultry products, reducing Salmonella counts from broiler carcasses after phage administration. Higgins et al. (2005) reduced Salmonella counts in 100% of broiler carcasses where phages were inoculated [13]. Moreover, Kang et al. (2013) decreased Salmonella counts on chicken skin by up to 3 logs after the application of a single phage [14]. Other research showed Salmonella decreasing counts by 1 log on fresh egg shells after application of the phage [7].
However, the impact that phages present in the environment may have against certain Salmonella serovars with relevance in food safety is not well understood. Thus, an improved understanding of Salmonella phage diversity may provide better insights into the role of phages in Salmonella ecology and diversity and facilitate an improved approach toward biocontrol and diagnostics [15,16].
The aim of this study was, therefore, to assess Salmonella phage diversity in commercial layer and broiler poultry farms in relation to the most prevalent serovars in the poultry production system in Eastern Spain. Thus, in this study we tested whether occurrence of phages against Salmonella was related to the poultry production type.

Salmonella Strain Selection for Phage Isolation
Salmonella strains used for phage detection were field strains selected from the strain collection repository from the Centro de Calidad Avícola y Alimentación Animal de la Comunidad Valenciana (CECAV), which is the benchmark laboratory for Salmonella field strains isolation from poultry farms throughout Spain. The origin of the field strains was the NSCP [4], and each selected strain used in this study was isolated from poultry farms. All selected serovars were those most prevalent in poultry production in Spain [3]: S. Enteritidis, S. Typhimurium, S. Typhimurium monophasic variant, S. Kentucky, S. Hadar, S. Senftenberg, S. Ohio, S. Infantis and S. Virchow. The strains were thawed and revived on nutrient agar (Oxoid Ltd., England, UK) and incubated at 37.5 ± 2 • C for 18 ± 4 h. For characterization of the strains, the antimicrobial susceptibility pattern was performed. To this end, Salmonella sensititre plates (Gram Negative MIC Plate) were used to assess antimicrobial susceptibility of isolated strains. A 10 µL aliquot of the inoculum was aseptically transferred to 10 mL sensititre cation-adjusted Mueller-Hinton broth, and plaques were inoculated according to manufacturer instructions. Plates were read at 18 h to 24 h manually by visualization of a growth button on the bottom of the microtitre well using a light box. Reading the results was performed according to the manufacturer's instructions.

Study Sample
A total of 141 poultry farms located in the Eastern Spain were sampled: 108 layer farms (from 41 to 64 weeks of rearing) and 33 broiler farms (ranging from 35 to 42 days of rearing), all of them belonging to three of the main companies in Spain that handle the majority of the broilers and layers reared in Spain (one company from broiler and two companies from laying hens). Farms selected for the study were conventional commercial poultry farms of broilers and layers. All flocks of laying hens analyzed (lines Lohmann and Hyline) were vaccinated against Salmonella according to the standard vaccination guidelines. To this end, Salmonella vaccination was performed with the vaccine Salmovac 440, a live vaccine given orally in a triple dose through water (day 1, week 6 and week 15) to protect against S. Enteritidis and S. Typhimurium serovars according to mandatory regulations in the Valencia region [19]. Moreover, layers were reared in cages with a density of 750 cm 2 /hen. With respect to broiler production, all analyzed flocks (lines Cobb and Ross) were reared on the floor in cages containing wood shavings to a depth around 10 cm, and with a density of 33 kg/m 2 . All the animals were kept indoors under controlled conditions equipped with programmable electrical lights, automated electric heating and forced ventilation [20].

Faeces Samples Collection
From each farm, two faeces samples of 150 g were taken from different points of the facility [4]. Once in the laboratory, faeces samples collected from each farm were pooled and placed in sterile pots: 25 g to assess Salmonella status of the farm and 10 g for phage detection (as described below).

Salmonella Phage Isolation
Phages were isolated from faeces collected by an enrichment procedure [22]. To do so, 10 g of each faeces sample were diluted in 90 mL of Luria Bertani (LB) (VWR Chemicals, Barcelona, Spain) and incubated along with each selected Salmonella serovar overnight at 37 • C. After incubation, 2 mL of this enrichment culture was centrifuged 16,000× g for 5 min. The supernatant was then filtered through a 0.22 µm membrane.
Phages were isolated and purified in a spot test by the double agar method. Briefly, bacterial suspensions of each serovar were adjusted to an optical density at 600 nm (OD = 600) of 0.2 (~10 8 CFU/mL) in LB and incubated at 37 • C for 4 h. Then, 200 µL of cultures were added to 5 mL of LB agar (LB with 0.6% agar) tempered to 45 • C and poured onto previously prepared and dried LB basal agar (with 1.6% agar). Then, 10 uL of each filtrate were spotted onto the surfaces of Salmonella lawns and incubated overnight at 37 • C. After the incubation, morphologically different plaques were selected and resuspended in 1 mL of PBS. Ten-fold serial dilutions of the phage suspension were plated by the double agar layer method, and phages that produced clear plaques were selected. This procedure was repeated three times to obtain a single type of phage [23].

Statistical Analysis
We tested whether occurrence of phages against Salmonella was related to the poultry production system. To do so, we fitted a generalized linear model (GLM) where occurrence of Salmonella phage was the response variable and the sample type (faeces from different broiler and layer farms), Salmonella serovar (n = 9), genetic lines (n = 2, for both poultry production type), poultry companies (n = 1 and n = 2, for broilers and layers, respectively), husbandry (n = 1), Salmonella vaccine strain (n = 1) were the factors.
For this analysis, the error was designated as having a binomial distribution and the probit link function was used. Binomial data for each sample were assigned a 1 if a Salmonella phage was isolated or a 0 if not. A p-value < 0.05 was considered to indicate a statistically significant difference. Differences in binomial traits for variables, genetic lines, poultry companies, husbandry and Salmonella vaccine strain, were not significant and were excluded from the model. Finally, a descriptive analysis of the patterns obtained against different Salmonella serovars per farm, and antimicrobial resistance of the strains, was carried out. Analyses were carried out using a commercially available software program (SPSS 21.0 software package; SPSS Inc., Chicago, IL, USA, 2002).

Results
In this study, a total of 141 faeces samples were collected from poultry farms. From each, 141 pools of 25 g were analyzed to assess Salmonella status of the farm, and 1269 analyses were done for specific phages isolation (farm × serovar) ( Figure 1). No Salmonella was detected in any farm, although Salmonella phages were detected in all farms sampled, at least against one of the serovars included in this study.
Animals 2020, 10, x 4 of 12 agar (with 1.6% agar). Then, 10 uL of each filtrate were spotted onto the surfaces of Salmonella lawns and incubated overnight at 37 °C. After the incubation, morphologically different plaques were selected and resuspended in 1 mL of PBS. Ten-fold serial dilutions of the phage suspension were plated by the double agar layer method, and phages that produced clear plaques were selected. This procedure was repeated three times to obtain a single type of phage [23].

Statistical Analysis
We tested whether occurrence of phages against Salmonella was related to the poultry production system. To do so, we fitted a generalized linear model (GLM) where occurrence of Salmonella phage was the response variable and the sample type (faeces from different broiler and layer farms), Salmonella serovar (n = 9), genetic lines (n = 2, for both poultry production type), poultry companies (n = 1 and n = 2, for broilers and layers, respectively), husbandry (n = 1), Salmonella vaccine strain (n = 1) were the factors.
For this analysis, the error was designated as having a binomial distribution and the probit link function was used. Binomial data for each sample were assigned a 1 if a Salmonella phage was isolated or a 0 if not. A p-value < 0.05 was considered to indicate a statistically significant difference. Differences in binomial traits for variables, genetic lines, poultry companies, husbandry and Salmonella vaccine strain, were not significant and were excluded from the model. Finally, a descriptive analysis of the patterns obtained against different Salmonella serovars per farm, and antimicrobial resistance of the strains, was carried out. Analyses were carried out using a commercially available software program (SPSS 21.0 software package; SPSS Inc., Chicago, IL, USA 2002).

Results
In this study, a total of 141 faeces samples were collected from poultry farms. From each, 141 pools of 25 g were analyzed to assess Salmonella status of the farm, and 1269 analyses were done for specific phages isolation (farm × serovar) (Figure 1). No Salmonella was detected in any farm, although Salmonella phages were detected in all farms sampled, at least against one of the serovars included in this study.

Salmonella Antimicrobial Susceptibility Characterization
From different Salmonella serovars from the poultry sector included in this study (n = 9), 56% were resistant to at least one of the fourteen antibiotics tested, and 44% were MDR to 3 or more of the groups of antibiotics tested. Salmonella serovars MDR were S. Typhimurium monophasic variant, S. Typhimurium and S. Virchow. The highest percentages of antimicrobial resistance (AMR) were found to be TET (44%) followed by AMP (33%), NAL (33%), SMX (22%) TMP (11%), and CHL (11%). Resistance to MERO, AZM, TGC, CAZ, COL, GN and CTX was not observed.

Discussion
The diversity of Salmonella phages in poultry farms regarding their production type (broilers or layers) and the most prevalent Salmonella serovars in the Eastern Spain were analyzed in this study. Although Salmonella spp. were not present in any of the farms assessed, phages from several serovars of public health and poultry production importance were present in 100% of the samples collected. These results showed that although the bacterium is not present in the farm environment, its specific phages can remain in it.
It is claimed that AMR will be the main cause of deaths worldwide by 2050, overtaking other major causes of deaths such as cancer or road traffic accidents [24,25]. For this reason, the reduction of antimicrobial use at the field level throughout Europe is one of the most important aims in the poultry sector [26]. Results of this study showed Salmonella strains with a high percentage of antimicrobial resistance, especially against TET, AMP and NAL, three of the antibiotics most frequently used to treat poultry, and also used against human diseases [27,28]. Although Salmonella treatment with antibiotics is banned in the EU, its resistance to antibiotics could be acquired from different sources, such as the environment or antibiotics used to control other infections (E. coli) [29]. Phage patterns obtained against different Salmonella serovars per farm in this study indicated that the environment of animal farms, especially poultry operations, could represent an important source of Salmonella phages against several serovars [30,31]. In this sense, the phages obtained could be effective to combat these antibiotic-resistant strains, with the aim of controlling Salmonella AMR and its spread to the food chain [32].
Regarding Salmonella phages per serovar and poultry production type, S. Enteritidis, S. Typhimurium and S. Typhimurium monophasic variant phages were three of the phages most frequently isolated in poultry farms. This could be explained by the strict vaccination programs implemented in the poultry production system. Vaccination against S. Enteritidis is mandatory in all commercial layer flocks, and optional for layer and broiler breeders [33]. In addition, the vaccination programme is stricter in the Valencia region where, since 2008, it is mandatory to vaccinate not only against S. Enteritidis, but also against S. Typhimurium [19]. Live vaccination in poultry maintains the Salmonella vaccine strain in birds, as well as the house environment [34][35][36], and could encourage phage presence in the field. In this context, the latest data recovered from official checks in the

Discussion
The diversity of Salmonella phages in poultry farms regarding their production type (broilers or layers) and the most prevalent Salmonella serovars in the Eastern Spain were analyzed in this study. Although Salmonella spp. were not present in any of the farms assessed, phages from several serovars of public health and poultry production importance were present in 100% of the samples collected. These results showed that although the bacterium is not present in the farm environment, its specific phages can remain in it.
It is claimed that AMR will be the main cause of deaths worldwide by 2050, overtaking other major causes of deaths such as cancer or road traffic accidents [24,25]. For this reason, the reduction of antimicrobial use at the field level throughout Europe is one of the most important aims in the poultry sector [26]. Results of this study showed Salmonella strains with a high percentage of antimicrobial resistance, especially against TET, AMP and NAL, three of the antibiotics most frequently used to treat poultry, and also used against human diseases [27,28]. Although Salmonella treatment with antibiotics is banned in the EU, its resistance to antibiotics could be acquired from different sources, such as the environment or antibiotics used to control other infections (E. coli) [29]. Phage patterns obtained against different Salmonella serovars per farm in this study indicated that the environment of animal farms, especially poultry operations, could represent an important source of Salmonella phages against several serovars [30,31]. In this sense, the phages obtained could be effective to combat these antibiotic-resistant strains, with the aim of controlling Salmonella AMR and its spread to the food chain [32].
Regarding Salmonella phages per serovar and poultry production type, S. Enteritidis, S. Typhimurium and S. Typhimurium monophasic variant phages were three of the phages most frequently isolated in poultry farms. This could be explained by the strict vaccination programs implemented in the poultry production system. Vaccination against S. Enteritidis is mandatory in all commercial layer flocks, and optional for layer and broiler breeders [33]. In addition, the vaccination programme is stricter in the Valencia region where, since 2008, it is mandatory to vaccinate not only against S. Enteritidis, but also against S. Typhimurium [19]. Live vaccination in poultry maintains the Salmonella vaccine strain in birds, as well as the house environment [34][35][36], and could encourage phage presence in the field. In this context, the latest data recovered from official checks in the Valencia region showed that 100% of S. Enteritidis strains isolated from rearing layers were S. Enteritidis vaccine strains (unpublished data). Moreover, specific phages against S. Typhimurium monophasic variant have been found, which may be explained by the mandatory oral administration of S. Typhimurium vaccine, which could provide cross-immunization against S. Typhimurium monophasic variant [37].
A high prevalence of phages against S. Ohio, S. Infantis and S. Virchow have been found in this study; these are three of the main serovars isolated in the Valencia region from the NSCP (unpublished data). These results are in line with other researchers, who stated that the presence of phages in the farm environment would suggest the bacterial strain has been present at some point in the recent past [12,38]. In addition, this fact could be used for the indirect detection of pathogens based on their specificity towards bacteria [32,39]. In this line, phages against S. Virchow, S. Hadar and S. Ohio were observed to be more prevalent in broilers than in layers. These results are in accordance with data recovered from the Salmonella control programme in the Valencia region, as neither S. Virchow nor S. Hadar were isolated from laying farms (unpublished data). Moreover, Marin and Lainez (2009) also demonstrated that the main serovars isolated from broiler farms in the Valencia Region were S. Virchow, S. Ohio and S. Hadar [40]. On the other hand, no statistical differences have been found between the poultry production type and the presence of phages against S. Enteritidis, S. Typhimurium, S. Infantis, and S. Senftenberg. This result could be related to the historically close relationship between these serotypes and both layer and broiler production systems [41].

Conclusions
In conclusion, the results of this study showed that the poultry farm environment could represent a valuable source of Salmonella phages. A wide Salmonella phage diversity was present in the broiler and layer farms analyzed, being more varied in broilers. Nevertheless, further studies are needed to study the epidemiology of phages against other serovars present in other countries and its diversity from the point of view of molecular studies.

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript: