Dogs as a source of Salmonella spp. in apparently healthy dogs in the Valencia Region. Could it be related with intestinal lactic acid bacteria?

Background Although salmonellosis is considered one of the most important food-borne zoonotic diseases in Europe, close contact between dogs and their owners can also be a potential source of Salmonella spp. for humans. This study assessed the prevalence and antimicrobial resistance of Salmonella spp. in apparently healthy dogs in the Valencian Region, eastern Spain. Moreover, a macroscopic comparison of lactic acid bacteria in both Salmonella-positive and Salmonella-negative dogs was carried out. Results Of a total of 325 dogs sampled, 6 (1.85%) were positive for Salmonella spp. with 3 different serotypes, Havana (3), Mikawasima (2) and monophasic Typhimurium (1). All isolates were susceptible to all antimicrobials tested except monophasic S. Typhimurium, which was resistant to ampicillin. Finally, macroscopic results revealed that lactic acid bacteria had higher heterogeneity in the Salmonella-negative dogs than in the Salmonella-positive dogs. Although the results in our study showed a low prevalence of Salmonella spp., raw food has been suggested as a risk factor for bacteria in dog faeces. Conclusions Public awareness campaigns on good hygiene practices, especially after handling canine faeces or raw food, are necessary. Furthermore, to reduce the potential transmission of bacteria, dogs should be fed food that has been properly cooked, as raw or undercooked food can be a source of zoonotic pathogens. Moreover, further studies must be performed to determine the relationship between lactic acid bacteria and Salmonella spp. in dog faeces.

with more prevalent serotypes in dogs [11,12]. In addition, companion animals could be an important source of multidrug-resistant Salmonella spp. [13]. Therefore, pet dogs should be considered a public health risk [4]. Children, elderly and immunocompromised individuals have a higher risk of salmonellosis because they are more susceptible to infection [2,11,14,15].
Since the 1970s, previous studies have reported the zoonotic transmission of Salmonella spp. from dogs [16]. Nevertheless, the prevalence of this bacterium in dogs varies considerably. First, dogs usually act as asymptomatic carriers, and they can shed one or more serotypes intermittently for more than 6 weeks [11,17]. Moreover, dogs can harbour Salmonella spp. in the intestine and mesenteric lymph nodes without evidence of clinical signs [10,11,18]. Different studies have shown a wide prevalence of Salmonella spp. that oscillates between 0 and 79% [11,12,[18][19][20][21][22][23][24][25]. This variability makes establishing the real prevalence among dog populations a challenge.
Other factors that have been reported to influence the prevalence of Salmonella spp. in dogs are the environment where animals live or contact with wild animals or other infected animals [11,18,26]. Additionally, animal feeding has been described as one of the main risk factors for the prevalence of Salmonella spp. in dogs [11]. Feeding dogs with raw food has also been related to the high prevalence of salmonellosis [12,20,21,27,28].
Another factor related to the increase in Salmonella spp. in animals could be associated with the alteration of normal microbiota, which protects the gastrointestinal tract from colonization by pathogens; alterations in the normal microbiota could provide a suitable environment for bacterial replication [29,30].
As mentioned above, the prevalence of Salmonella serotypes in clinically healthy dogs varies notably and may even be different depending on the country [11]. As far as we know, no studies on the prevalence of Salmonella spp. in dogs have been performed in Spain. Therefore, the objective of this study was to determine the presence of Salmonella spp. in asymptomatic dogs housed in different environmental conditions in the Valencian Region (southern Spain). Moreover, the antimicrobial drug susceptibility of the isolates was determined, and macroscopic analysis of gastrointestinal lactic acid bacteria (LAB) in Salmonella spp.-positive and Salmonella spp.negative animals was also evaluated.

Results
The prevalence of Salmonella spp. in apparently healthy dogs The number of dogs sampled and the prevalence of Salmonella spp. is shown in Table 1 Regarding feeding, of the 325 dogs analysed, 196 ate commercial pet food (60.3%), and the remaining dogs were also fed home-prepared food (39.7%). Moreover, 36 dogs (11.1%) were fed raw food. The type of water source was also asked in the questionnaire; 11 dogs (3.4%) drank bottled water, while 269 (82.8%) drank running water and 45 (13.8%) drank from a water well. Finally, whether the dogs were in contact with other animal species was also asked in the questionnaire, and 178 dogs of 325 (54.8%) were in contact with wild animals. Independent of the environment where dogs lived, only 6 (1.85%) of the 325 rectal swabs analysed were positive for Salmonella spp. No significant differences were found between the percentage of Salmonella spp. and the age, sex, type of food, water source or contact with wild animals. However, the prevalence of Salmonella spp. was higher in animals that were fed raw food compared to those that did not eat raw food, and this difference was statistically significant (p value ≤0.01).

Antimicrobial susceptibility of Salmonella isolates
The antimicrobial susceptibility of Salmonella serotypes is shown in Table 2. All serotypes were susceptible to all antimicrobials tested except monophasic S. Typhimurium, which was resistant to ampicillin.
Macroscopic LAB differences between Salmonella-positive and Salmonella-negative dogs LAB from Salmonella-negative dogs (n = 319) and Salmonella-positive dogs (n = 6) were grown on MRS agar. Macroscopic analysis was carried out, and subjective differences were observed. Figures 1 and 2 show the macroscopic profile of LAB from the Salmonella-negative and Salmonella-positive dogs, respectively. As an initial approximation, Salmonella-negative animals seemed to have more LAB than Salmonella-positive animals.

Discussion
Although dogs may not play as large of a role in the transmission of salmonellosis to people as food, dogs have long been known to be asymptomatic carriers of Salmonella serotypes [31]. Their close relationship and frequent contact with humans, especially children, may make these animals a potential source of Salmonella spp. for humans and thereby represent an important public health issue [2,10,11,18].
The isolation of Salmonella spp. from clinically healthy dogs has been reported to be between 0 and 79% [11,21,23,24,27]. The results obtained in the present study showed a prevalence of Salmonella spp. of 1.8% in apparently healthy dogs in Valencian Region (Southern Spain). Our results were in accordance with the literature. In most studies, the prevalence of Salmonella-positive dogs was below 3% [12]. Nevertheless, previous studies have shown that the prevalence of Salmonella in asymptomatic dogs can vary geographically [10,11,18,19,22,23,25]. Factors associated with these variations could be in part related to differences in the sample sizes, faecal sampling conditions and isolation and detection methods employed [11,23]. In the present study, dogs were sampled only once. The low prevalence observed in our study could be related to the fact that only one faecal sample was collected, as intermittent shedding of Salmonella spp. has been well documented [11,17]. However, rectal swab samples were employed in other studies to determine the presence of Salmonella in dog faeces; these studies observed similar results and a higher prevalence of Salmonella [32,33].
The prevalence of Salmonella spp. in dogs has also been reported to depend on the immediate environment in which the animals live [11]. For example, the rates of isolating Salmonella from stray dogs have been reported to be significantly higher than those from household dogs [32]. In the current study, samples were collected from dogs housed in three different environments: households, animal shelters and hunting kennels. The    [12]. They found that there was no significant statistical association with Salmonella status and living or exposure with other animals, age, exposure to water sources, hunting or performing sport activities. However, different studies have shown that the prevalence of Salmonella in dogs exposed to contaminated environments including hunting dogs or stray dogs is higher than that in household dogs [12,23,32,33]; this result could be because dogs are more likely to roam free, scavenge, be in contact with carcasses or offals of wildlife and be fed raw or undercooked food [23,33,34]. As mentioned, differences in the prevalence of Salmonella can also vary depending on feeding practices [11]. Feeding pet raw meat-based diets (RMBDs) has received increasing attention in recent years [35]. Our results showed that 4 of 6 dogs positive for Salmonella were fed raw poultry carcasses. These results are in accordance with other studies that concluded that consuming raw diets increases the risk of carrying Salmonella spp. [12,20,21,24,27,36].
In our study, S. Havana were found in the faeces of Salmonella-positive dogs fed raw chicken carcasses. This serotype has been isolated previously from poultry products [46]. Pace et al. [38] determined a possible human infection caused by S. Havana from dog food contamination. S. Mikawasima has also been found in both wild boar from cattle-free and cattle-grazed areas in Spain [47], and in our study, S. Mikawasima was isolated from one hunting dog in contact with these species. Moreover, there has been an unusual increase in the number of S. Mikawasima infections in humans [39,44]. Therefore, these results underline the potential public health risk of dog-and pet-contaminated foods [12,22,34].
In addition, a major problem in terms of both animal and human health is the emergence and increase in antibiotic resistance [34]. In fact, Salmonella has been listed by the World Health Organization (WHO) as one of the antibiotic-resistant priority pathogens and has become a worldwide health issue [48]. The results obtained in the present study showed that all serotypes isolated in our study were susceptible to all antibiotics tested. This is in accordance with other studies [11,12,[49][50][51]. Only monophasic S. Typhimurium was resistant to ampicillin.

Other authors have reported the isolation of resistant
Salmonella strains from dogs [11,12,32,52]. In addition, adding raw animal products to pet dog diets has been identified as a risk factor for the presence of antimicrobial-resistant Salmonella spp. [11,53].
New strategies are urgently needed to manage antimicrobial resistant infections [34,48,53]. Among the various methods to reduce the use of antibiotics [54], probiotics are expected to be an alternative intervention measure to prevent bacterial infection in dogs [55]. To identify beneficial probiotic bacteria that inhibit pathogens, a better understanding of the dog microbiome is needed [56][57][58]. As a preliminary approach, we compared LAB in samples from both Salmonella spp.-positive and Salmonella spp.-negative animals, and it seems that macroscopic differences were observed. Commensal bacteria maintain the stability of the digestive tract and can help prevent intestinal infections by modulating the immune response and inhibiting certain pathogens such as Salmonella spp. that cause infections [58,59]. The subjective differences observed at the macroscopic level in our study could suggest that some LAB may protect against the pathogen or, on the other hand, the presence of Salmonella spp. in dogs could inhibit beneficial LAB strains. Our preliminary results could be in accordance with other studies that observed that gut LAB in pigs inhibit the growth of Salmonella and prevent the adhesion of bacteria in the intestinal tract [60].
In that sense, further studies must be carried out to address the limitations of our study and analyse whether these differences could be related to Salmonella spp. infection. It would be interesting to identify whether some species of LAB in Salmonella spp.-negative dogs are able to inhibit Salmonella spp., which could explain the absence of infection. Host species specificity is considered a requirement for probiotics, especially as a therapy for GI disorders [58]. In that sense, the study of dog microbiota may lead to the discovery of new therapies that can act against pathogens such as Salmonella spp.

Conclusion
In conclusion, the results obtained in this study indicate that apparently healthy dogs can act as subclinical carriers of Salmonella spp. in the Valencian region, eastern Spain. This could be significant to public health, as dogs are in close contact with humans. However, further studies must be performed because it is known that the shedding of Salmonella in dog faeces is influenced by several factors, such as diet, sampling procedures, and geographic area. Moreover, additional studies must be performed to determine the relationship between LAB and Salmonella spp. in dog faeces.

Sample collection and questionnaire
All animals were handled according to the principles of animal care published by Spanish Royal Decree 53/2013 [61]. Sample collection was carried out in the Valencian Region (eastern Spain) between October 2017 and June 2018. A total of 325 healthy dogs housed in three different environments were sampled (dogs were housed in households or animal shelters or were hunting dogs). Rectal specimens were collected using sterile cotton swabs (Cary Blair sterile transport swabs, Deltalab, Barcelona, Spain) by rotating the swab inside the rectum of the dog and then transported under refrigeration to the Laboratory of "Agentes microbiológicos asociados a la Reproducción Animal-ProVaginBIO", UCH-CEU University, for Salmonella spp. isolation.
Data from each dog were also collected by a questionnaire developed for this study (Additional file 1) to determine the possible risk factors for Salmonella spp. infection, especially those related to the environment where animals were housed, the diet or type of food, contact with other animal species and the source of water. Other data were also included in this questionnaire, including gender and age. All questionnaires were completed and submitted together with the samples to the laboratory.
Salmonella spp. isolation, serotyping and determination of antibiotic susceptibility.

Lactic acid bacteria isolation
LAB isolation was carried out using the same rectal samples collected for Salmonella spp. isolation. Swabs were homogenized in BPW, and 100 μl was inoculated in the medium used for identification of LAB, especially Lactobacillus Man, Rogosa and Sharpe (MRS agar) (Scharlab, Barcelona, Spain) [65,66], and incubated for 24-48 h under anaerobic conditions. After incubation, subjective macroscopic observation was carried out.

Statistical analysis
Statistical analysis was performed with the statistical package R Commander and RcmdrPlugin. The associations between Salmonella occurrence; categorical factors were compared using Pearson's χ2 test, and the confidence intervals for prevalence estimates were calculated using the Wilson score interval method. A p-value < 0.01 was reported as statistically significant.