Zoonotic potential and prevalence of Salmonella serovars isolated from pets

ABSTRACT Salmonellosis is a global health problem, affecting approximately 1.3 billion people annually. Most of these cases are related to food contamination. However, although the majority of Salmonella serovars are pathogenic to humans, animals can be asymptomatic carriers of these bacteria. Nowadays, a wide range of animals is present in human households as pets, including reptiles, amphibians, dogs, cats, ornamental birds, and rodents. Pets contaminate the environment of their owners by shedding the bacteria intermittently in their feaces. In consequence, theyare thought to cause salmonellosis through pet-to-human transmission. Each Salmonella serovar has a different zoonotic potential, which is strongly regulated by stress factors such as transportation, crowding, food deprivation, or temperature. In this review, we summarize the latest reports concerning Salmonella-prevalence and distribution in pets as well as the risk factors and means of prevention of human salmonellosis caused by contact with their pets. Our literature analysis (based on PubMed and Google Scholar databases) is limited to the distribution of Salmonella serovars found in commonly owned pet species. We collected the recent results of studies concerning testing for Salmonella spp. in biological samples, indicating their prevalence in pets, with regard to clinical cases of human salmonellosis.

Pet animals; zoonotic transmission; salmonella serovars distribution; pet regulations

Background
Keeping pets provides health, emotional and social benefits to their owners. According to the American Pet Products Association's 2019-2020 National Pet Owners Survey, approximately 67% of households and 85 million families own at least one pet [1,2]. It was estimated that 63.4 million and 42.7 million households own dogs and cats, respectively, with 5.7 million and 4.5 million pet owners keeping ornamental birds and reptiles [3]. In contrast, the European Pet Food Industry reported that cats are the most popular pets in Europe (about 110,000,000 in 2020), followed by dogs (about 90,000,000 in 2020) and ornamental birds (52,000,000 in 2020). Furthermore, in Europe, the pet-reptile population is approximately 9 000 000, with the UK ranked as the top country [4].
In spite of the beneficial effects of pets on human's health, these animals may be asymptomatic carriers of different bacterial (e.g., Pasteurella, Salmonella, Brucella, Yersinia, Campylobacter, Capnocytophaga, Coxiella, Leptospira, Chlamydia) [5,6], fungal (e.g., Candida sp. Aspergillus sp.) [7], parasitic (e.g., arthropods, helminths, protozoa) [8], and viral (e.g., rabies, norovirus, lymphocytic choriomeningitis virus) pathogens. Salmonellosis is one of the most serious zoonotic diseases in the world. Its etiologic agents -Salmonella spp. are in most cases pathogenic for people affecting primarily young children (<5 years), older adults (age >64 years), immunocompromised people, and pregnant women [9]. For some animals, Salmonella spp. is thought to be an opportunistic pathogen or even a part of natural gut microbiota. For instance, up to 80-90% of reptiles are asymptomatic Salmonella spp carriers [10][11][12][13]. It is well documented that reptiles may lead to cases of human salmonellosis, and these infections gained a separate disease entity and abbreviation -reptile associated salmonellosis (RAS) and reptile-exotic-pet associated salmonellosis (REPAS) to indicate the role of domestic reptiles living at households in spreading a Salmonella spp [14].

Methods
In order to address the globally increasing pet-tohuman transmission of Salmonella spp., it is crucial to establish the background of Salmonella-prevalence and their distribution in pets. This review aimed to describe the epidemiology of pet-associated salmonellosis and determine the retail sources of pets linked to human illness. We identified primary literature, reviews, and consensus guidelines through the National Library of Medicine's PubMed using the following search terms: 'Salmonella' or 'salmonellosis' AND 'pets' or 'companion animals' or 'zoonosis' or 'zoonotic infection' or 'amphibians' or 'reptiles' or 'dogs' or 'cats' or 'ornamental birds' or 'rodents' or 'guinea pigs' or 'international trade' or 'diet'. In order to provide pivotal and novel insights into the topic of zoonotic salmonellosis linked to the contact with companion animals and including the increasing in recent years tendency to keep exotic animals in households, we mainly focused on articles published in the second half of the last decade (2015-2020). However, the existing literature included in this review has been actualized by performing a literature search to add new relevant publications published in 2021. Original articles in English and different national languages (Polish, French, Spanish, if available) were included. Articles were screened by reading titles and abstracts and were initially excluded if they did not refer to zoonotic salmonellosis or were related to human salmonellosis caused by food or water contamination or due to human-tohuman transmission. Articles were then read in full, especially those aiming to detect Salmonella spp. samples in the feaces of different companion animals and clinical cases of Salmonella infection linked to contact with them (for instance, RAS salmonellosis). If a selected article was a review, we searched for relevant citations to find primary literature on the subject. Occasionally, reviews were directly used as sources, mainly to convey background information that is not in the core focus of this review or to additionally confirm the usage of data from a specific citation. We identified more than 500 articles of interest, of which we included 147 articles in this review based on their content. Furthermore, if the additional information was not available in scientific publication, we referred to the internet. We sorted out all the crucial information that we were willing to provide in this review in the following order, recapitulating the number of available literature in this scientific area: amphibians and reptiles, dogs, cats, pet birds and pet rodents (18 articles for 'Salmonella' and 'amphibians', 156 articles for 'Salmonella' and 'reptiles', 114 articles for 'Salmonella' and 'dogs', 46 articles for 'Salmonella' and 'cats', 15 articles for 'Salmonella' and 'pet birds', and 11 articles for 'Salmonella' and 'pet rodents' in the last 6 years searched in the PubMed database).

Classification of Salmonella spp
Salmonella spp. are a global problem of public health, as they cause almost 1.3 billion cases of illness each year, leading to more than 3 million deaths [15]. In the USA (US) alone, approximately 1.2 million human infections, 23,000 hospitalizations and 450 deaths occur each year. In contrast, in European Union (EU) countries, salmonellosis is the second most commonly reported gastrointestinal infection, followed by Campylobacter sp. In 2018, approximately 92,000 confirmed cases [16] of human salmonellosis were documented. Including total notification of human salmonellosis for the last 6 years, the stabilized tendency after a long period of a declining trend is observed [16].

Typhoidal salmonellosis
Based on the ability to develop specific pathologies in humans, all known Salmonella serovars are classified into typhoidal and non-typhoidal salmonellosis. Typhoidal Salmonella serovars including Typhi, Sendai and Paratyphi are highly adapted to humans; animals are not their carriers. These pathogens are the causative agents of enteric fever (also known as typhoid or paratyphoid fever if caused by serovar, Typhi or Paratyphi, respectively). This disease is characterized by low morbidity and high mortality displaying several symptoms, such as high fever, diarrhea, vomiting and headache [21]. Worldwide, enteric fever is the most prevalent in impoverished areas that are overcrowded with poor access to sanitation. To date, the highest incidences of typhoidal Salmonella infection in the world occurred in southcentral Asia, southeast Asia, and southern Africa [22].

Non-typhoidal salmonellosis
Non-typhoidal salmonellosis (NTS) is a zoonotic disease caused by multiple Salmonella serovars other than Typhi, Sendai, and Paratyphi. Due to differential disease symptoms, NTS can be divided into noninvasive and invasive (iNTS). The vast majority of the non-invasive NTS proceed as gastrointestinal selflimiting infections and do not require antibiotic treatment [23]. Invasive salmonellosis is a more severe disease with sepsis, septic aortic aneurysm, and septic arthritis, meningitidis, and are thought to result in the patient's death. Most of the iNTS are caused by the same serovars as non-invasive infections and affect people at higher risk group as children and elderly, people with health defects (AIDS or liver cirrhosis patients) and pregnant women; antimicrobial treatment is always needed. Contact with pets can result in both non-iNTS and iNTS [24]. In general, NTS salmonellosis is considered as a foodborne disease (about 80% of all cases were caused by food contamination, reaching 94% in the US in 2012 [25,26]). According to reports from EFSA (European Food Safety Authority) and ECDC (European Centre for Disease Prevention and Control), there are more than 90,000 of NTS cases in Europe annually, with the highest number in Germany, Czech Republic, UK, and Poland [16,27]. It is worth noting that non-typhoidal salmonellosis cases are still underestimated as some mild infections are unreported. Also, the epidemiological investigation is not always properly conducted or not conducted at all.

Routes of Salmonella spp. transmission
It was estimated that about 9% of human salmonellosis is caused by direct contact with animals. Considering only pets, these cases are significantly lower, accounting for approximately 1% of morbidity of human salmonellosis per year [28]. As Salmonella spp.are assummed to belong to the natural microbiota of animal's intestine or gallbladder [29], these animals may also potentially lead to indirect or direct transmission of the pathogen to human. Pets may contaminate the environment and other foodproducing animals by shedding the bacteria intermittently in their faeces [30]. Thus, Salmonella spp. is thought to spread by fecal-oral route during consumption of contaminated food or water. Stress factors such as transportation, mixing or crowding, food deprivation, parturition, exposure to cold, concurrent viral or parasitic disease, sudden change of feed or overfeeding, can lead to an increase in shedding load of Salmonella spp. to the environment [31]. For instance, De Lucia et al. [32] showed evidences of increased Salmonella spp. shedding by insufficient separation of wild birds from one outdoor pig farm. Salmonella spp. was isolated from pig faeces, environmental samples and wild bird droppings. The wild bird population increased considerably once the pigs had left the farm and the proportion of Salmonellapositive wild bird droppings increased over time with 7.4%, 15.8% and 44.3% at the first, second and third visit, respectively. The levels of Salmonella spp. identified in some of the wild bird droppings were high (10 5 -10 6 CFU (colony-forming unit)/gram (g)) indicating that this pathogen was actively replicating in the gastrointestinal tract of wild birds, leading to soil and outdoor pig farm contamination [32]. Furthermore, fomites such as houseflies (Musca domestica) can also spread Salmonella spp. For instance, in the US, Xu et al. [33] determined Salmonella-prevalence in flies captured from 33 cattle farms, including 5 beef and 28 dairy farms, and characterized antibiotic resistance profiles of the isolated Salmonella spp. 26 out of the 33 cattle farms (79%) and 185 out of the 1650 flies (11%) tested Salmonella-positive. These incidences varied from farm to farm, ranging from 0% to 78%, suggesting that flies are thought to be effective vehicles of transmitting antibiotic resistant Salmonella spp, posing risks to both human and animal health [33]. Another route, in which Salmonella spp. may spread is through vertical transmission, a phenomenon that occurs commonly in birds and reptiles. Avian eggs can be contaminated on the outer shell surface or internally. Internal contamination can be caused by the pathogen's penetration through the eggshell or by direct contamination of egg contents, before oviposition, originating from infection of the reproductive organs [34]. In contrast, reptilian eggs are more permeable than avian eggs, due to their low calcium and high fibre contents. Furthermore, water uptake by the egg after it has been laid is crucial in the development of the reptilian embryo. For this reason, reptilian eggs are usually laid in locations with high humidity. Thus, both the permeable eggshell and the high humidity are factors that increase the likelihood of Salmonella penetrating the shell [35].
Indirect route of Salmonella spp. transmission from animals to humans is possible due to the ability of Salmonella cells to survive in the environment. One of these abilities is the biofilm production. Salmonella spp. can attach to many different spaces; they may be found e.g on vegetables, chicken eggs, and stainless steel or plastic [36][37][38]. Biofilm structures with cellulose and curli fimbriae as main components of the extracellular matrix promote the vertical transmission of S. enterica ser. Enteritidis in chicken [37]. Salmonella spp. was also found as a contamination of many surfaces near the animal living area, e.g. in vacuum cleaner bag, sink drain or on the door knob in household in which bearded dragon was kept, and in the kitchen, service area and public space of Antwerp Zoo [39,40].
Regardless of the route of Salmonella transmission in the environment, the faecal-oral route remains the most common that leads to human infection. It seems that the ingestion dose required to induce the infection depends on the Salmonella serovar, type and the way the food is handled, and the susceptibility of the host. Hara-Kudo et al. [41] indicated results for five different Salmonella serovars from eleven outbreaks in Japan [41]. From 7 outbreaks caused by S. enterica ser. Enteritidis, in two of them the infection rate was 100%. Based on their observation, it was determined that the ingested dose of this serovar was at least 3.51 × 10 6 CFU (3.9×10 4 per 1 g) but only 4 and 6 people were exposed to infection. In other outbreak caused by Salmonella enterica ser. Cerro, the minimal dose of pathogen was 1.6 × 10 3 MPN (most probably number method) with 10% infection rate. The other analyzed Salmonella enterica serovars were: O4:H:eh, NT (ingested dose 2.6 × 10 6 CFU, 7 × 10 2 per 1 g), Montevideo (363 MPN, 66 and 960 MPN per 1 g) and Agona (<1500 CFU, <30 CFU per 1 g). Salmonella enterica ser. Montevideo and Salmonella enterica ser. Agona were the serovars with the lowest dose of pathogen needed to cause a disease after consuming salad with radish sprouts and fried soy pulp with egg, respectively [41].
During transmission to humans or animals, Salmonella spp. are faced with multiple stress factors such as a temperature, pH, salinity, metal ion stress, osmolarity, limiting nutrients and host immune defences. However, these pathogens are efficient enough to respond and adapt to these environments not only to survive but also to disseminate and retain its pathogenicity [42]. The ability of the bacteria to adapt to their host's environment is regulated by many microbial features, which are responsible for the expression of clinical manifestations in specific host species. Host adaptation by Salmonella serovars occurs through two mechanisms: the acquisition of novel genetic elements encoding specific virulence factors, and loss of genes. Kisiela et al. [43] suggested that activation or inactivation of mannose-specific type 1 fimbrial adhesin FimH in different Salmonella serovars reflects their dynamic ability and course of adaptation to their specific host's environment. The authors demonstrated that point mutations, horizontal gene transfer and genome degradation are responsible for differential pathoadaptive evolution of some Salmonella serovars [43]. Furthermore, Salmonella spp. can adapt to human hosts by changing their outer structures, such as lipopolysaccharide (LPS) and specific outer membrane proteins (OMPs). Those changes can lead to resistance of Salmonella spp. to human serum, especially the complement system which is part of the innate immune response [43][44][45][46]. Salmonella serovars isolated from reptiles are often resistant to human serum, which enables them to cause disease in humans. Strains isolated from the cloaca of the grass snake (Natrix natrix) from urban and touristic areas in Poland [30] have shown to be highly resistant to 50% human serum [data not published]. The possibility of carrying more than one Salmonella serovar by the same animal significantly increases the risk of genetic material exchange, which could lead to the acquisition of new virulence genes or other genetic factors.

The role of diet in Salmonella spp. prevalence in pets -amphibians, reptiles, birds, rodents, cats and dogs
In spite of the widely discussed Salmonellaprevalence in animals, relatively little is known about the impact of diet that drives the possibility of the animal to become a carrier of this pathogen [47]. In this section, we reviewed the literature concerning the role of diet in Salmonella-prevalence in pets and showed risk factors associated with the increase of Salmonella spp. transmission from pets to humans.
Amphibians are carnivores and eat mostly earthworms, crickets, flies, moths, honeycomb moths, and small cockroaches, albeit bigger species of amphibians could be fed with small fish or mice [29]. Salmonella carrying in amphibians is generally asymptomatic. These pathogens are often isolated from frogs and toads than newts and salamanders, which can be the result of wider human contact with them in the environment (wild animals) or frequent breeding (pets). As amphibians obligatorily require access to water to live, contaminated water may become an indirect threat for humans. Additionally, amphibians are assumed to become carriers or suffer from a Salmonella spp. infection by consumption of contaminated feed or insects, which could be vectors of these bacteria, as described previously [48].
The diet of reptiles may be subjected to a variation depending on genera and species. Salmonellaprevalence in reptiles is reported to be higher in turtles than in lizards and snakes. In general, diets of omnivorous reptiles are usually balanced, containing both plant (herbs, crushed fruits, and vegetables) and animal components (mostly insects, young mice, snails). In contrast, herbivorous tortoises are characterized by a plant diet (herbs, fruits, and vegetables). Snakes and crocodiles are undoubtedly carnivores; in their diet animal feed is almost exclusive, consisting mainly of fish, chicks, rodents and other small mammals [33,49,50]. It is also important to note, that microbiota in the digestive tract of some reptiles may change periodically after long periods of fasting. Investigation of intestinal microbiota among herbivorous, omnivorous and carnivorous reptiles has shown that certain bacteria may become dominant, depending on diet, especially in animals kept in captivity. For instance, in herbivorous reptiles, highly varied gram-negative bacteria showed the highest prevalence, including Salmonella spp. Furthermore, Jiang et al. [51] observed the significant difference between the gut microbial community in loach-fed crocodile lizards than in the earthworm-fed and wild lizards. In addition, they found that the captive lizards fed loaches resulted in the enrichment of Elizabethkingia, Halomonas, Morganella, and Salmonella spp. Thus, this study proved that a diet promoting colonization of Salmonella spp. in the intestine of captive lizards may lead to the increased likelihood to transmit the pathogen from reptiles to humans [51]. The impact of diet in Salmonellaprevalence in captive reptiles were also reported in the US by Clancy et al. [52]. From a total of 175 samples isolated from 182 reptiles housed in Bronx Zoo, Salmonella enterica subsp. enterica was the most predominant (78/175; 45%). However, other nonenterica serovars were also identified, including Salmonella enterica subsp. diarizonae (42/175; 24%), many of which were clinically ill showing bony changes, dermatitis and anorexia. Authors determined that the strongest factors associated with an increased risk of illness in reptiles were carnivorous diet and prior confiscation [52].
Depending on the species, the diets of birds consist of animal and/or plant elements. Herbivorous birds forge on seeds, herbs, fruits, vegetables and special factory-made feed, while predatory birds such as eagles, hawks, falcons and owls are carnivores killing their prey by talons. Diet-dependent spreading of Salmonella spp. is associated with the contamination of feed by faeces, or in the case of omnivorous and carnivorous birds, with consumption of contaminated carrion, as well as colonized or ill prey [53]. The risk derives especially from the practice of releasing birds of prey during hunting. A similar risk occurs in outdoor pens of parrots or pigeons (i.e., kept on the balcony), these pens often have contact with wild, free-living birds, which can easily lead to contaminated feed ( Figure 1). Much in this regard depends on the decisions of individual pet bird owners [51,53]. It is also worth bearing in mind that commercially available feeds used by pet owners may not provide sufficient nutrients, or consist of illbalanced nutrients for a given species, leading to poor health and a higher susceptibility to infection or asymptotic carrying [53].
Young rodents, as all mammals, first thrive on their mother's milk and upon reaching appropriate age become omnivorous. Their diets become very varied and contain many plant materials, as seeds, herbs, fruits, vegetables and animal components, as invertebrates, eggs or carrion [54]. Infected by Salmonella spp. rodents are involved in spreading the pathogen in their environment through faeces, which stay contagious up to three months. This may contaminate feed of other rodents, like fruits, vegetables, hay, fodder or water, and consequently lead to increased Salmonella-prevalence among other pet rodents [54,55].
Diet of cats and dogs is as varied as any omnivores; however, animal feed predominates, including raw meat [56]. Newborn cats and dogs consume their mother's milk; however, dogs may also consume placenta and colostrum, which is beneficial for the formation of a healthy microbiota. Owners of older animals often introduce a diet of raw meat, including poultry [57]. According to a large, structured, 2016 survey in the US, 3% of dog and 4% of cat owners feed their pets in raw products, and raw or cooked human food was purchased for pets by 17% of dog owners [3]. Such a diet, despite its many benefits, carries a high risk of Salmonella spp. infection. For instance, Finley et al. [58] observed that when dogs are fed with Salmonella-contaminated food, they can become infected and consequently shed the bacteria in their faeces to contaminate the environment, other domestic animals, and even pet owners [58]. Experimental addition to dog's diet probiotics containing Lactobacillus led to inhibition of Salmonella spp. growth. Probiotic lessened the gastrointestinal symptoms in ill dogs, albeit it also induced increased release of Salmonella to the environment, potentially leading to increased risk of infections in other animals. As Lowden et al. [9] have shown, commercial feed including dry food lowered the risk of asymptomatic carrying of Salmonella spp., but did not exclude it [9].
Other than diet, additional factors can influence Salmonella-prevalence among pets, including coexistence in limited space, environmental conditions, polygamy, presence of arthropods, contamination of paraphernalia, contact with wild animals, and others ( Figure 2).

Salmonella spp. prevalence in amphibians and reptiles
Amphibians and reptiles have become increasingly popular as pets worldwide. In the US alone, 4.5 million households own at least one reptile [3]. The most predominant are turtles, lizards and snakes. Nevertheless, up to 90% of reptiles are carriers of one or more Salmonella serovars [3,47]. In contrast, within the EU countries, less than 1% of human cases of salmonellosis are associated with exposure to reptiles [59]. Including amphibians kept at households, the most popular are frogs, salamanders and caecilians. In these animals, Salmonella-prevalence and associated cases of transmission to human are very limited compared to that of reptiles. However, their role is significant.
For instance, Ribas et al. [61] isolated 67 Salmonella strains from 97 frogs and toads (67/97, 69%) breed on Thailand farms and urban and protected areas; Salmonella-prevalence was 90%, 0% and 44.8%, respectively. The high Salmonella-prevalence in amphibians kept in farms (90%) confirms their significant role as vectors for the spread of salmonellosis to livestock. In this case, transmission to humans was considered as a result of indirect contact with amphibians. Of the eight identified in amphibians serovars, six of them (S. enterica ser. Hvittingfoss, S. enterica ser. Newport, S. enterica ser. Panama, S. enterica ser. Stanley, S. enterica ser. Thompson, and S. enterica ser. Wandsworth) led to human salmonellosis in Thailand. Farm-reared Chinese edible frogs (H. rugulosus) showed the highest Salmonellaprevalence (62%) [63]. In another study, Williams et al. [62] isolated 21 Salmonella serovars from 47 frogs (21/47, 45%). In this case, amphibian-associated salmonellosis appeared in 3 children keeping amphibians at households (3/15, 20%) [62]. These reports lead to the conclusion that awareness among amphibians'owners about potential risks of amphibianassociated salmonellosis is still required.

Cases of reptile-associated salmonellosis in humans
The determination of the zoonotic potential of Salmonella spp. is important to highlight the problem of public health, particularly due to the increasing tendency of keeping such exotic animals as reptiles at households. For instance, in the US, of the 8389 non-typhoid salmonellosis case-patients, 290 (3.5%) reported reptile exposure. Including faecal samples of 60 reptiles, 36 (60%) yielded the same Salmonella serovar as the human isolate [78]. Krishnasamy et al. [85] described five Salmonella Paratyphi B variant L(+) tartrate + (Java) isolates in four American inhabitants keeping ball pythons (Python regius) as pets. The median patient age was 10 years (range 1-40 years). No patient was hospitalized, and no deaths occurred [85]. In the US, the main outbreaks of human salmonellosis caused by turtle exposure occurred in 2015 and 2016. In 2015, based on the interview of 104 patients, 50 (48%) had contact with turtles. 18 (40%) of them were hospitalized, but no death occurred. The median age was 3 years (range <1-77 years). 21 positive Salmonella isolates were detected in turtles and 17 isolates matched the outbreak strains [85]. In 2016, a total of 133 patients with human salmonellosis were reported; 41% of them were children 5 years of age or younger. 55 (50%) of the 110 interviewed people reported contact with turtles or their environments; 38 patients were hospitalized, and no death was reported [86,87]. In Spain, Ricard et al. [88] reported a case of meningitis caused by S. enterica ser. Vitkin in a 1-month-old child after exposure to an aquatic pet turtle [88]. A very similar case was also reported in Spain. The same Salmonella serovar was isolated from the turtle faecal sample and blood of a two-year-old girl who had severe complications including high fever, sunken eyes and, pasty mucosa [89]. Furthermore, in Italy, Corrente et al. [13] conducted a crosssectional study among reptile owners in order to assess a potential link between the presence of Salmonella spp. in their pets and the hygiene practices. From a total of 100 families, in 26 of them the potential risk of RAS occurred. Including 100 pet animals tested, Salmonella-prevalence was 57%. It was determined that co-habitation of the animals with other reptiles in the same terrarium was associated with a 2-fold increase in the risk of Salmonella spp infection. Animals handled by owners that did not report washing their hands after the cleaning procedures or the handling were exposed to a 3-fold increase in the risk of infection [13].
Considering reports from last 6 years and published by European scientists, reptile-associated salmonellosis with detection of the same Salmonella serovar in both patient's blood and reptile faeces was observed in Switzerland -2016 (years of publication) [90], UK (UK) -2015 [91], Romania -2017 [92], France -2015 [88,93] and Spain -2015 [88,89]. In Switzerland, the first case of reptile-associated sinusitis due to S. enterica subsp. diarizonae was reported in a 29-year-old snake handler who owned five pet snakes. In three snakes, the same Salmonella serovar was detected as in the blood of the owner. It was suggested that Salmonella spp. reached the upper respiratory tract hematogenously after oral inoculation or perhaps via inhalation [90]. In the UK, from 175 cases of human salmonellosis reported in the period 2010-2013, 48 patients had exposure to reptiles (48/175, 27.4%); 8 patients reported RAS salmonellosis with severe symptoms such as bacteraemia, meningitis and colitis requiring surgery. Almost half of RAS patients were hospitalized (23/48), but no deaths occurred [91]. Furthermore, in Romania, Gavrilovici et al. [92] reported a rare case of otitis with Salmonella spp. in a healthy 16-year-old adolescent, who was bathing in a village lake, where turtles were common. After taxonomic speciation, it turned out that the etiologic agent of this ear infection was S. enterica subsp diarizonae. Otitis was also associated with mastoiditis. Audiometric testing showed a moderately conductive hearing loss [92]. In France, the first isolation of S. enterica subsp. arizonae was reported in the bronchial aspirate from a patient suffering from pneumonia. The patient, a 73-year-old man kept snakes as pets [93].
Infants and children <5 years old are the most frequently exposed to RAS infections [11,14,78,89,94,95]. One study performed in Taiwan revealed that 31% of RAS cases occurred in children less than 5 years of age and 17% occurred in children aged 1 year or younger [94]. In other study, Kiebler et al. [96] investigated an outbreak of human salmonellosis in 133 people with exposure to pet bearded dragon lizards. The median patient age was 3 years (range, <1-79 years), 57% were aged ≤5 years, and 37% were aged ≤1 year [96]. Nevertheless, cases of RAS infection in adults and elderly people are also occurring, but in a lower frequency; mostly these infections are escorted with other, secondary infections. For instance, a 42-year-old patient from Equatorial Guinea experienced symptoms such as malaise, weakness, fever, and mild diarrhea. Based on the faecal sample analysis, S. enterica subsp. salamae was identified. During medical consultation, the patient reported regular consumption of sea turtle meat [97]. Furthermore, in Japan, Suzuki et al. [98] reported a case of pericarditis caused by S. enterica subsp. arizonae in a 36-year-old man with a history of type 2 diabetes mellitus. The patient was infected by pathogen transmission from pet snakes: a ball python (Python regius) and a Mexican black kingsnake (Lampropeltis getula nigrita) [98]. Nevertheless, these findings highlight the heightened risk in children and the potential for RAS to be transmitted without direct contact with the animal or its enclosure [99][100][101]. Furthermore, more hospitalizations occurred in RAS patients than non-RAS cases, suggesting that reptile-associated infection carries a higher likelihood of more severe symptoms with bloodstream infection [59].

Salmonella spp. prevalence in dogs
Dogs usually act as asymptomatic carriers of Salmonella spp; they are thought to shed one or more serovars intermittently for more than 6 weeks [102]. Rarely occurring clinical signs of salmonellosis in adult dogs and puppies include fever, loss of appetite, diarrhea, bloody diarrhea, abdominal pain, and abortion [103]. Other factors that may increase Salmonella-prevalence in dogs are the environment where animals live, contact with wild animals or other infected animals, differences in pet sanitary practices, feeding habits, public awareness about dog zoonosis, socioeconomic status of the owners, sample size, sampling strategies, and isolation methods performed [102,103].
Including 2015-2021, Salmonella-prevalence in household dogs was reported in different continents, indicating significant geographic variation in global perspective ( Table 2). A study of 436 faecal samples from healthy dogs, including 126 samples from dogs kept in UK homes, reported Salmonella spp. only in one female terrier breed (1/4366, 0.23%) [9]. In another study, from a total of 325 healthy dogs across Spain, Salmonella-prevalence was 1.85% (6/325, 1.85%) [103]. Furthermore, Reimschuessel et al. described that 60 diarrheic and non-diarrheic dogs from a total of 2422 dog population were Salmonellapositive (60/2422, 2.5%). Faecal samples were solicited from different geographically dispersed veterinary laboratories in the US. This study confirmed statistically higher prevalence in diarrheic dogs (3.8%) than in non-diarrheic dogs (1.8%) [104], which is in concordance with other reports [8,[105][106][107]. Faecal samples collected from 144 nondiarrhoeic dogs in Grenada revealed that 5.6% (8/ 144) of them were Salmonella positive [105]. A similar percentage was also observed in Western Australia. Of the 405 faecal samples obtained from dogs placed from the different environment: animal shelters, racing greyhounds or households, 5.4% were Salmonella-positive (22/405, 5.4%) [108]. A slightly higher percentage of Salmonella-prevalence in 243 dogs was observed in China (23/243, 9.47%) [28]. Furthermore, investigations undertaken in Ethiopia and Equador represent even higher prevalences, with the percentage of 11.7% [109] and 12.5% [110], respectively. The goal of Wu et al. [106] study was to investigate the association between Salmonella spp. infection, pet dogs and their caregivers in Thailand. Salmonella-prevalence was observed in 18 companion dogs from a total of 140 analyzed (18/140, 12.86%) [106]. As conclusion, dogs may be potential agents of salmonellosis, especially when multiple different factors (e.g. weakened immune system, improper diet, rich in raw food, and indecent environmental and animal welfare commitments) contribute to the increased risk of pathogen transmission to dog owners.
Salmonella prevalence in dogs is also highly variable depending on the environment in which the animals live. For instance, Salmonella isolation rates from stray dogs and shelter dogs are higher than those from household dogs. This phenomenon may be due to the increased freedom to roam and scavenge, possible close contact with carcasses or offals of wildlife and raw and undercooked food [103]. In Spain, Bataller et al. [103] obtained 1 Salmonellapositive rectal swab from 85 dogs kept in households (1.17%) and 3 Salmonella-positive samples from 84 dogs kept in animal shelters (3.57%) [103]. Furthermore, in Texas, US, Salmonella prevalence  [111]. Moreover, in Iran, a total of 100 faecal swabs and blood samples were obtained from symptomatic and apparently healthy shelter dogs; 11 samples (11%) of them were Salmonellapositive [112]. These observations indicate the serious problem of public health especially in urban communities, where a massive population of stray dogs in cities exists with no certain monitoring and control system over their nutritional habits, potentially leading, in consequence, to transmission of Salmonella spp. to humans [112,113].

Salmonella spp. prevalence in cats
Several reports published in 2015-2021 have concluded that contact with healthy cats kept in homes does not constitute a major zoonotic risk of salmonellosis. Only a few cases were reported, in which salmonellosis was passed on from cats to humans [27,28,104,114,115]. For instance, in China, Wei et al. [28] collected faecal samples from 113 cats and only two cats (with and without diarrhoea) were Salmonella-positive (2/113, 1.77%) [28]. In Western Australia, Aeh and Stayt [108]. reported the prevalence of faecal pathogens in the microbiome of cats with diarrhoea. Of 289 feline faecal samples reviewed, Salmonella spp. (1.7%) were detected, mostly in young cats (range 14 weeks to 2 years and 10 months) [108]. Interestingly, Vercelli et al. [116] detected S. enterica ser. Typhimurium in the urine culture of a cat suffering from endocarditis and myocarditis [116]. Introduced above reports show a relatively low Salmonella-prevalence in cats.

Salmonella spp. prevalence in ornamental birds
Among birds most often kept by humans are parrots, canaries, European goldfinches, pigeons, and increasingly popular birds of prey like owls and falcons. However, including 2015-2021, a low frequency of Salmonellaprevalence was reported in ornamental birds kept in households. Most of the studies relate to Salmonella spp. transmission to humans by indirect contact of pet birds with other companion or wild animals. For instance, when pet birds are gathered in an exhibition in open-air aviaries, other animals having access to these places come in direct contact with them (like in the example shown in Figure 1). This contact may be a source of indirect Salmonella spp. carriage to humans, especially including the fact that Salmonella can survive for extended periods on wood and dust and can live for 28 months in avian faeces [117]. In one study, it was determined that domestic cats and dogs were linked to Salmonella spp. transmission from wild birds (81% and 52% of cat and dog cat isolates, respectively, shared a common Salmonella serovar with birds) [118]. Furthermore, Mather et al. [119] determined that some subtypes of S. enterica ser. Typhimurium -definitive phage types (DTs) 40, 56 variant and 160 -are hostadapted to wild passerine birds (e.g. finches, sparrows), and these birds may represent a reservoir of infection for humans and other companion animals, especially those kept outdoor (for example as shown in Figure 1 or dogs/ cats partly allowed to roam outdoor) [120,119]. Moreover, de Oliviera et al. [119,120] obtained cloacal swabs from 156 free-ranging urban birds including synanthropic great egrets (Ardea alba) and feral pigeons (Columba domestica) that inhabited the surroundings of an urban zoo in Brazil to assess shelter and food. By defecating in these areas, they potentially contribute to the Salmonella-transmission to the captive in zoo animals. A total of 11 birds were positive for S. enterica ser. Typhimurium (11/156; 7%) showing that these freeranging birds are possible sources of infection to other animals [119,120]. In urban infrastructure, synanthropic birds such as domestic pigeons, house sparrows or common chaffinches find abundant food and places for roosting and nesting. This phenomenon may create opportunities for frequent contact with humans and other animals. Pigeon dropping may be a potential risk of Salmonella spp. transmission through contamination of drinking water sources or agricultural crops [121]. Sharing the same environmental condition where outdoor pets have contact with pigeon droppings may lead to Salmonella spp. passage and, in that way, pets become asymptomatic carriers of this pathogen. In conclusion, although cases of Salmonella-transmission from pet birds to humans are rare, caution should still be exercised when engaging in contact with these companion animals. Furthermore, limiting the contact between wild birds and pet birds and their foods is another valid measure to prevent unnecessary transmission.

Salmonella spp. prevalence in rodents
Due to their small sizes and relatively low purchase and maintenance costs, rodents (e.g. hamsters, rats, mice, gerbils and guinea pigs) did not lose their popularity as pets in recent years. However, including 2015-2021, more cases of Salmonella spp. transmission to humans were associated with wild rodents rather than their captive counterparts [122]. In one study, Himsworth et al. [123] [124]. Altogether, including the fact that since 2015 we did not find the literature detecting cases of Salmonella spp. in pet rodents and a low number of articles determined Salmonella spp. exposure in wild rodents, we are inclined to ascertain that there is a significantly low possibility to be Salmonella-infected by contact with these animals. However, these cases may occur and should not be omitted. Nevertheless, considering rodents individually, guinea pigs are highly susceptible to Salmonella spp. and hence, they need more attention. These animals are the most frequently kept as pet rodents, with 0.8 million in the UK and 1.36 million in the US in 2019 [125]. They are often selected as pets due to their placid, docile temperament and ease of handling [54]. However, Salmonella-infected guinea pigs exhibit reduced physical activity, social interaction progressing, lethargy, and anorexia. Reduced physical activity can induce gut stasis which can cause rapid deterioration resulting in sudden death [57]. The incubation period is 5-7 days [58]. Aging, other diseases, malnutrition, and environmental stress are predisposing factors to develop severe clinical symptoms of salmonellosis in guinea pigs [57].
Due to the high susceptibility to Salmonella spp., guinea pigs are thought to become carriers, which in turn make them a potential source of Salmonella spp. transmission to humans. For instance, in 2017, two S. enterica ser. Enteritidis isolates were detected in 9 American inhabitants who reported exposure to pet guinea pigs, which were purchased from two pet stores. Five Salmonella isolates from guinea pigs matched the outbreak strain. The median patient age was 12 years (range = 1-70 years). One patient was hospitalized, and no deaths were reported [126][127][128][129][130][131]. In conclusion, guinea pigs may act as potential sources of human salmonellosis caused by direct or indirect contact with humans. However, it is worth bearing in mind that household guinea pigs as rodents are not likely to be a source of human salmonellosis, even if they are highly susceptible to be Salmonella spp.

The importance of wildlife trade
International importations of free-living animals are one of the major drivers of salmonellosis emergence and results in its globalization. Illegal wildlife trade (for example, for companion or ornamental pets), is the world's fourth largest illegal business after narcotics, counterfeiting and human trafficking [132]. Although the scale of the illegal market is unknown, it was calculated that approximately 5.9-9.8 million reptiles were (legally) imported to the EU in 2009 alone, a substantial rise from the 1.6 million imported in 2005 [133]. Including European countries, Germany is by far the largest importer of live reptiles within the EU. In this country, 1532 valid reptile species and 352 valid amphibian species had been recorded in the German pet trade in 2017-2018 [134]. Another report showed that, from 2013 to 2014, about 490,750 exotic individual animals were legally imported to the Netherlands. 43% of them were destined for the Netherlands, a small number (4%) was destined for other EU countries and the rest (53%) were in transit to other non-European countries. The majority of the animals imported in the Netherlands were reptiles (93.8%), followed by amphibians (5.8%), birds (0.06%) and mammals (0.4%). The animals originated predominantly from the US (78.8%), Vietnam (5.1%), Indonesia (3.5%) and Tanzania (3.1%) [134]. Furthermore, Green et al. [135] evaluated the trade in live wild animals entering the UK in 2014-2018 using data reported by the Animal and Plant Health Agency (APHA). Over 8 million individual animals were imported into the UK from 90 countries across nine global regions. Amphibians were the most commonly imported group (73%), followed by reptiles (17%), mammals (4%), and birds (3%). The highest number of import records came from Europe and Africa, but the largest volume of animals came from North America and Asia [135]. Since exotic amphibians and reptiles are not tested for Salmonella spp. and a large number of them are imported by trade companies (99.8%) and mostly destined for the pet industry, the probability of exposure of humans to Salmonella spp. is high [134]. The scale of international trade is likely to be even greater than current estimates due to incomplete record-keeping and widespread illegal activity throughout the industry [135]. Thus, due to observed more interactions with humans by international trade of free-living amphibians and reptiles and human urbanization resulting in increasing human encroachment into natural ecosystems, the role of these animals in Salmonella distribution is incontrovertible [135]. A special field where a wildlife trade takes place are wet markets. These types of markets are especially popular among low-income communities of Asia, Africa and Latin America. While countries have drawn the attention to wet markets due to COVID-19 pandemic, these areas can be also an important sources of other zoonoses such as salmonellosis [136]. Factors predisposing Salmonella spp. occurrence in wet markets are poor animal keeping conditions (overcrowding, cramped cages, transport mortality, wrong or insufficient food, proximity to other animal and species, stress, injuries and diseases), poor sanitation (lack of toilets and hand washing stations) and the possibility to contaminate fresh food and meat by shedding bacteria from wildlife animals [137,138]. To date, a lot of studies relied on the contamination of different types of meat including chicken, beef and pork by Salmonella enterica in China, Philippines, Malaysia and Vietnam [139][140][141]. The studies from Asia also confirmed presence of virulence genes and multi drug resistant and ESBL producing (extended-spectrum beta-lactamases) fenotypes of S. enterica isolated from meat sampled in wet markets [142,143]

Pet regulations and guidelines for pet owners
Animal-human relationships may reduce human stress and ailments. However, these interactions may also have harmful effects, including the spread of salmonellosis. A study conducted among 401 Canadian pet owners revealed a range of practices that increase Salmonella-disease risk, for instance: allowing dogs and cats to sleep in a child's bed, allowing dogs to lick a child's face, and allowing a reptile to roam through the kitchen. Although the hand washing by children was high (76% washed hands after touching the pet, its feces or housing), the authors concluded there is still a high need to educate people on Salmonella-diseaseprevention practices [144]. Different national and international organizations, including the World Health Organization (WHO), CDC, the Association of Reptilian and Amphibian Veterinarians (ARAV) and the American Pet Products Association (APPA) are providing pet owners in the recommendation on how to prevent or at least minimize salmonellosis well as to promote and develop responsible pet ownership and the pet products industry [145]. These organizations support and monitor the industry legislations and regulations. Although Salmonella occurs globally, these pathogens are most commonly detected in areas, where intensive animal husbandry is practiced. In some countries, Salmonella infections were eliminated in domestic animals due to Salmonella eradication programs. In Sweden, according to the Swedish law on zoonoses (Zoonoslagen, SFS 2006 1039), every case of Salmonella spp. isolation from domestic animal, animal product or feed should be reported and measures to eradicate Salmonella should be taken at any positive finding [145].

Conclusion
Bacteria Salmonella spp. are still one of the most serious global problems of public health affecting approximately 1.3 billion cases of illness every year. To date, several different routes of Salmonella spp. transmission to human were reported, both indirect (for example by environment) and direct (by consumption of contaminated food or close contact with infected animals). Due to increasing frequency of keeping exotic animals like amphibians, reptiles and ornamental birds at households, their role in the transmission of Salmonella spp is growing. Based on the current literature regarding Salmonella spp. isolation and characterization in pets, we indicated bacterial zoonotic potential of pet-to-human transmission. It is worth noting that Salmonella-prevalence in pets depends on many aspects including diet, co-existence with other animals in limited space, environmental conditions, potential contact with wild animals and others. Based on collected data of Salmonella-prevalence in pets, we emphasize that when considering adopting and keeping companion animals, it is important to be aware of potential routes of Salmonella spp. transmission and their consequences of human health.

Funding
This manuscript is supported by the Freie Universität Berlin Open Access Publication Fund.

Notes on contributor
Mateusz Dróżdż holds an M.Sc. in Microbiology with a major focus on Salmonella-prevalence in the environment and is currently PhD student at Freie Universität Berlin, Germany. His main focus is temporal control of spliceosome activity to modulate splicing switches. Michał Małaszczuk is the PhD student at Department of Microbiology, University of Wroclaw, Poland. His project is related to the characterization of Salmonella sp. pathogenicity and the determination of antibiotic resistance of bacteria. Emil Paluch is the assistant professor at Department of Microbiology, Wrocław Medical University in Poland. His main focus is to understand the molecular mechanisms of different compounds from plants and fungi that may be used in medicine and industry, with special regard to nano and microcomposites. Aleksandra Pawlak is currently working as a researcher and teacher at Department of Microbiology, University of Wrocław, Poland. Her scientific interests focus on Salmonella spp. virulence factors, biofilm formation and microbiota of free-living reptiles.

Data availability statement
The authors confirm that the data supporting the finding of this study are available within the article (http://dx.doi. org/10.1080/20008686.2021.1975530).