Staphylococcus aureus Epidemiology in Wildlife: A Systematic Review

Staphylococcus aureus is a common bacterial colonizer of humans and a variety of animal species. Many strains have zoonotic potential, moving between humans and animals, including livestock, pets, and wildlife. We examined publications reporting on S. aureus presence in a variety of wildlife species in order to more cohesively review distribution of strains and antibiotic resistance in wildlife. Fifty-one studies were included in the final qualitative synthesis. The most common types documented included ST398, ST425, ST1, ST133, ST130, and ST15. A mix of methicillin-resistant and methicillin-susceptible strains were noted. A number of molecular types were identified that were likely to be found in wildlife species, including those that are commonly found in humans or other animal species (including livestock). Additional research should include follow-up in geographic areas that are under-sampled in this study, which is dominated by European studies.


Introduction
Staphylococcus aureus is a common commensal bacterium that lives within the nares, throat, and on the skin of humans and a wide variety of animal species. S. aureus can spread via person-to-person contact (directly or mediated by fomites) and can be transmitted zoonotically via direct contact with animals or animal products, including raw meats [1]. It can also be maintained in the environment in manure, water, or the air [2].
Because of its frequency in various environments and species, it is critical to understand movement within and between communities. S. aureus is frequently resistant to one or more classes of antibiotics, and the continued spread of methicillin-resistant S. aureus (MRSA) over the past several decades in both human and animal species has increased the risk of acquiring a resistant infection that makes treatment more difficult and costly [3].
The epidemiology of MRSA in particular has changed over the past 30 years [4]. Once primarily a hospital-associated pathogen, the rise of novel strains of MRSA in the 1990s outside of the nosocomial environment led to the recognition of "community-associated MRSA" (CA-MRSA), in contrast with the historic hospital-associated (HA-MRSA) strains [5]. In the mid-2000s, a third genre of MRSA was recognized, as colonization and infection of livestock and livestock workers led to the designation of livestock-associated MRSA (LA-MRSA) [6]. It should be noted that we know less about the changes in methicillin-susceptible S. aureus during this period (MSSA), as the bulk of surveillance is dedicated to MRSA and does not always capture MSSA epidemiology.
Wildlife are a special case and often under-studied in the epidemiology of antibiotic resistance in the community and environment. Wildlife can act as reservoirs for intrinsic resistance elements or organisms (those that are naturally occurring in the environment, including environmental bacteria and fungi living in soil and water) and may also be exposed to resistant organisms or resistance genes in the environment amplified via human activity. This may be via treated humans who excrete resistant bacteria, antimicrobials that eventually end up in sewage effluent dispersed into the environment, or from sludge or waste from humans or livestock dispersed on fields as fertilizer. This may lead to further dissemination into streams, rivers, and larger waterways and also allow for airborne transmission of dried materials. Resistance may also be generated in the environment due to spraying of antibiotics on citrus groves and other plants [7] as well as by use in aquaculture [8].
S. aureus is a commensal organism that is able to effectively colonize a wide variety of host species, including many mammalian species but also birds and fish. As such, animals besides human have the potential to harbor novel strains of S. aureus, which could enter the human population, or conversely, humans may also transmit strains of S. aureus to other animal species [9], which can then acquire additional resistance genes.
The clearest evidence of zoonotic transmission of S. aureus has been in livestock. Isolates of clonal complex 398 appear to have originated as a human-adapted lineage but were transmitted to livestock including pigs and cattle and have become both more antibiotic-resistant (including MRSA strains) and have also typically lost some human virulence factors [9]. A similar situation appears to have happened with CC5 in poultry [10]. Recent research also suggests an emerging lineage in humans, strains of CC130, originated in cattle, and typically carry a novel methicillin resistance gene (mecC, originally called mecA LGA251 ) [11]. The role other animals may play in such cross-species transmission is less defined. Systematic collection and molecular typing of S. aureus from animal species has not been a priority; as such, cross-species transmission events from such animals to humans or wildlife to better-studied animal populations (including livestock and poultry) have likely been missed. This review examines the epidemiology of S. aureus in wildlife, including molecular typing and antibiotic resistance data where available.

Search Results
Searching within Pubmed resulted in 856 hits, Web of Science in 58, and peer-reviewed materials within ProQuest in 918, for a total of 1832 publications. Upon searching references for additional studies that had been missed by our search terms, another nine were added. Titles and abstracts were examined to exclude duplicates (956); this left 885 remaining. Additional publications were excluded if they used animals only as an experimental model rather than examined epidemiology in wild species (such as rats, guinea pigs, and rabbits) and those that only mentioned S. aureus within the discussion or otherwise rather than consisted of a study focused on S. aureus epidemiology in wildlife. This left 69 for analysis ( Figure 1).

Full-Text Articles Excluded
Eighteen studies were included within the initial analysis but excluded from further analysis Table 1 due to lack of detail reported regarding the S. aureus detected. These studies included the identification of S. aureus in a white ibis in Egypt [12], captive bustards in the United Arab Emirates [13], a peregrine falcon (Falco peregrinus) in Spain [14], from "free living insectivores" including the common shrew, lesser shrew, bank vole, root vole, and field mouse [15]; S. aureus was reported in this study but were not typed nor reported which species were positive. In Brazil, an opossum with mastitis was described but neither molecular typing nor antibiotic resistance phenotype was provided [16]. Similarly, a systemic S. aureus infection in a raccoon was also reported but not further characterized [17]. A 2013 study suggests that S. aureus infection is an important skin disease of red squirrels (Sciurus vulgaris) in Great Britain [18], and a Canadian publication determined that S. aureus was a common organism found in bite wounds from Norway and Black rats (Rattus norvegicus and Rattus rattus, respectively) [19], but no details were provided in either paper. S. aureus was also found along with other mecA-positive staphylococci in foxes in the United Kingdom, but samples were not typed [20].
S. aureus-positive Spanish Ibex (Capra pyrenaica hispanica) were identified in Spain but not typed or examined phenotypically [21]. A black rhinocerous (Diceros bicornis) in Kenya was reported to have an S. aureus infection (a possible cause of mortality) but also lacks details [22]. Bighorn sheep (Ovis canadensis nelson and Ovis canadensis mexicana) in Arizona were found to be colonized with S. aureus [23], but it was not characterized. S. aureus was identified in fecal samples taken from red deer in Poland [24] and from fecal samples from slaughtered reindeer in Finland and Norway [25] but was not further characterized. Finally, S. aureus was isolated from bottlenose dolphins (Tursiops truncatus) in the southeastern United States, but it was not further characterized [26,27]. Multiple zoo animals in Belgium were tested for MRSA, but no positive samples were reported [28]. Schaumburg [29] was not included in Table 1 because species are not specific (monkey, goat, etc. rather than exact species types) but demonstrates some sharing of spa types between humans, domestic animals, and wildlife (more for the former than the latter). to have an S. aureus infection (a possible cause of mortality) but also lacks details [22]. Bighorn sheep (Ovis canadensis nelson and Ovis canadensis mexicana) in Arizona were found to be colonized with S. aureus [23], but it was not characterized. S. aureus was identified in fecal samples taken from red deer in Poland [24] and from fecal samples from slaughtered reindeer in Finland and Norway [25] but was not further characterized. Finally, S. aureus was isolated from bottlenose dolphins (Tursiops truncatus) in the southeastern United States, but it was not further characterized [26,27]. Multiple zoo animals in Belgium were tested for MRSA, but no positive samples were reported [28]. Schaumburg [29] was not included in Table 1 because species are not specific (monkey, goat, etc. rather than exact species types) but demonstrates some sharing of spa types between humans, domestic animals, and wildlife (more for the former than the latter).   Brown rat (Rattus norvegicus) CC130 MSSA Germany [40] Yellow-necked mouse t208/ST49, t4189/ST49, t1773/ST890, t843/ST130 ND Germany [42] House mouse t843/ST130 ND Germany [42] Bank vole t208/ST49, t4189/ST49 ND Germany [42] Bank vole (Myodes glareolus) CC49, ST890, ST1959 MSSA Germany [40] Common vole t1773/ST890, t15027/ST3252, t3058/ST3252, t3830/ST1956 ND Germany, Czech Republic [42] Field vole t1736/ST890, t2311/ST88, t3830/ST1956 ND Germany [42]

Molecular Types
An examination of the molecular types found in wildlife demonstrates an extensive diversity of types of S. aureus. However, some broad conclusions can be suggested. Though comparisons across publications are difficult due to divergent methodology of sampling, testing, and geography, Figure 2 illustrates the most common molecular types, according to the total count publications identifying them. These molecular types include a mix of human pandemic types (ST5, ST8, ST1, ST30, ST22) [82] and molecular types that have been more commonly described in animals or at the animal-human interface (ST398, ST130, ST133, ST425) [83,84].

Molecular Types
An examination of the molecular types found in wildlife demonstrates an extensive diversity of types of S. aureus. However, some broad conclusions can be suggested. Though comparisons across publications are difficult due to divergent methodology of sampling, testing, and geography, Figure  2 illustrates the most common molecular types, according to the total count publications identifying them. These molecular types include a mix of human pandemic types (ST5, ST8, ST1, ST30, ST22) [82] and molecular types that have been more commonly described in animals or at the animal-human interface (ST398, ST130, ST133, ST425) [83,84].

Discussion
This review demonstrates a significant amount of diversity in Staphylococcus aureus sampled from a wide variety of wildlife species across several continents. Populations of S. aureus present in wildlife may serve as reservoirs that could be transmitted to nearby domestic livestock or poultry or directly or indirectly to humans. Such a reservoir of S. aureus in the environment may also contribute to the exchange of antibiotic resistance or virulence genes among human or animal S. aureus, potentially leading to novel strains.
The continuing encroachment of humans into animal spaces due to agriculture, deforestation, climate change can lead to "spillovers" of pathogens from one species to another [85]. Most commonly we examine this with wildlife as a reservoir and humans as the affected species (e.g., Ebola, Nipah, MERS, SARS). However, transmission may also occur in reverse, with humans seeding wildlife with pathogens [86,87]. In the case of S. aureus, it appears both may be occurring, as has been previously documented among livestock [9]. In the case of antibiotic-resistant pathogens, such bidirectional transmission may be direct, via contact between human and animal species. More likely in the case of wildlife species, transmission may be indirect, such as via environmental reservoirs of pathogens including water sources, soil, exposure to manure, air, and contact with contaminated fomites [88,89]. Transmission may also occur due to consumption of meat products contaminated with S. aureus, but sampling wildlife meat products is exceedingly difficult and has not been done in

Discussion
This review demonstrates a significant amount of diversity in Staphylococcus aureus sampled from a wide variety of wildlife species across several continents. Populations of S. aureus present in wildlife may serve as reservoirs that could be transmitted to nearby domestic livestock or poultry or directly or indirectly to humans. Such a reservoir of S. aureus in the environment may also contribute to the exchange of antibiotic resistance or virulence genes among human or animal S. aureus, potentially leading to novel strains.
The continuing encroachment of humans into animal spaces due to agriculture, deforestation, climate change can lead to "spillovers" of pathogens from one species to another [85]. Most commonly we examine this with wildlife as a reservoir and humans as the affected species (e.g., Ebola, Nipah, MERS, SARS). However, transmission may also occur in reverse, with humans seeding wildlife with pathogens [86,87]. In the case of S. aureus, it appears both may be occurring, as has been previously documented among livestock [9]. In the case of antibiotic-resistant pathogens, such bidirectional transmission may be direct, via contact between human and animal species. More likely in the case of wildlife species, transmission may be indirect, such as via environmental reservoirs of pathogens including water sources, soil, exposure to manure, air, and contact with contaminated fomites [88,89].
Transmission may also occur due to consumption of meat products contaminated with S. aureus, but sampling wildlife meat products is exceedingly difficult and has not been done in a systematic manner. Meat products from livestock are a potential way that livestock-associated strains of S. aureus may spread from farms to communities [1,89], but the impact of meat from wildlife sources (including various deer species and wild boar) which may play a role in transmission of S. aureus bacteria or resistant genes is less clear.
While few studies reviewed here examine the environment and wildlife at the same time, a study by Porrero et al. [90] found mecC-positive S. aureus in river water after the area had been found to be positive for ST425-mecC in wild boar and fallow deer at the same location [54], suggesting a shared source of exposure or transmission between the various animal species and/or the environment.
Interestingly, bats and non-human primates seem to have little overlap with other animal strains. Bat molecular types consisted primarily of newly identified spa and/or MLST, though ST15 was reported twice-in a straw-colored fruit bat in Nigeria, and a captive Egyptian fruit bat sampled in Denmark [49,50].
For primate S. aureus, the papers reviewed here represent a mix of primates raised in captivity (including zoos and research facilities) and those sampled in sanctuaries and parks. As such, intensity of contact with humans who may be carrying typical human strains of S. aureus varies widely, and the importance of cross-species transmission remains in debate. Human-to-primate transmission was suggested in a study of wild primates MRSA in Nepal [63] and primates in Gabon [70]. The reverse was suggested by examination of an ST3268 strain found in macaques in primate research facilities in Singapore [68] and the United States [66,67]; this molecular type was also found in macaques in a New York research facility [65], suggesting the need for screening of animals prior to export/import. While most reports suggest preponderance of primate-associated strains, testing in a Texas facility found that their animals were colonized primarily with USA300/ST8 strains, which are common in humans and suggestive of human-to-animal transmission. However, workers at the facility were not tested for carriage [72].
Though S. aureus strains were typically taken as colonizers from healthy animals, several primates were actively infected with S. aureus. A gorilla in a primate center in Gabon was found to have a large lesion on his back; the gorilla died suddenly, and autopsy also found S. aureus in tissue samples; all were spa type t148 [70]. Though this is a human-associated strain, sampling of caretakers did not show any colonized humans involved in the animal's care. In the Washington state facility, S. aureus was cultured from the wounds of two macaques, but both were likely primate strains (t15469/ST3268 and t13638/ST3268) [66]. Another publication from Korea documented a macaque with an acute necrotic lesion caused by MRSA, but molecular typing was not carried out [64].
How may exposure to human pathogens, including S. aureus, in great ape populations affect release of them back into wild from captivity? This is addressed in several publications, suggesting that primates from captivity may pose a risk to their wild brethren [74] due to carriage of organisms such as drug-resistant S. aureus. Others argue release still should be possible but caregivers should be screened, and those positive for S. aureus carriage should not have contact with infant apes, and post-release monitoring of animals should include screening for this bacterium [93]. This may be difficult given the high level of carriage found in wild primates (up to 100% of chimpanzees tested and 72% of lemurs) [73].
While most studies examined asymptomatic colonization of wildlife, in some reports, such as those from captive zoo animals ( [44]), a number of clinical infections could be examined. These infections included abscesses, bacteremia, bite wounds, and dermatitis, among other conditions. Common animal-associated lineages were found, including CC130, CC133, and bacteremia caused by CC398 in an African wildcat. There was considerable diversity among the infection isolates, though a few did share spa or ST/CC types including two cases of t208/ST49/CC49 infections in red squirrels, two cases of t1166/ST3269/CC133 infections in a black swan and white-face whistling duck, and two cases of t15307/ST133/CC133 in another white-face whistling duck and a Baikal teal. This again suggests the potential for exposure to a contaminated environmental source for some of these animals, including water or other shared habitats within the facility.
Other captivity-based studies document the potential for bidirectional transmission between humans and animals in these facilities. In the San Diego zoo, a MRSA outbreak was noted in 2008, with pustules documented on both an elephant calf and three caretakers. Twenty total caretakers were infected over the next month, and the calf was euthanized. Investigation determined that the calf's infection with MRSA type USA300 likely came from a colonized caretaker, as the other elephants tested were colonization-negative [60].
Isolates examined in collected studies include methicillin-resistant and methicillin -susceptibile S. aureus. This testing included a mix of phenotypic and molecular methods, with some studies employing both. With the discovery of mecC [11,94], some early papers examining phenotypic testing alone should be looked at with some skepticism, as mecC-positive S. aureus isolates do not always show up as MRSA phenotypically, which can hinder the detection of mecC-carrying isolates [55]. Indeed, wildlife may be a key reservoir for mecC, as its presence was noted in a number of European reports (see Table 1). Interestingly, mecC has not been reported in any isolates originating in the United States to date.
There are a number of limitations to this review. Sampling was concentrated in a small number of countries and a relatively limited number of animal species have been sampled in different geographic areas, making large-scale comparisons difficult. Sampling techniques and anatomical locations tested within animal species vary among research groups. Most studies employed some sort of live animal swabbing (of noses, throats, skin, cloaca, etc.), but several used feces or scat instead of live animal testing. The studies also differed significantly in molecular and antibiotic resistance testing reported, making generalizations across publications difficult. Access to many animal species is also likely a function of convenience rather than a systematic sampling of all organisms in a particular environment, leading to over-representation of some animals relative to their abundance and an under-representation of others. Additional sampling should be carried out in order to examine the continued evolution of S. aureus in wildlife, and to track any strains that may have an increased propensity for zoonotic spread and threat to human health.

Eligibility Criteria
Studies that reported the presence of S. aureus (methicillin-resistant or susceptible) in any species of wildlife were eligible for inclusion.

Information Sources and Search Strategies
PubMed, Web of Science, and peer-reviewed materials within ProQuest databases were searched in May 2019 to identify eligible studies. The following search terms were used "MRSA OR Methicillin Resistant Staphylococcus aureus OR Staphylococcus aureus AND wildlife." Reference lists of the identified studies were also checked for additional studies. "Wildlife" was defined as wild animals but also captive animals who would typically be wild (such as zoo elephants) and those on nature preserves. Captive animals used as livestock or poultry or otherwise farmed or used as pets or work animals were also excluded. Articles were limited to English language only. Articles were examined and duplicate articles were removed.
Titles and abstracts were examined and articles were retained when there was evidence of S. aureus colonization or infection reported within wildlife species as defined above. Citations which included information on S. aureus antibiotic resistance and/or molecular typing were included in Table 1 and were grouped by animal species type.

Conflicts of Interest:
The authors declare no conflict of interest.
Appendix A Table A1. Picture credits.