Nasal Staphylococcus aureus and S. pseudintermedius carriage in healthy dogs and cats: a systematic review of their antibiotic resistance, virulence and genetic lineages of zoonotic relevance

Abstract The molecular ecology of Staphylococcus aureus, Staphylococcus pseudintermedius and their methicillin‐resistant strains in healthy dogs and cats could serve as good models to understand the concept of bacterial zoonosis due to animal companionship. This study aims to provide insights into pooled prevalence, genetic lineages, virulence and antimicrobial resistance (AMR) among healthy dogs and cats. Original research and brief communication articles published from 2001 to 2021 that reported the nasal detection of S. aureus and S. pseudintermedius in healthy dogs and cats in the community, homes and outside veterinary clinics were examined and analysed. Forty‐nine studies were eligible and included in this systematic review. The pooled prevalence of nasal carriage of S. aureus/methicillin‐resistant S. aureus (MRSA) in healthy dogs and cats were 10.9% (95% CI: 10.1–11.9)/2.8% (95% CI: 2.4–3.2) and 3.2% (95% CI: 1.9–4.8)/0.5% (95% CI: 0.0–1.1), respectively. Conversely, the pooled prevalence of S. pseudintermedius/methicillin‐resistant S. pseudintermedius (MRSP) in healthy dogs and cats were 18.3% (95% CI: 17.1–19.7)/3.1% (95% CI: 2.5–3.7) and 1.3% (95% CI: 0.6–2.4)/1.2% (95% CI: 0.6–2.3), respectively. Although highly diverse genetic lineages of S. aureus were detected in healthy dogs and cats, MSSA‐CC1/CC5/CC22/CC45/CC121/CC398 and MRSA‐CC5/CC93/CC22/CC30 were mostly reported in dogs; and MSSA‐CC5/CC8/CC15/CC48 and MRSA‐CC22/CC30/CC80 in cats. Of note, MSSA‐CC398 isolates (spa‐types t034 and t5883) were detected in dogs. Genetic lineages often associated with MSSP/MRSP were ST20/ST71, highlighting the frequent detection of the epidemic European MRSP‐ST71 clone in dogs. S. aureus isolates carrying the luk‐S/F‐PV, tst, eta, etb and etd genes were seldomly detected in dogs, and luk‐S/F‐PV was the unique virulence factor reported in isolates of cats. S. pseudintermedius isolates harbouring the luk‐S/F‐I, seint and expA genes were frequently found, especially in dogs. High and diverse rates of AMR were noted, especially among MRSA/MRSP isolates. There is a need for additional studies on the molecular characterization of isolates from countries with under‐studied nasal staphylococci isolates.


INTRODUCTION
Staphylococcus aureus (S. aureus) and Staphylococcus pseudintermedius (S. pseudintermedius) are considered part of the nasal microbiota of healthy humans and dogs/cats, respectively, but can also become opportunistic pathogens (Abdullahi, Lozano, et al., 2021;. In the nasal cavities of these pets, S. pseudintermedius is more frequently detected than S. aureus (Bannoehr & Guardabassi, 2012;. S. pseudintermedius is a significant aetiological agent of pyoderma, soft tissue, urinary tract and ear infections in dogs (Bannoehr & Guardabassi, 2012;Lynch & Helbig, 2021). Although humans are not the natural host of S. pseudintermedius, this microorganism could cause infections in persons with close contact with infected/colonized dogs (zoonotic transmission) (Carroll et al., 2021;Lozano et al., 2017).
S. pseudintermedius, described in Devriese et al. (2005), belongs to the Staphylococcus intermedius group (SIG) which encompasses four additional species, viz; S. intermedius, S. delphini, S. cornubiensis and S. ursi (Carroll et al., 2021). The last two species have been very recently described (Murray et al., 2018;Perreten et al., 2020). S. pseudintermedius is the most common skin commensal in dogs (Bannoehr & Guardabassi, 2012;, but it is less prevalent in cats due to differences in skin cell adherence factors (Bierowiec et al., 2021;Mališová et al., 2019). This species can also be isolated in healthy donkeys or wild boar, among others Mama et al., 2019). In the earlier 2000s, S. pseudintermedius could not be accurately distinguished from S. intermedius using biochemical methods. Therefore, it was often incorrectly identified as S. intermedius (Börjesson et al., 2015).
Methicillin-resistant S. aureus (MRSA) represents a major challenge in human and animal infection control. Methicillin-resistant S. pseudintermedius (MRSP) has also become a global health problem in veterinary medicine and both S. aureus/MRSA and S. pseudintermedius/MRSP are of zoonotic relevance. The mecA or mecC genes are responsible for methicillin resistance and confer reduced affinity for beta-lactam antimicrobials, and mecA is found in most MRSA/MRSP isolates.
Staphylococcus aureus/MRSA nasal carriage in dogs and cats could be an indicator of increased risk of colonization/infection for persons in-contact with these animals by virtue of transmission processes in their shared household . Available molecular typing data revealed that companion animals could be carriers of S. aureus/MRSA human-associated clonal lineages of livestock-associated (LA) MRSA clones, which suggests reverse zoonosis and jump among hosts . In addition to MRSA, available literature has also highlighted MRSP as one of the most concerning multidrug-resistant staphylococci in pets' health. There have been increasing reports of MRSP colonization and infection with high multidrug resistance in dogs and cats (Bierowiec et al., 2021;Lord et al., 2022;Papić et al., 2021).
It is crucial to have close surveillance on MRSP and MRSA in companion animals in place to determine their burden and direct adequate infection control in veterinary hospitals and the human-pet interface. While the molecular epidemiology of clinical MRSA and MRSP isolates in small animals has often been explored, very few have assessed the nasal carriage of S. aureus and S. pseudintermedius in healthy dogs and cats. Regarding S. pseudintermedius, molecular analysis has revealed that the global MRSP population are very diverse and have been identified in pets (Pires Dos Santos et al., 2016). Using MultiLocus Sequence Typing (MLST), studies have demonstrated between-country variations in common sequence types (STs) (Perreten et al., 2010;Pires Dos Santos et al., 2016). For S. aureus in pets, the mainly reported genetic lineages of MRSA in dogs and cats often depend on the geographical location, health status of pets and occupation of pets' owners. For instance, communityassociated MRSA (eg. CC22, CC30, CC5) and LA-MRSA (e.g. CC398) clones have been widely reported (Aires-de-Sousa, 2017;Chueahiran et al., 2021;Ma, Worthing, Gottlieb, et al., 2020;Quitoco et al., 2013).
A recent global health research strategy emphasized the imperative need for an integrated surveillance system in the 'One Health' context of antimicrobial resistance (AMR) profiles of zoonotic bacteria including Staphylococcus spp (Dafale et al., 2020). Given the significant role of animals in the transmission of Staphylococcus the molecular characterization of isolates from countries with under-studied nasal staphylococci isolates.

K E Y W O R D S
healthy cats, healthy dogs, MRSA, MRSP, MRSP-ST71, MSSA-CC398, nasal carriage, Staphylococcus aureus, Staphylococcus pseudintermedius, zoonosis sp. across other ecosystems (human and environment), the present study was designed to investigate the pattern and pooled prevalence of S. aureus, MRSA, S. pseudintermedius and MRSP in healthy dogs and cats and to provide a comprehensive review, analysing extracted data on their AMR, virulence and genetic lineages in the past two decades over the world.

Study design
This systematic review was developed and executed on cross-sectional studies that reported S. aureus, S. pseudintermedius, MRSA and/or MRSP in the nasal cavities of healthy dogs and cats. This study did not require formal ethical approval, as the data used were obtained from previously published findings. This systematic review was based on the guidelines of Preferred Reporting Items for Systematic reviews and Meta-analyses (PRISMA) (http://prism a-state ment.org/PRISM Astat ement/ check list.aspx, accessed on 1 October 2021). All literature describing canine and feline isolates formerly identified as S. intermedius will be referred to as S. pseudintermedius, unless otherwise shown by genomic investigation. Special focus was given to MRSA, MRSA-CC398, MRSA-CC22, MRSA-CC1, MRSA-CC5 and MSSA-CC398 among dogs/ cats. Also, data about the AMR phenotypes and virulence genes were extracted.

Articles search strategy
This was a systematic and synthesizing review on nasal carriage of healthy dogs and cats by S. aureus, S. pseudintermedius, MRSA and/or MRSP. The entire literature search strategy, the selection of suitable published articles, the data extraction, the presentation of results and the statistical analyses were performed following the PRISMA guidelines. Published articles covering any or a combination of S. aureus, S. pseudintermedius and their methicillin-susceptible/resistant strains were reviewed using PubMed/Medline, Scopus, Google scholar, Ajol, Embase, ScieLo and Web of Science databases. Original researches and brief communications (reports) published (in English) from 1 January 2001 to 1 October 2021 were searched in these databases. Studies were searched and identified using the medical subject headings (MeSH) or article titles or abstracts. For that, the following MeSH key words combination were used: "nasal S. aureus carriage in dogs", "nasal S. aureus carriage in cats", "Methicillin-Resistant Staphylococcus aureus in dogs", "Methicillin-Resistant Staphylococcus aureus in cats", "nasal MRSA in healthy dogs", "nasal MSSA in healthy dogs", "nasal MRSA in healthy cats", "nasal MSSA in healthy cats", "Methicillin-Resistant Staphylococcus pseudintermedius in healthy cats", "Methicillin-Resistant Staphylococcus pseudintermedius in healthy dogs", "nasal MRSP in healthy dogs", "nasal MRSP in healthy cats", "nasal Staphylococcus pseudintermedius in healthy dogs" and "nasal Staphylococcus pseudintermedius in healthy cats". All eligible studies were scrutinized to identify potential articles from their reference lists. This study did not limit the search by any geographical location.

Inclusion criteria
Articles with appropriate and sufficient data about the prevalence of 'S. aureus, S. pseudintermedius, MRSA and/ or MRSP nasal carriage', in healthy dogs/cats (pets in the community, outside veterinary clinics) were selected and extensively reviewed. Data from these studies were extracted and utilized to determine the pooled prevalences of S. aureus, S. pseudintermedius, MRSA and MRSP nasal carriages in healthy dogs and cats (studies with sample size higher than 10). To avoid error from data selections, the pooled prevalence values were computed from only studies with a cross-sectional design. A dog or cat was considered healthy if there was no reported history of current disease or illness and no abnormalities detected on full physical examination in the eligible study. The MRSA or MRSP status was defined by the identification of resistance by phenotypic methods (using oxacillin or cefoxitin, according to the susceptibility method used and the microorganism) and/or mecA positivity by polymerase chain reaction (PCR) or agglutination tests.
All original or brief communication studies on the molecular typing of S. aureus, S. pseudintermedius, MRSA and/or MRSP in healthy dogs and cats regardless of whether they were cross-sectional or not were considered and included in the systematic review of genetic lineages of data of the isolates. Also, this was regardless of the number of animals tested in contrast to the criterion for determination of pooled prevalence (i.e. a sample size of >10). The search also included S. intermedius as well for the dates prior to the accurate discrimination of S. intermedius and S. pseudintermedius.

Exclusion criteria
(i) Studies that contained duplicate data or were overlapping articles; (ii) reviews and conference abstracts; (iii) cross-sectional studies that included fewer than 10 dogs or cats; (iv) studies on sick or healthy dogs and cats in the veterinary clinics and hospitals, (v) studies on dogs or cats owned by a veterinarian, (iv) studies on wound scrapings, buccal, tracheal, skin, ear, faecal and other animal samples were excluded.

Data extraction process
Where possible and appropriate, authors; year of study; study design; study setting or location; the number of S. aureus, S. pseudintermedius, MRSP, MRSA, MSSP, and/or MSSA isolates; type of specimen; laboratory method employed for detection; antimicrobial susceptibility phenotypes and corresponding genotypes and molecular types were reviewed and extracted. When there was disagreement, the relevant paper was reviewed, and the differences were resolved by the authors' consensus. Finally, 49 full-text articles were included. These directly focused on the distribution pattern of the S. aureus, MRSA, S. pseudintermedius, MRSP, genetic lineages, AMR phenotypes and genotypes, and/or virulence genes in nasal cavities of healthy dogs and cats.

Statistical analysis
The pooled prevalence of nasal carriage of S. aureus, MRSA, S. pseudintermedius and MRSP was calculated. MedCalc version 20.013 was used for all statistical analyses. Where possible, analyses of pooled prevalence and corresponding 95% confidence intervals (CI) were carried out using the random-effects model. The pooled prevalence was computed by combining the results of the numbers of positive cases and the total number of animals (studied) obtained from eligible multiple cross-sectional studies. Essentially, it involved dividing net sum positives (for S. aureus, S. pseudintermedius, MRSA, MRSP) by the entire population of healthy dogs or cats (included in each study).

Studies characteristics
Out of the 49 eligible and analysed studies (Figure 1), 47 and 13 were on S. aureus and/or S. pseudintermedius/S. intermedius nasal carriage in healthy dogs and cats, respectively (Table S1). Of these, three studies on healthy dogs or cats reported S. intermedius instead of S. pseudintermedius (Chasioti et al., 2019;Elnageh et al., 2021;Epstein et al., 2009); however, due to suspected misidentification, all the S. intermedius isolates for dogs/cats were considered as S. pseudintermedius for this systematic review. It is important to remark that some studies simultaneously investigated staphylococci in both dogs and cats. Table S1 shows the characteristics and data of the eligible studies, countries of the studies, type of pets (divided into dogs and cats), number of pets tested, number of S. aureus, S. pseudintermedius and MRSA/MRSP obtained from the studies, and the AMR phenotypes and genes, as well as the virulence factors detected from S. aureus and/or S. pseudintermedius isolates.

The pooled prevalence of nasal carriage of S. aureus, S. pseudintermedius, MRSA and MRSP in healthy dogs and cats
The pooled prevalence of nasal carriage of S. aureus in healthy dogs and cats was 10.9% (95% CI: 10.1-11.9) and 3.2% (95% CI: 1.9-4.8) respectively (Table 1). Although, relatively low when compared to the pooled prevalence of nasal S. aureus carriage in healthy humans (15.9%) and nasotracheal carriage in healthy wild animals (18.5%) Abdullahi, Lozano, et al., 2021), the pooled prevalence of nasal S. aureus carriage in healthy dogs was higher than in cats. The reason for this observation is not fully understood as both dogs and cats frequently come in close contact with humans in the household environment, showing a similar risk of S. aureus acquisition from humans.
From our analysis, the pooled prevalence of nasal carriage of S. pseudintermedius in dogs and cats was 18.3% (95% CI: 17.1-19.7) and 1.3% (95% CI: 0.6-2.4), respectively (Table 1). S. pseudintermedius carriage is frequently higher in healthy dogs than in healthy cats. Moreover, comparing S. pseudintemedius and S. aureus carriage, S. psedintermedius carriage is also higher than S. aureus in these animals. S. pseudintemedius is considered an opportunistic pathogen that could cause severe and necrotizing infections of the skin, ears, bones and post-surgical abscesses (Bannoehr & Guardabassi, 2012). Moreover, S. pseudintermedius could also be found in humans acquired by zoonotic transmission from colonized dogs (Carroll et al., 2021;Lozano et al., 2017;Somayaji et al., 2016) and may be misdiagnosed as S. aureus in human infections (Börjesson et al., 2015). To the best of our knowledge, no study determined the pooled prevalence of S. pseudintermedius nasal carriage in healthy dogs. Differences were observed in the prevalence of S. pseudintermedius nasal carriage among the eligible studies; for instance, some reported a very high prevalence (74.7%) (Garbacz et al., 2013). This difference in prevalence is multifactorial, it could be related to the environment of the dogs, and owners' and animals' health statuses.
In relation to MRSA, the growing concern about the spread of MRSA in communities has led to the establishment of recommendations for surveillance, including research into the rate of colonization in healthy dogs, including stray dogs present in public areas. The pooled prevalence of canine MRSA nasal carriage (2.8% [95% CI: 2.4-3.2]) was relatively higher than that of the feline carriage (0.5% [95% CI: 0.0-1.1]) ( Table 1). The exposure of pets to their living environment or persistently colonized humans or other animals can be considered a risk factor for colonization with MRSA strains; thus, it is essential to identify community and environmental reservoirs and sources of MRSA in healthy dogs and cats (Mohamed et al., 2020). MRSA nasal carriage in healthy cats was detected by five studies in Libya, the United States, Brazil, the United Kingdom and Saudi Arabia (Elmoslemany et al., 2021;Elnageh et al., 2021;Gingrich et al., 2011;Loeffler et al., 2011;Quitoco et al., 2013) (Table S1). Moreover, nasal MRSA in healthy dogs has been studied and detected in all continents (Table S1). From our analysis, the pooled prevalence of nasal carriage of MRSP in dogs and cats was 3.1% (95% CI: 2.5-3.7) and 1.2% (95% CI: 0.6-2.3) respectively (Table 1).
The complex relationships between bacterial species from different hosts facilitate genetic flow, extending AMR between humans, animals and the environment, resulting in public health concerns. Of this, methicillin resistance is one of the major and significant AMR of zoonotic importance. A companion animal may play an important role in the maintenance and transmission of MRSP or MRSA in the household .
It is important to remark that there was a variation in the pooled prevalence of S. aureus/S. pseudintermedius and MRSA/MRSP across the continents ( Figure 2). Nevertheless, when interpreting these data, it should be noted that they may be influenced by regional or local epidemiologies of S. aureus/S. pseudintermedius. Particularly, the pooled MRSA and MRSP prevalence were highest in Africa and Asia respectively. In these cases, these variations could also be due to differences in used methodologies, and other potential reasons such as pets care (especially animal hygiene), environmental sanitation or antibiotic use in animals and humans in these continents (Collignon & Voss, 2015;Fletcher, 2015;Valiakos et al., 2020).

Antibiotic resistance phenotypes and genotypes of S. aureus and S. pseudintermedius detected from nasal cavities of healthy dogs and cats
In the studies included in this systematic review, it appeared that both MRSA and MSSA had resistance against non-beta-lactam antimicrobials, such as fluoroquinolones, aminoglycosides, tetracyclines, sulphonamides, macrolides, and/or lincosamides (Tables 2 and 3). Based on the WHO classification of antimicrobials into highly important versus critically important classes of antimicrobial agents (World Health Organization, 2017), some antibiotics of zoonotic relevance in each class are discussed below.
Regarding methicillin resistance, out of the 21 studies that reported nasal MRSA in healthy dogs or cats, 17 investigated and detected the mecA gene. However, none reported the mecC gene (Tables 2 and 3). In dogs, S. aureus isolates were frequently resistant to tetracycline, aminoglycosides, rifampicin, sulphonamides and/or chloramphenicol (Table 2). S. pseudintermedius strains presented wider AMR patterns (phenotype and genotype) than those reported in nasal S. aureus of healthy dogs. In a recent study, S. pseudintermedius from dogs showed more frequently a MDR phenotype (45.5%) than S. aureus (40.9%) in a collection of isolates obtained between 2007 and 2016 (Lord et al., 2022).
Aside from the cefoxitin/oxacillin resistance detected in MRSA and MRSP, sulfamethoxazole-trimethoprim and ciprofloxacin resistances were the most predominant in healthy cats (reported in 4 and 3 out of 5 studies respectively). Other antibiotic resistances detected included erythromycin and clindamycin (inducible), amikacin, enrofloxacin and tetracycline (Table 3). Remarkably,  detected the resistance genes tet(M), erm(B), aphA, aadE and sat4 in MSSP isolates from healthy cats in Spain. Moreover, Rynhoud et al. (2021) detected some genes responsible for sulfamethoxazole-trimethoprim and aminoglycoside resistance in MRSP (dfrG and aac(6′)-aph(2″), aadD, ant6-Ia, and/or aph3-III respectively). A wide variation in S. aureus resistance patterns was detected in many studies, but penicillin and ciprofloxacin resistances were frequently detected in S. aureus and S. pseudintermedius that colonize healthy cats (Table 3). MDR phenotype has also been reported to be common among S. pseudintermedius strains from diseased pets (100.0%) than healthy ones (73.0%) (Rynhoud et al., 2021). Focusing on some antimicrobials of great importance for human and animal health (or for epidemiological purposes), tetracycline resistance was reported in healthy cats in only one study related to MSSA isolates in Saudi Arabia (Elmoslemany et al., 2021). In contrast, tetracycline resistance was detected in nasal MRSA and MSSA isolates from healthy dogs in 12 studies (Table 2). Specifically, out of the 61 pooled MRSA isolates from nasal cavities of healthy dogs, 24 (39.3%) were tetracycline resistant. However, eight (8.9%) out of the 89 pooled MSSA isolates from eligible studies were tetracycline resistant (data obtained from Table 2). The lineages MSSA-ST1212/t582, MSSA-ST121/t159 and MRSA-ST59 were detected among tetracycline-resistant isolates (Gharsa, Ben Slama, et al., 2015;Wan et al., 2012). The elevated frequency of tetracycline resistance found may be related to the frequent use of this antimicrobial agent (firstline) in dogs and cats. This phenotype of resistance has been proposed as a good marker of MRSA isolates belonging to clonal lineages associated with livestock animals, especially CC398 (Benito et al., 2014;Ceballos et al., 2020). Tetracycline resistance was also very common among S. pseudintermedius isolates.
In relation to nasal S. aureus strains from healthy dogs, many studies reported the detection of erythromycin resistance (Table 2). Similarly, two studies reported the erythromycin-clindamycin inducible resistance phenotype in nasal S. pseudintermedius isolates from healthy dogs and cats Van Balen et al., 2017).
Recently, an increasing penicillin susceptibility rate is being detected among invasive MSSA human isolates, opening therapeutic opportunities for these infections; this phenotype has been frequently found among scnnegative or CC398 isolates suggesting a potential animal association (Mama, Aspiroz, Lozano, et al., 2021). In this review, penicillin susceptibility was reported in 79.2% (n = 19) of the MSSA isolates detected from healthy dogs by Gómez-Sanz, Torres, Benito, et al. (2013), and seven of these isolates (36.8%) belonged to the CC398 clone.
In healthy dogs, four studies also reported varying ciprofloxacin resistance among nasal S. aureus isolates, with a range of 40.7%-100.0% (Abdel-Moein et al., 2012;Coelho et al., 2011;Myaing et al., 2016;Wedley et al., 2014). Of the four studies, the pooled prevalence of ciprofloxacin resistance in S. aureus nasal isolates from healthy dogs was 53.9% (55 out of 102 isolates). Ciprofloxacin resistance was reported in S. aureus isolates in two studies of healthy cats (Elnageh et al., 2021;Gharsa, Ben Slama, et al., 2015). None of the S. aureus studies reported the molecular mechanisms implicated. In the case of S. pseudintermedius, the GyrA [Ser84Leu, Glu714Lys, Ser84Leu] and GrlA [Ser80Ile] amino acids substitutions were reported from MRSP isolates of healthy dogs in one unique study ; these mutations are associated with decreased susceptibility to ciprofloxacin and other fluoroquinolones in staphylococci (Rynhoud et al., 2021).
Nine studies of healthy dogs or cats reported gentamicin resistance in S. aureus (Table 2), but only one of these studies indicated the bifunctional gene, aac(6′)-aph(2″) as the mechanism implicated (Coelho et al., 2011).

Genetic lineages associated with S. aureus isolates from the nasal cavities of healthy dogs and cats
In relation to nasal MSSA isolates from healthy dogs, the lineages CC1, CC5, CC22, CC45, CC121 and CC398 were predominant, although CC15, CC30, CC88, CC97 and CC133 were also reported; these lineages are generally associated with both humans and animals. Moreover, the lineages CC22 and CC30, that were identified among canine MRSA isolates, are known human-associated clones (Table 4). On the other hand, the lineages CC5, CC15, CC45 and CC8 were found in MSSA isolates of healthy cats, and CC22, CC80 and CC30 among feline MRSA isolates (Table 4). CC398 is a relevant clonal lineage due to its possible relationship with farm animals or veterinary farmers. In recent years, two CC398 clades have been proposed, one animal subclade more associated with MRSA and one human subclade related to MSSA. The CC398 lineage was reported in healthy dogs in only one study performed in Spain (Gómez-Sanz, Torres, Benito, et al., 2013), in which all isolates were MSSA. The MSSA-ST398 lineage represented 29.2% of the total S. aureus isolates recovered from the dogs in this study and no data about the possible contact with livestock were available for these animals. These MSSA-ST398 isolates presented the spa-types t034, t108 and t5883 (Table 4). Despite the methicillin susceptibility of these strains (that could suggest the human-related livestock-independent subclade), the detection of the spatype t034 and the absence of the IEC system (scn-marker) infer an animal (or livestock) origin . MSSA-CC398 isolates have also been detected in wild birds (such as storks), also in Spain, but the isolates presented the spa-types t571 and t6606; and were scn-positive (Gómez et al., 2016).
Regarding CC133, it is a well-known small ruminantassociated lineage (Guinane et al., 2010), that has been detected as a predominant S. aureus lineage in cats in Japan (Sasaki et al., 2012), and in a collection of large felines from a Danish Zoo (Espinosa-Gongora et al., 2012).
Among other STs isolated from dogs, ST1/CC1 was the most frequently reported in most eligible dogs' studies. It is a lineage that has no host specificity, as it has also been associated with rabbits, horses, cattle and swine (Alba et al., 2015;Feltrin et al., 2016;Islam et al., 2017;Silva et al., 2020). Conversely, MRSA-CC1 has often been linked to hospitalassociated outbreaks, especially in the intensive care unit (Monecke et al., 2020). The role of pets in the dissemination of S. aureus-CC1 isolates should be evaluated in the future. The ST22 lineage is found in many of the MRSA isolates from dogs and it is known as a human epidemic clone (EMRSA-15) (Katakweba et al., 2016). The EMRSA-15 (ST22) is predominantly a pandemic MRSA isolated both in hospital and community settings. It was estimated that ∼98% of all ST22 isolates reported in the literature have been associated with humans (Roberts et al., 2018). However, this ST has previously been isolated from dogs in other geographical regions, supporting the notion that this S. aureus lineage may have adapted to the canine host (Katakweba et al., 2016). Nevertheless, this clone may be of human origin and transmitted to dogs in the household environment. The other predominant lineage detected among MRSA from dogs was ST36, which is one of the three pandemic lineages that have been described within CC30 causing epidemic waves (Di Gregorio et al., 2021). Moreover, MRSA isolates of the genetic lineages ST5 and ST93 were also reported in nasal strains from healthy dogs in Australia .
The definition of community-and hospital-associated strains is no longer a strict one, as CA-MRSA strains have been described to be responsible for outbreaks in hospital settings, while HA-MRSA strains have been reported spreading in the community (Bal et al., 2016;Coll et al., 2017). However, it is crucial to delineate the pattern of predominant transmission and degree of spillover from community to hospital and vice-versa. The ability of MRSA-ST80 to disseminate at the community level has been reported (Drougka et al., 2016;Mairi et al., 2020). Furthermore, the transmission of MRSA-ST80 among different host species in the context of households and veterinary practices has been described by Drougka et al. (2016), who reported that dogs and cats can be the reservoir of MRSA-ST80 which can cause severe infection in humans. The so-called paediatric clone, CC5, has globally been associated with community outbreaks (McManus et al., 2021).

Genetic lineages associated with S. pseudintermedius isolates from the nasal cavities of healthy dogs and cats
Out of the eight studies with molecular typing on S. pseudintermedius in healthy dogs, a great diversity of lineages was reported among MSSP isolates, and no specific predominance was observed ( Table 5). As observed in the current review, high MLST clonal diversity is a common characteristic of MSSP (Pires Dos Santos et al., 2016). The diverse global population structures of S. pseudintermedius are evident in the world map comparing shared STs (Perreten et al., 2010).
Moreover, diverse lineages were also detected among MRSP of healthy dogs, although ST71 lineage was predominant (14.6% of isolates), detected in studies carried out in Spain, Brazil and Australia (Table 5). MRSP isolates of the lineage ST71 have largely been associated with sick dogs. Although the ST71 was considered a major MRSP-clone in Europe, recent studies have reported a downward trend in the prevalence of the MRSP-ST71 lineage among companion animals in France, Netherlands and Finland (Bergot et al., 2018;Duim et al., 2016;Grönthal et al., 2017). In relation to MRSP-ST71 from nasal swabs of healthy cats and dogs, other non-European countries have reported this clone, such as Australia (Rynhoud et al., 2021). It could be that veterinary personnel acted as carriers and facilitated the transmission of the MRSP-ST71 clone between animals they handle, or pet owners who interact with other dogs in the community such as in dogs' parks. The MRSP-ST71 has already been incriminated in potential zoonotic diseases (Kitagawa et al., 2021), reflecting the recent emergence of these important multidrug-resistant bacteria in animal and public health. Conversely, the other clone of high relevance, especially in the USA, MRSP-ST68, was not reported from nasal cavities of healthy dogs and cats in the studies included in this systematic review.
Although none of the MRSP isolates from cats was genetically characterized, an MSSP isolate was found to be ST142 by  ( Table 5). Although the MSSP-ST142 is a rarely reported lineage in cats, it has been detected in household dogs, and their owner; and could represent a case of transmission between dogs and cats in the same household  Gómez-Sanz et al., 2019).

Major virulence factors of S. aureus detected from nasal cavities of healthy dogs and cats
Genes of several virulence factors including leukotoxins, enterotoxins and exfoliatins were reported in some of the reviewed and eligible studies of healthy dogs and cats (Table 6). Eight MRSA strains carrying luk-S/F-PV, three MSSA eta-positive, one MSSA etb-positive, two MSSA etdpositive and one MRSA tst-positive were reported in five studies on healthy dogs. However, only one MRSA luk--S/F-PV-positive strain (of the lineage ST80) was detected in one study on healthy cats (Elmoslemany et al., 2021). Previous results have shown that luk-S/F-PV (encoding the Panton-Valentine Leucocidin, PVL) is considered a stable genetic marker for CA-MRSA and has been associated with necrotic skin lesions and necrotic pneumonia (Kong et al., 2016;Labandeira-Rey et al., 2007;Shallcross et al., 2013). The luk-S/F-PV gene was detected in three studies on healthy dogs; in one of them, three PVL-positive MRSA-ST30 isolates were reported in Brazil (Penna et al., 2021). Another study from Australia demonstrated MRSA-luk-S/F-PV in 4 out of 5 strains (3 ST93 and 1 ST5) from healthy dogs . However, the genetic lineage of the MRSA-luk-S/F-PVpositive strain from the study of Wachtmeister (2018) was not determined. Moreover, one MRSA-tst-positive strain (spa-type t022) was detected in a healthy dog in the United Kingdom (Wedley et al., 2014).

Major virulence factors of S. pseudintermedius detected from nasal cavities of healthy dogs and cats
Most of the S. pseudintermedius isolates in which virulence was analysed, carried a diverse range of virulence factors that are responsible for host-specific clinical symptoms such as the pyodermic infection on dogs and cats ( Table 7). The lukS/F-I, siet, expA genes were particularly often detected in nasal S. pseudintermedius isolates. MSSP-lukS/F-I positive isolates were reported in three studies involving dogs and cats (Gharsa et al., 2013;Gómez-Sanz, Torres, Benito, et al., 2013;. Similarly, siet-carrying S. pseudintermedius isolates were detected in eight studies on healthy dogs and cats. Moreover, staphylococcal enterotoxins genes (sea, seb, sec, sed, sei, sej, sek and ser) were also reported in some studies (Table 7). Biofilms produced by S. pseudintermedius play an important role in the pathophysiology of infection and colonization (Jain & Agarwal, 2009;Singh et al., 2013). Some studies reported that MRSP-ST71 can produce biofilms, suggesting the necessity of investigating the prevalence of biofilm-producing S. pseudintermedius and their implications on AMR in veterinary medical practices (Dicicco et al., 2012;Osland et al., 2012;Singh et al., 2013). The icaA and icaD genes responsible for biofilm production were detected among healthy dogs' isolates in one study (Han et al., 2015).

Host adaptation system of S. aureus isolates from nasal cavities of healthy dogs and cats
from the nasal samples of healthy dogs in the study of Ma, Worthing, Gottlieb, et al. (2020) were scn-positive.
As a limitation of this study, analysis of global data was performed using studies from low-resources and developed countries, that used different techniques for bacterial isolation and identification. The studies' methodologies differences might to some extent affect the prevalence values of S. aureus or S. pseudintermedius reported in healthy dogs and cats. Consequently, the pooled prevalence data obtained could be misreported (i.e. under-reported or over-reported).

CONCLUSION
It can be inferred that the nasal carriage of S. aureus and S. pseudintermedius in healthy dogs were moderate (10.9% and 18.3%) but low in healthy cats (3.2% and 1.2%). However, the nasal carriage of MRSA and MRSP in both healthy dogs and cats was low (<3.5%). Moreover, a high diversity of lineages and clones was detected in both staphylococcal species, especially among MSSP and MSSA. The MRSA strains recovered from healthy dogs and cats frequently belonged to human-associated genetic lineages which indicates the transmission of these clones within the household niche. Conversely, the MSSA strains corresponded to genetic lineages associated with both humans and animals. Both S. aureus and S. pseudintermedius showed high levels of AMR to commonly used human antibiotics. MRSP isolates from dogs and cats showed more frequently a MDR phenotype than MRSA. Moreover, some S. aureus isolates carried virulence factors of human health importance, as is the case of luk--S/F-PV, tst, eta, etb and etd. It is necessary to implement adequate epidemiological and microbiological surveillance on S. aureus/S. pseudintermedius and MRSA/MRSP in companion animals. Additional studies should include follow-up of molecular characterization of isolates from countries with under-studied nasal staphylococci isolates. By extension, studies should be more tailored towards the 'One Health' approach (humans-animals-and their shared environment).

FUNDING INFORMATION
This work was supported by the project PID2019-106158RB-I00 of the MCIN/AEI/10.13039/501100011033 of Spain. Also, it received funding from the European Union's H2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement N° 801586.

CONFLICT OF INTEREST
No conflict of interest declared.

DATA AVAILABILITY STATEMENT
The data extracted and utilized for this study have been thoroughly referenced. However, data related to the statistical analyses can be made available on request through the corresponding author (C.T.)