Staphylococcus aureus is one of the most common human pathogens, generally associated with skin and soft tissue infections (SSTIs) due to its affinity for skin and mucous membranes, and may cause bacteremia and other persistent chronic diseases, leading to high morbidity and mortality in children (Tong et al., 2015; Chalmers and Wylam, 2020). In children, S. aureus is not only one of the leading pathogens causing SSTIs but also a common cause of bloodstream infection (BSI), commonly as a secondary infection or being associated with more virulent S. aureus strains, such as Panton–Valentine leukocidin (PVL) producing strains (Kaplan, 2006; Kalu et al., 2022).

S. aureus is continually evolving and progressively acquiring resistance mechanisms to the majority of antibiotics (Cassat and Thomsen, 2021; Pantosti et al., 2007). After the spread of penicillin-resistant S. aureus strains worldwide, the semi-synthetic penicillin, methicillin, was introduced in clinical practice. Unfortunately, strains resistant not only to methicillin but to the entire class of β-lactam antibiotics emerged shortly thereafter (Chambers and DeLeo, 2009), becoming one of the primary threats of antibiotic resistance worldwide, as recognized by the World Health Organization (WHO) (Tacconelli et al., 2018). Methicillin resistance in S. aureus is due to the production of an additional penicillin-binding protein (PBP), designated PBP2a, which is not inhibited by β-lactams. PBP2a is codified by the mecA gene and transported in a 30–60 kb element, denominated staphylococcal chromosomal cassette (SCC)mec (Pantosti et al., 2007). In addition, homologs of mecA, such as mecC, have been described. SCCmec elements are classified into types and subtypes. SCCmec types are based on the combination of the mec gene complex and the ccr gene complex types, with fourteen SCCmec types up to date, described using Roman numerals (Uehara, 2022). The classification of SCCmec subtypes is based on J (“junkyard”) regions, which may carry additional antimicrobial resistance determinants and may be located in different positions (Uehara, 2022).

In Spain, methicillin-resistant S. aureus (MRSA) prevalence in blood samples ranged between 23 and 25.8% in the last years, according to the data reported by the European Antimicrobial Resistance Surveillance Network (EARS-Net) (The European Committee on Antimicrobial Susceptibility Testing, 2023). MRSA isolates are commonly associated with hospitalization (HA-MRSA). However, in the mid-1990s, clinically and genetically distinct MRSA strains emerged, affecting healthy children and adults lacking traditional risk factors (Gorak et al., 1999; Elston, 2007). These strains, called community-associated MRSA (CA-MRSA), were mostly related to SSTIs and could be more virulent, often producing PVL cytotoxin (Boyle-Vavra and Daum, 2007; DeLeo et al., 2010). CA-MRSA isolates are typically associated with SCCmec types IV and V, which are usually composed solely of methicillin-resistant genes and thus are smaller SCCmec elements, being more susceptible to non-β-lactam antibiotics, including several orally available agents, compared to HA-MRSA (Herold, 1998; David and Daum, 2010).

Different CA-MRSA lineages have been described worldwide, with sequence type 80 (ST80) usually carrying an SCCmec type IV as the most common clone in Europe, North Africa, and the Middle East (Mairi et al., 2020). Other common lineages described are ST30-IV (Southwest Pacific clone) in East Asia and Oceania, ST1-IV (USA400 clone) and ST8-IVa (USA300 clone) in the United States (Mairi et al., 2020). In Spain, the most common clone is the PVL-positive ST8-IVc, while ST30-IVc, ST80-IVc, and ST5-IVc are less prevalent (Vindel et al., 2014).

The prevalence of CA-MRSA in the pediatric population varies among countries. However, large-scale multicenter studies are limited, and the majority of the studies did not provide detailed molecular-level information. A retrospective study in a hospital in North Carolina reported a 75.9% prevalence of CA-MRSA in children over 7.5 months in 2006 (Shapiro et al., 2009). A study including 39 pediatric hospitals in the USA observed a decrease in MRSA infections by 52% from 2009 to 2016. However, the study design did not allow for distinguishing between hospital-onset vs. community-onset infections (Spaulding et al., 2018). A retrospective observational study of a 10-year review in a hospital in Taiwan reported 56.8% (363/639) of CA-MRSA among children (Yueh et al., 2022). The prevalence of CA-MRSA colonizations/infections was 62% (31/50) in Birmingham, UK, data from three hospitals with pediatric services (Adedeji et al., 2007). A cross-sectional study in Paraguay reported an increase of CA-MRSA prevalence in children from 21% (24/113) between 2009 and 2010 to 54% (91/168) between 2012 and 2013, with a dominant clone, ST30-IV-t019 (Rodriguez et al., 2020).

In Spain, CA-MRSA infections in children were reported for the first time in 2006, these isolates were PVL positive and had the SCCmec-IV (Broseta et al., 2006). Another study in 2007 in an emergency department in Madrid, Spain, found 13.2% (7/53) cases of CA-MRSA causing SSTI. PVL-positive isolates were more likely to cause severe local disease and belonged to ST8 (Daskalaki et al., 2010). A recent multicenter study involving children from community settings across Spain (the COSACO study—Colonization by S. aureus in the Community) showed a prevalence of MRSA colonization at 1.4%, with no strains producing PVL and associated with rural settings (Del Rosal et al., 2020).

Nevertheless, up-to-date data regarding CA-MRSA infections in children in Spain are scarce, and still, little information is available on the genetic traits of these strains. This study aimed to characterize CA-MRSA isolates causing pediatric infections in Spain using whole-genome sequencing (WGS) and to move forward to the implementation of WGS in the surveillance of MRSA infections following the roadmap set by the European Centre for Disease Prevention and Control (ECDC) (European Centre for Disease Prevention and Control, 2019). This study gathers genomic information about (i) genetic relatedness of CA-MRSA isolates causing pediatric infections in Spain using WGS-based typing methods, including spa-type, multilocus sequencing typing (MLST), core-genome multilocus sequence typing (cgMLST), and core-genome single-nucleotide polymorphism (SNP)-based analysis; (ii) antibiotic resistance mechanisms; and (iii) key virulence factors.

Ninety-eight CA-MRSA isolates from pediatric infections, mostly SSTIs, in Spain were analyzed using WGS in this study. ST8, t008, and SCCmec type IV (2B) were the most common lineages, but still, the number of different STs found in this collection highlights the diversity of CA-MRSA isolates able to cause infections among children. The ST8-SCCmec IVc lineage was previously described as the most common in Spain in a study during 2004–2012, including the children and adult population (Vindel et al., 2014). Other common lineages within this collection, such as the Southwest Pacific ST30-IVc clone and the African ST88-IVa clone, have also been previously reported in Spain (Vindel et al., 2014) and other countries (Otter and French, 2010; Breurec et al., 2011). The ST1472-IVa clone has been previously described in Spain but associated with methicillin-susceptible S. aureus (MSSA) (Gasch et al., 2012). Two ST398-Vc isolates, an originally livestock-associated clade, were identified in this collection; however, one of these isolates was positive for scn virulence gene—a human-specific immune evasion genetic marker (Zhao et al., 2020)—and negative for tet(M) tetracycline resistance gene—a genetic marker of the pig-associated clade (Price et al., 2012)—what indicates human-adapted ST398 MRSA (Laumay et al., 2021). Recently, an increase in macrolide resistance in MSSA isolates from blood has been described in Spain, associated with the spreading of ST398/ermT isolates, 78.6% of them had the scn virulence gene, and 14.3% had tetracycline resistance tet(M) gene (El Mammery et al., 2023).

In contrast, finding clusters of genetically related CA-MRSA isolates (≤5 allele differences by cgMLST and 0–6 core-genome SNPs), within the same hospital and from geographically separated hospitals, suggests a possible common origin and patient-to-patient transmission of CA-MRSA. A study about household transmission of USA300 CA-MRSA in the United States using WGS demonstrated that single strains were transmitted within households and could persist for 2.5–8.5 years (Alam et al., 2015). In another study, an agent-based model was used to represent population behavior, locations, and contact patterns using CA-MRSA data from Chicago, IL, USA, which indicated that contact with colonized individuals in households was probably the primary source of CA-MRSA acquisition. The authors suggested that interrupting household transmission should be the main target to control CA-MRSA (Macal et al., 2014). Because of this scenario, in which an MRSA spreads to multiple members of the patient’s household or community, Salgado et al. suggested a different term, “community-onset,” to describe the patient’s location at the time of identification of MRSA and to avoid a wrong origin assignment (Salgado et al., 2003).

Approximately half of the CA-MRSA isolates in this study were additionally resistant to erythromycin, and approximately one-third were additionally resistant to tobramycin and tetracycline. Classically, CA-MRSA is usually susceptible to the majority of antibiotics other than methicillin and β-lactams, while multi-resistance is common in HA-MRSA isolates (Otto, 2013). The increase in macrolide consumption in Spain, mainly outpatient use of azithromycin, has been related to the increase in macrolide resistance in MSSA isolates from blood (El Mammery et al., 2023), and could also explain the high percentage of co-resistance to erythromycin in CA-MRSA in this study. Erythromycin resistance could be explained mainly by the plasmid-located erm(C), mph(C), and msr(A) genes. Both mph(C) and msr(A) genes were mainly found together and in the same contig; this association has been previously described (El Mammery et al., 2023). Tetracyclines are a possible treatment option for patients with community-onset MRSA SSTI. In this study, the plasmid-located tetK gene was the primary mechanism of tetracycline resistance, and the plasmid-located tetL and the chromosomal or transposonal tetM genes were rarely detected. The tetM gene (76%) followed by the tetK gene (73%) were vastly present in MRSA isolates in the European SENTRY program published in 2001, whereas tetK gene was the most common in MSSA isolates (Schmitz, 2001). In a recent study on MRSA CC398 from Spanish hospitals, all tetracycline-resistant isolates carried the tetM gene, and 75% of them carried the tetK gene as well (Ceballos et al., 2020). In this collection of CA-MRSA isolates causing infections in children, almost all isolates (96%) had a plasmid replicon gene, RCR plasmid type (84.7%) being slightly more prevalent than the tetha-replicating plasmid type (74.5%), whereas in a previous study analyzing 278 non-identical S. aureus plasmids, theta-type mechanism was more prevalent (57%) (Kwong et al., 2017). In both studies, the prevalence trend for RCR initiator types was Rep_trans > Rep_1 > RepL. Still, for tetha-replication plasmids, those using Rep_3 initiator were the highest in our study compared to RepA_N initiators in the study by Kwong et al. (2017).

The success of CA-MRSA causing infections is partly due to their enhanced virulence (Otto, 2013; Otto, 2010), such as via the acquisition of the phage-encoded PVL genes, which were highly present (75%) in this collection, as previously described in CA-MRSA isolates (Boyle-Vavra and Daum, 2007; DeLeo et al., 2010). The successful spread of USA300 CA-MRSA was related to genes present in the ACME. Only two (2%) ST8-SCCmec IVa isolates were ACME-positive in this study, representing the USA300 clone, since this element is limited to strain USA300 (Otto, 2013). A study on CA-MRSA in adults in Spain between 2004 and 2012 showed 8.9% of the USA300 clone (ST8-IVa-ACME-positive) (Vindel et al., 2014). The evolution of CA-MRSA is related to the combination of different events, on one side, the acquisition of the SSCmec type IV, which is shorter and thus believed to cause less of a fitness burden, on the other side, an increased toxin expression and the acquisition of novel virulence genes on mobile genetic elements (MGE) (Otto, 2013; Otto, 2010). In this regard, all eta (exfoliative toxin A) and sea genes (enterotoxin A) were phage-located in this study as previously described (Turner et al., 2019). In contrast, the ɑ-hemolysin (Hla, ɑ-toxin) encoding gene, a cytolysin with pro-inflammatory effects that is produced by most S. aureus strains (Otto, 2013), was present in all CA-MRSA isolates in this collection, in addition to other cytolytic and pore-forming proteins, such as γ-hemolysin (HlgAB, HlgCB) and δ-hemolysin (δ-toxin). CA-MRSA strains commonly have high Agr activity, a regulator that may contribute to the increased expression of virulence factors (Otto, 2010). In this study, agr type I was the most prevalent as previously described in a Spanish study during the period 2004–2012 (Vindel et al., 2014), but agr type III is higher in the present study (32.6%) than before (10.2%) (Vindel et al., 2014).

This study has some limitations. Despite being one of the most extensive series to date on WGS-based typing of CA-MRSA causing infections in children, the sample size may not be sufficient to generalize the results to other regions or contexts. Although this is a multicenter study, our findings are limited to the population covered by those centers included in the study. The retrospective design, based on reviewing hospital records that were not standardized for this study, may have limited our findings. A more extensive prospective study at a national level, including all country regions, similar to the nationwide surveillance study on CA-MRSA nasal colonization in Spanish children (Del Rosal et al., 2020), would improve our knowledge on the molecular epidemiology and characteristics of MRSA causing infections in children. Furthermore, collecting epidemiological data on contacts, including households as important reservoirs, could help us understand the dynamics of S aureus acquisition and transmission, especially in SSTI among children.

Nasal colonization with MRSA has been correlated with a higher chance of MRSA infections (Von Eiff et al., 2001; Wertheim et al., 2005); nevertheless, colonization is mainly due to MSSA as described in individuals in the USA (Otto, 2010) and in Spanish children (Román et al., 2021), and only 1.5% of MRSA nasal carriage was found among individuals in the United States (Otto, 2010) and 1.4% among Spanish children (Román et al., 2021). Interestingly, when we comparatively analyzed previous CA-MRSA from nasal colonization in Spanish children (Letunic and Bork, 2021) with CA-MRSA isolates causing infections included in this study, we observed differences in the lineages and presence of virulence factors, with t002 and ST5 being the most prevalent in nasal carriers, with no PVL-positive isolates. Although both studies have been carried out at different time and in different populations and therefore comparisons should be carefully taken, these findings rise the possibility that infection in the absence of colonization could be possible, such as by direct body contact with infected individuals or contaminated fomites, and that even other body sites than the nostrils, such as throat, axilla, groin and perirectal area could be source of MRSA colonization (DeLeo et al., 2010).

In summary, we described a diverse collection of CA-MRSA isolates causing infections in children, mostly SSTIs, with ST8-IVc as the most prevalent lineage. The phylogenetic analysis results suggested possible contact transmission due to the presence of genetically related ST8-IVc isolates from the same hospital, but also possible common origin of ST5-IVc isolates from different hospitals and different geographic areas. A correlation between nasal colonization and infection has been accepted in the epidemiology and etiology of CA-MRSA isolates. However, we observed different lineages among CA-MRSA isolates causing infections and CA-MRSA isolates from nasal colonization in Spanish children. Future studies on CA-MRSA isolates causing infections and their colonization rate in the same population should be performed to better understand the extent to which MRSA colonization is involved in the development of CA-MRSA infections. Finally, this study represents a first step toward implementing WGS in CA-MRSA surveillance in Spain, following the roadmap set by the ECDC at the European level.