The prevalence, antibiotic resistance and mecA characterization of coagulase negative staphylococci recovered from non-healthcare settings in London, UK

Four hundred seventy-nine samples were collected from different environmental sites in various locations (n = 355) and from the hands of volunteers (n = 124) in London UK between April 2013 and Nov 2014. Environmental sites included hotels (n = 100), baby care facilities (n = 65), handbags (n = 43), supermarkets (n = 37), restaurants (n = 36), public transport (n = 54), and a public library (n = 20). All specimens were plated on Mannitol Salt Agar (Oxoid, Basingstoke, UK), and then incubated aerobically at 37 °C for 24–72 h). One or two colonies for each site were selected based on staphylococci morphology [4]. The colonies were then purified on Nutrient Agar (Oxoid, Basingstoke, UK).

All isolates were initially screened using Gram staining, catalase and coagulase tests. Those that demonstrated potential staphylococci characteristics were identified by Matrix-assisted laser desorption ionization time flight mass-spectroscopy (MALDI-TOF-MS, Microflex LT, Bruker Daltonics, Coventry, UK) in a positive linear mode (2000–20,000 m/z range) as described previously [12]. The resulting spectra were compared with reference spectra by using the Biotyper 3.0 software (Bruker Daltonics, Coventry, UK). Escherichia. coli DH5α (Bruker Daltonics, Coventry, UK) was used as a standard for calibration and quality control.

A panel of 11 antibiotics was used to determine the antibiotic susceptibility of all the isolates. The standard disk diffusion method was used to test AM: amoxicillin (10 μg); CEP: cefepime (30 μg); CHL: chloramphenicol (30 μg); ERY: erythromycin (5 μg); FC: fusidic acid (10 μg); GEN: gentamicin (10 μg); MUP: mupirocin (20 μg); OX: oxacillin (1 μg); PEN: penicillin (1 unit); STR: streptomycin (10 μg); TET: tetracycline (10 μg);. The susceptible, intermediate resistant or resistant were determined by the Guidelines for Susceptibility Testing [14]. The Minimum Inhibitory Concentrations (MIC) for oxacillin were additionally evaluated using “M.I.C. evaluators” (Oxoid Ltd., Basingstoke, UK).

The mecA gene was determined by using PCR method as described previously [15]. For mecA positive isolates, SCCmec types were determined by evaluating mec and ccr complexes [15].

Multi-locus sequence typing (MLST) was used to determine the sequence types of S. epidermidis [16]. Sequence types were assigned using the S. epidermidis database (www.​mlst.​net).

A total of 643 staphylococci isolates were recovered in this study, including those from hotels (n = 74), baby care facilities (n = 46), handbags (n = 17), supermarkets (n = 89), restaurants (n = 96), public transportation (n = 94), human hands (n = 192) and public libraries (n = 35) (Additional file 1: Table S1).

Six hundred forty-three staphylococcal isolates belonging to 19 staphylococcal species were identified in this study. This included: S. epidermidis (n = 193), S. hominis (n = 161), S. capitis (n = 77), S. warneri (n = 63), S. haemolyticus (n = 45), S. pasteuri (n = 33), S. saprophyticus (n = 20), S. aureus (n = 12), S. simiae (n = 10), S. cohnii (n = 9), S. sciuri (n = 5), S. pettenkoferi (n = 3), S. auricularis (n = 2), S. caprae (n = 2), S. equorum (n = 2), S. lugdunensis (n = 2), S. xylosus (n = 2), S. arlettae (n = 1), and S. simulans (n = 1). S. epidermidis was the predominant species, followed by S. hominis, S. capitis, S. warneri, S. haemolyticus, S. pasteuri, and S. saprophyticus However, the occurrence of the species varied for different sites. S. epidermidis was predominant among the isolates recovered from restaurants, public transport, hands and handbags, whereas S. hominis was predominant among the isolates recovered from supermarkets, baby care facilities and hotels and S. haemolyticus was predominantly isolated from the library (Table 1).

The disc diffusion method was used to test 606 isolates against a panel of 11 antibiotics. 572 (94%) isolates were resistant to at least one antibiotic, and only 34 isolates were fully susceptible. Resistance to penicillin, and fusidic acid was observed in more than 65% of all staphylococcal isolates tested. 202 (33%) isolates were resistant to streptomycin, 190 (31%) to erythromycin, 161 (27%) to amoxicillin, 98 (16%) to tetracycline, 87 (14%) to mupirocin, 59 (10%) to gentamicin, 48 (8%) cefepime, 36 (6%) oxacillin, and 21(3%) chloramphenicol (Table 2).

Sixty-eight (11%) mecA positive staphylococcal isolates were determined, however, no MRSA was determined in this study. S. sciuri had the highest mecA gene carriage (80%) among all 19 staphylococcal species, followed by S. cohnii (33%), S. haemolyticus (22%), and S. saprothyticus (20%). Other isolates demonstrated relatively lower carriage of mecA gene, including S.hominis (3%), S.capitis (8%), S. epidermidis (11%), S.warneri (11%), S.pasteuri (13%). No mecA gene was found in the remaining 10 species, including S. aureus, S. simiae, S. equorum, S. caprae, S. xylosus, S. auricularis, S. simulans, S. arlettae S.pettenkoferi, and S. lugdunensis.

SCCmec types were fully determined in forty-six isolates. Twenty-two out of 68 isolates lacked either the mec gene complex or the ccr gene complex. Thirteen staphylococci (19%) carried SCCmec type V, followed by 8 isolates carrying SCCmec type I (2%), 5 isolates SCCmec type IV (7%), 4 isolates SCCmec type II (6%), 1 isolate SCCmec type III (2%), 1 isolate SCCmec type VI (2%), and 1 isolate SCCmec type VIII (2%). In addition, three isolates harboured a new SCCmec type 1A, which carried combination of class A mec complex and ccr type 1. Of the ten isolates that were non-typeable, three carried a combination of class A mec complex and ccrC, six carried a combination of class B mec and ccrC, and one carried class B mec and ccr type 3 (Table 3).

MLST was performed to determine the housekeeping genes of 13 oxacillin resistant and mecA positive S. epidermidis. MLST typing revealed that all S. epidermidis strains possess new MLST types. MLST types of S. epidermidis isolates with in house numbers of 279, 133, 134, 135, 126, 259, 124, 127, 234, 187, 308, 153 and 191 were respectively assigned as ST599, ST600, ST600, ST600, ST601, ST602, ST602, ST603, ST604, ST605, ST606, ST607 and ST608 (Table 4). Three S. epidermidis isolates shared the same sequence types (ST), including S. epidermidis 133, 134 and 135 that were isolated from different sites of a library (DSL) possessed ST600 whereas S. epidermidis 259, and S. epidermidis 124 that had ST602 sequence type were isolated from the human hands (HH) and different sites of hotels (DSH) respectively.

Although antibiotic resistance is commonly linked to the clinic, recent studies from different ecological niches revealed multidrug resistant bacteria is widespread in the environment [11, 12, 17].

We have previously reported on high levels of antibiotic resistance in staphylococci isolated from different environmental/public settings [11, 12]. In this study we evaluated the dissemination of antibiotic resistant staphylococci recovered from a wide range of environmental settings, and characterised the carriage of the mecA gene and the diversity of SCCmec elements in these isolates.

Six hundred and forty-three staphylococci isolates belonging to 19 species, including S. epidermidis, S. hominis, S. haemolyticus, S. capitis, S. warneri, S. pasteuri, S. saprophyticus, S. cohnii, S. aureus, S. simiae, S. sciuri, S. pettenkoferi, S. lugdunensis, S. equorum, S. caprae, S. xylosus, S. auricularis, S. simulans, and S. arlettae, were identified in this study. Interestingly, many of the staphylococci species recovered in our study have previously been associated with the community, preserved food, and wildlife [4, 10].

Antibiotic resistance of staphylococci associated with healthcare settings is well documented, however, little is known about the antibiotic resistance in staphylococci isolated from different ecological niches [4]. In this study, the majority of staphylococci were resistant to penicillin (65%) and fusidic acid (66%) (Fig. 1). Despite that 80% of hospital associated CoNS (across Europe) were reported to be resistant to oxacillin [18], only 6% of CoNS were resistant to oxacillin in this study. In addition, the levels of resistance to chloramphenicol (3%), cefepime (8%), gentamicin (10%), mupirocin (14%), tetracycline (16%), and erythromycin (31%) were lower compared to those reported in clinical settings [19‐22]. In contrast, the rates of resistance to fusidic acid (66%), amoxicillin (27%) and streptomycin (33%) in environmental staphylococcal isolates were higher than those reported in clinical staphylococci isolates [21, 23, 24]. It is widely accepted that higher levels of antibiotic resistance in clinical isolates are due to consistent antibiotic exposure [25]. The environment may also contribute to the development of antibiotic resistance in microorganisms due to human/ animal therapeutics, sewage, agriculture and industrial use of antibiotics [26]. Therefore, the wide dissemination of multidrug resistant CoNS in non-healthcare associated environments is a disturbing finding. In our study, 94% of staphylococcal isolates were phenotypically resistant to at least 1 antibiotic, 18% were resistant to five or more antibiotics and only 6% staphylococcal isolates were fully susceptible. The study also revealed that the number of isolates resistant to multiple antibiotics varied between the different isolation sites. The least number of multiple antibiotic resistant CoNS isolates were recovered from the public transport (58%), the highest was isolated from hotels (78%).

Methicillin resistant staphylococci pose a major public health threat, and cause severe economic and health consequences [27]. Methicillin resistance is determined by the mecA gene, which encodes for penicillin binding protein 2a (PBP2a) that has a low affinity to β-lactam antibiotics [28]. Hussain et al. assessed the correlation between mecA gene and oxacillin susceptibility breakpoints (0.5 mg l^− 1) of 493 clinical CoNS belonging to and classified into 4 categories [29]. The mecA gene positive staphylococci were categorized into groups I and II, and demonstrated that group I (S. haemolyticus (83.3%), S. epidermidis (61.9%), S. hominis (51.8%)) differs from group II (S. cohnii (28.5%), S. warneri (27.3%), S. saprophyticus (9.0%)) by their high levels of mecA-carriage [29]. Interestingly, S. hominis (38%), S. haemolyticus (22%), and S. epidermidis (7%) isolated in this study harboured significantly lower levels of the mecA gene. Moreover, in this study S. cohnii (33%) and S. saprophyticus (10%) showed higher mecA gene carriage than clinical isolates reported by Hussain, et al. [29], whereas the levels of mecA gene carriage in S. warneri (6%) were lower than in clinical isolates. No mecA gene was detected in staphylococcal species of groups III and IV, which included S. xylosus, S. lugdunensis, S. capitis, S.simulans, and S. schleiferi [29]. Similarly, in this study S. lugdunensis, S. xylosus and S. simulans were determined to be susceptible to oxacillin and lacked mecA gene. However, in contrast to the reports by Hussain, et al. [29] we found that mecA gene was present in 8% of S. capitis isolates.

Oxacillin susceptible mecA gene positive S. aureus (OS-MRSA) has been reported worldwide, and the risk of induced high levels of oxacillin resistance was determined in OS-MRSA [30, 31]. In this study, 68 (46%) staphylococcal isolates were confirmed by PCR to carry the mecA gene, however, they were phenotypically susceptible to oxacillin with the MICs (oxacillin) varying from 0.015 to 2 mg l^− 1. This study demonstrates the prevalence of mecA positive but oxacillin susceptible CoNS (OS-CoNS) in the environment. Little is known about OS-CoNS isolates recovered from the environment and their epidemiological data are limited. Additional studies are necessary to further our understanding of the prevalence and molecular epidemiology of OS-CoNS in the environment.

SCCmec is a mobile genetic element with two essential components: the mec gene complex, and the cassette chromosome recombinase (ccr) gene complex [32]. The combination of the mec gene complex and ccr gene complex confers different SCCmec types [32]. SCCmec type I, II, III are reported to be associated with MRSA recovered from healthcare settings, whereas SCCmec type IV and V are mainly associated with the community [32]. Moreover, it has been shown that the size of SCCmec types IV and V are smaller than SCCmec types I, II and III, thus conferring increased mobility by their smaller size and contributing the spread of these smaller SCCmec elements [33] . In this study, SCCmec type I, II or III were found in 19% (n = 13) of mecA-positive CoNS, whereas 27% (n = 18) of CoNS were determined to harbour SCCmec type IV or V. SCCmec type VI and VIII were previously identified in Portugal (2006) and Canada (2009) in hospital associated MRSA (HA-MRSA) [33, 34]. In this study, we identified one of each type, however, we did not detect SCCmec types IX.

Becker et al., have previously summarized the community and livestock associated staphylococcal species and their SCCmec types, which included S. capitis (I, IA, II, III, IV, IVa, V, non-typeable: (NT)), S. cohnii (NT), S. epidermidis (I, IIa, IIb, III, III (variant), IV, IVa, IVb, IVc, IVd, IVe, IVg, V, VI, NT), S. haemolyticus (I, II, II.1, III, III (variant), IV, V, NT), S. honomis (I, III, IV, NT), S. pasteuri (IVc), S. saprophyticus (III, NT), S. sciuri (I, III, IIIA, V, VII, NT) and S. warneri (IV, IV.1, IVb, IVE) [4]. In this study, species associated SCCmec types differed and included the following: S. capitis (I, NT), S. haemolyticus (I, II, V, NT) and S. hominis (I, V, NT), S. cohnii (I, V, NT), S. pasteuri (NT), S. saprophyticus (IV, NT), S. sciuri (II, VIII), S. warneri (I, V, NT). S. epidermidis possessed SCCmec types similar to those reported previously [4].

Thirteen unclassified SCCmec types were determined in this study, including three carrying class A mec complex and ccrC, six had a combination of class B mec and ccrC, one carried class B mec and ccr3, and three had a combination of class A mec complex and ccr type 1. The 1A was previously defined as a new SCCmec type 1A by others [35]. Pseudo (ψ)-SCCmec harbours the mec complex but lacks ccr, while, SCCmec12263 is reported to carry the ccr complex but lacks mec complex [36, 37]. In this study, 21 isolates (29%) were categorized as (ψ)-SCCmec and SCCmec12263 since they lacked either mec complex or ccr complexes. ψ SCC element is characterized by lacking genes for ccr and mec [4]. One of S. saprophyticus isolates in this study was found to possess the ψ SCC element (Table 5).

Whilst many studies have reported on the changing epidemiology of S. aureus, epidemiological data of other staphylococcal species are limited [38, 39]. In this study, 10 new MLST types were determined in 13 S. epidermidis isolates. Interestingly, although isolates recovered from human hands (S. epidermidis 259/ SCCmec V) and hotels (S. epidermidis 124/ SCCmec IV) harboured different SCCmec types, they shared the same MLST type ST602. In addition, three S. epidermidis isolates recovered from libraries (S. epidermidis 133, S. epidermidis 134, S. epidermidis 135) shared the same MLST type ST600 (Table 4). However, despite sharing the same MLST type S. epidermidis 133, S. epidermidis 134 and S. epidermidis 135 harbored SCCmec type 3B, I, IV respectively. Others reported that S. epidermidis ST2 was associated with type II, III, IV and non-typable SCCmec, and S. epidermidis ST22 harboured SCCmec type III, IV and V [40].