Introduction

Staphylococcus aureus, particularly methicillin-resistant S. aureus (MRSA), is a major human pathogen which causes both hospital- and community-acquired infections worldwide [1]. MRSA is commonly associated with skin and soft tissue infections (SSTIs), as well as pneumonia, bacteremia, and sepsis [2]. β-lactam antibiotic resistance in MRSA is attributed to the acquisition of mecA gene encoding the transpeptidase penicillin-binding protein 2a [3], and is a molecular hallmark for MRSA strains. The mecA gene is located on a mobile genetic element, the staphylococcal cassette chromosome mec (SCCmec), with at least 13 different types of SCCmec reported to date [4]. More recently, new mecA homologues mecB, mecC and mecD have also been identified to confer resistance to β-lactam antibiotics [5,6,7]. MRSA infections have been traditionally classified as either community-associated (CA-MRSA) or healthcare-associated MRSA (HA-MRSA). From a molecular standpoint, the CA-MRSA and HA-MRSA classifications are no longer distinct as patients can be colonized by MRSA in one setting and develop manifestations of infection in another. HA- and CA-MRSA infections have been conveniently used as epidemiological terms, however the line between them is significantly blurred. Although there are established healthcare-associated risk factors for HA-MRSA infection including recent hospitalization or surgery, residence in a long-term–care facility, dialysis, and indwelling percutaneous medical devices and catheters, community-onset HA-MRSA infections have been observed with increasing frequency among patients in community settings. Similarly, highly successful community-based clones have invaded the healthcare setting and have become successful nosocomial pathogens [8,9,10,11]. While demarcation of MRSA isolates as HA-MRSA or CA-MRSA can be confusing, there are clear differences in phenotypes and genetic background of MRSA strains associated with infection in either setting. HA-MRSA typically harbour SCCmec types I, II, or III, whereas CA-MRSA isolates carry SCCmec types IV, V and VI [12, 13].

Clinically, CA-MRSA is mainly associated with SSTIs [14] and often produce Panton-Valentine leukocidin (PVL), encoded by two co-transcribed genes, LukS-PV and LukF-PV [15, 16]. PVL is a bicomponent pore-forming cytotoxin assembled by LukS-PV and LukF-PV, has been demonstrated to have a significant role in the pathogenesis of MRSA by selectively targeting polymorphonuclear cells, macrophages and monocytes [17]. Epidemiological and clinical data provide substantial evidence that the high virulence potential of CA-MRSA isolates is associated with the expression of PVL [18, 19], which was also shown to be an important contributing factor in CA-MSSA infections, particularly in severe SSTIs and most notably in necrotizing pneumonia [14, 20, 21]. While HA-MRSA strains were initially observed not to be associated with PVL production [18], PVL genes carried by HA-MRSA strains have also been recently described in cutaneous and invasive infections [22, 23].

Evidently, the epidemiology of MRSA strains is constantly changing and their prevalence as well as molecular characteristics are known to vary between hospitals in different countries, cities within a country, or among wards of a hospital. Therefore, surveillance of the changing epidemiology of MRSA in local healthcare facilities with unique patient population is crucial for obtaining data that may aid empirical therapy and patient management. There are no reports on the molecular detection of virulence characteristics or their prevalence in MRSA isolates in Northern Cyprus (NC) in literature. This study aimed to investigate the prevalence of pvl in MRSA clinical isolates from patients admitted to a university hospital in NC using molecular methods.

Main text

Methods

Bacterial isolates

In this study, 50 S. aureus clinical isolates from patients admitted to Near East University Hospital between January 2012 and December 2019 which were processed at the hospital microbiology laboratory were randomly selected and included. Demographic data from each patient was anonymously collected and digitally stored. Bacterial isolates were initially cultured on blood agar media and were incubated at 37 °C for 18–24 h. All isolates were screened with coagulase test. Isolates were consequently processed with BD Phoenix 100 (Becton–Dickinson, BD Diagnostic Instrument Systems, USA) automated bacterial identification system according to supplier recommendations.

Antimicrobial susceptibility testing

Antimicrobial susceptibility of all isolates was initially performed by BD Phoenix 100 system. Consequently, susceptibility to cefoxitin in all S. aureus isolates identified as methicillin resistant with BD Phoenix 100 system was confirmed by agar disc diffusion assay and interpreted according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines [24]. Briefly, isolates were cultured on Mueller Hinton agar using 0.5 McFarland inoculum and cefoxitin disc (30 µg) was used for screening. All plates were incubated at 35 °C 5% CO2 for 24 h. Bacterial isolates with a zone of inhibition diameter < 22 mm were determined to be MRSA, while isolates with a zone of inhibition diameter ≥ 22 mm were recorded as MSSA.

Rapid DNA extraction method

DNA extraction was performed using the boiling method described previously [25]. Briefly, one to two colonies from overnight cultures on blood agar plates were suspended in 500 µl of sterile distilled water, and the suspension was heated at 100 °C for 15 min. After centrifugation for 5 min at 14,000 rpm to sediment the debris, 2 µl of the clear supernatant was used as template for PCR amplification.

Multiplex PCR detection of mecA and nuc genes

A multiplex PCR assay for the detection of S. aureus species-specific thermonuclease gene (nuc) and mecA gene was performed for all isolates. S. aureus SCCmec type IV strain (nuc +, mecA +, pvl −) was used as amplification control and pure water was used as negative control. Multiplex PCR assay included 2 µl of the DNA template added to a 25 µl final reaction mixture containing: 2X PCR Master Mix (Thermo Fischer Scientific, Waltham, MA USA) containing reaction buffer, Taq DNA polymerase (0.05 U/µL), 4 mM MgCl2, 0.4 mM of each dNTP and 10 pmol of each primer. The primer sets were used for amplification are shown in Additional file 1: Table S1 [26] and amplification was performed as described before [27] with a modification of an initial denaturation step at 94 °C for 10 min. PCR products were analyzed by agarose gel electrophoresis using 1.5% agarose gel. Gels were stained with ethidium bromide and visualized using MiniBIS Pro Gel Imaging System (DNR, Israel).

PCR detection of pvl gene

The presence of pvl gene was investigated for each sample using conventional single target PCR. 25 µl reaction mixture was prepared as described above using primer sets shown in Additional file 1: Table S1 [14]. S. aureus SCCmec type II strain (nuc +, mecA +, pvl +) was used as amplification control and pure water was used as negative control. PCR amplification was performed using conditions described before [28] with a modification of an initial denaturation step at 94 °C for 10 min. Product amplification was analyzed using gel electrophoresis.

Results

Fifty S. aureus strains isolated from different isolation sites from patients admitted to various hospital departments were investigated in this study. Within the patient group, 72% were inpatients (36/50) and 28% (14/50) were outpatients (Additional file 2: Figure S1). Of all the patients, 56% (28/50) were males and 44% (22/50) were females (Additional file 3: Figure S2). Clinical samples from patients were obtained from a range of departments were included in this study. The highest number of S. aureus cases were observed in Cardiology and Respiratory Medicine departments. Distribution of cases according to hospital departments is shown in Fig. 1a. Clinical samples were from collected from diverse patient body sites. Analysis of sample isolation sites indicated that 22% (n = 11) of the samples were taken from abscess/wound, 18% (n = 9) deep tracheal aspirate (DTA), 18% (n = 9) nasal swab, 16% (n = 8) of the samples were from blood, 10% (n = 5) urine, 4% (n = 2) sputum, 4% (n = 2) catheter tip, 4% (n = 2) body fluids, 2% (n = 1) bronchioalveolar lavage (BAL) and 2% (n = 1) were from urethral origin. Data suggested that the highest number of S. aureus cases were observed in samples isolated from abscess/wound, followed by DTA and nasal swabs. Analysis of different isolation sites for all samples is shown in Fig. 1b.

Fig. 1
figure 1

Distribution of Staphylococcus aureus infections a according to hospital departments b according to body sites

Full size image

Out of 50 S. aureus isolates, previously identified as methicillin-resistant by the BD Phoenix 100 automated bacterial identification system, 3 isolates were detected to be susceptible to methicillin based on cefoxitin disc diffusion assay and were confirmed as MSSA. Among the 50 isolates, 100% (50/50) were nuc positive by PCR. mecA gene was detected in 68% (34/50) of the isolates as confirmed with the multiplex PCR assay (Fig. 2a). Isolates which were detected as cefoxitin resistant by disc diffusion assay (n = 47) but did not carry mecA gene required further screening to identify alternative mechanisms of resistance by mecB and mecC PCR. Overall, 13 isolates out of 47 were positive for pvl gene, and the prevalence of pvl in confirmed MRSA isolates was 27.7% (13/47) (Fig. 2b). Two PVL-positive isolates which were cefoxitin susceptible by disc diffusion assay and negative for the presence of mecA were excluded from prevalence analysis. The clinical and molecular characteristics of the PVL-positive isolates are summarized in Table 1.

Fig. 2
figure 2

PCR detection of nuc, mecA and pvl genes a multiplex PCR for simultaneous detection of nuc and mecA in MRSA isolates. Lanes 1 and 3: nuc, mecA positive, Lane 2: nuc positive, PC: positive control, NC: negative control, M: 100 bp DNA ladder (Hibrigen); b detection of pvl in MRSA isolates by single target PCR. Lanes 1–4, 6–8, 10: pvl positive, Lanes 5, 9: pvl negative, PC: positive control, NC: negative control, M: 100 bp DNA ladder (Hibrigen)

Full size image
Table 1 Characteristics of Panton-Valentine leukocidin-positive isolates of Staphylococcus aureus
Full size table

Among the PVL-positive MRSA isolates 69.2% (9/13) were inpatients. PVL positivity was more common in MRSA isolated from DTA (30.8%, 4/13) and abscess/wound (23.1%, 3/13). The PVL-positive isolates were from nine ward areas, but were mainly from the Respiratory Medicine, Cardiology and Dermatology departments, as well as the Intensive Care Unit (ICU).

Discussion

Spread of PVL-producing MRSA from the community into healthcare setting poses a great public health risk and may result in outbreaks affecting vulnerable populations such as neonatal and intensive care units. Screening of virulence characteristics of patient-derived isolates is therefore essential for timely identification of patients carrying multi-drug resistant and virulent bacterial strains that would enable their isolation in hospitals. This pilot study reports novel findings in relation to MRSA virulence traits in a hospital setting in NC, in which the prevalence of PVL-positive MRSA was determined to be 27.7%.

Epidemiological characteristics of bacterial strains vary among countries, cities, hospitals as well as different wards within a hospital. So far, multiple studies have reported the presence of PVL in MRSA strains circulating in the hospital setting. In Turkey, PVL-positivity in MRSA isolates in hospitals was reported to range between 1.7 and 20% [29,30,31,32]. In a 2010 study in England and Wales, programme of enhanced surveillance of PVL-MRSA indicated an increased trend in clinical specimens [33], with reports of PVL-positive MRSA outbreak in a regional neonatal unit in the UK in 2012 [34]. Large-scale analysis of pediatric patients with invasive S. aureus infections recruited from 13 centers in 7 European countries has revealed that overall PVL positivity rate in MRSA isolates in 2016 was 7.8% [21]. On the other hand, reports show that the prevalence of pvl gene among MRSA strains in a Japanese hospital in 2019 was 13.5% [35], whereas PVL-positivity was higher at a rate of 39% among MRSA isolates collected from pediatric patients in China within the same year [36]. Epidemiological surveillance of MRSA strains collected between 2012 and 2017 in Australia showed that 28% of the isolates were PVL positive [37]. In an attempt to investigate MRSA isolates in developing countries, Shrestha et al. screened MRSA isolates at a tertiary hospital in Nepal, and indicated pvl gene positivity rate in nosocomial isolates to be 26.1% [38]. In another study, Singh-Moodley et al. have characterized MRSA isolates from blood cultures from 2013 to 2016 in South Africa, and reported pvl positivity rate to be 25% [39]. PVL-MRSA has been also linked to nosocomial infections in which hospital transmission of CA-MRSA had occurred in the US [40], highlighting the important source of CA-MRSA as a public health threat in hospitals. In low MRSA prevalence settings, healthcare workers may also serve as a reservoir of MRSA and an important potential source of transmission to patients as demonstrated by several studies [41, 42].

Several EU-EEA countries are still reporting high levels of MRSA. Between 2003 and 2005, Cyprus was reported to have highest proportion of MRSA among many countries in the Mediterranean region with a prevalence of 64%, while the prevalence of invasive MRSA cases between 2015 and 2018 were reported to range between 25 and < 50% [43], indicative of a high burden to the healthcare system. There is a shortage of data in terms of molecular characteristics of clinical S. aureus isolates in Cyprus. Our findings show, for the first time, that 27.7% of MRSA strains isolated from patients in a university hospital in NC were pvl positive. Continued surveillance and characterization of MRSA isolates in hospitals in the country is imperative for the prevention of spread of virulent nosocomial infections and the implementation of enhanced infection control strategies. Further in-depth molecular typing of clinical MRSA isolates should also be pursued to identify MRSA SCCmec types circulating both in the community and hospital setting.

Limitations

This was a pilot study with a small sample size of 50 isolates from a single center, and therefore does not represent an overall prevalence of PVL-MRSA in hospitals in Cyprus. Studies with a larger number of samples collected from patients admitted to different hospitals across the country are needed to determine the overall prevalence and distribution of virulent strains. Another major limitation of the study was the lack of genotyping to demonstrate SCCmec, Multi-Locus Sequence Typing (MLST), protein A (spa) and accessory gene regulator (agr) types of isolates and further analysis of additional virulence factors associated with MRSA infection.