Nasal Carriage by Staphylococcus aureus among Healthcare Workers and Students Attending a University Hospital in Southern Brazil: Prevalence, Phenotypic, and Molecular Characteristics

Background Staphylococcus aureus can asymptomatically colonize the human anterior nares and skin, and nasal colonization by this bacterium represents a potential risk for development of invasive infections. The aim of this study was to determine the prevalence of S. aureus nasal carriage among healthcare workers and students attending a university hospital and to characterize the isolates phenotypically and molecularly. Methods A cross-sectional study was performed with 324 volunteers. Cultures from nasal samples were obtained and S. aureus isolates were characterized according to their antimicrobial susceptibility profile and four virulence factors-encoding genes. MRSA isolates were characterized regarding their oxacillin/cefoxitin susceptibility, SCCmec, and REP-PCR types. Potential risks for S. aureus and MRSA carriage were analyzed. Results Of 324 nasal samples, 42.9% were identified as S. aureus, of which 28.8% were MRSA. S. aureus carriers were significantly higher in males and students (OR = 2.898, 95%CI 1.553–5.410); however, no variables were associated with MRSA carriage. All isolates were susceptible to vancomycin and the highest rate of resistance was observed for penicillin (90.6%). All isolates harbored the coa gene, and 97.8%, the icaA gene; 15.8% and 6.5% were positive for tst and lukS-PV/lukF-PV genes, respectively. Among MRSA isolates, 45% carried the mecA gene but were phenotypically susceptible to oxacillin/cefoxitin; two harbored the tst and none had lukS-PV/lukF-PV genes. All MRSAs were distributed into six SCCmec types and type I (62.5%) was the most frequent. REP-PCR typing identified four main clusters among MRSA isolates. Conclusion High prevalence of healthcare workers and students were identified as nasal carriers of S. aureus exhibiting different antimicrobial resistance profiles, including mecA-positive oxacillin-susceptible S. aureus (OS-MRSA) and the presence of virulence-encoding genes. Both cohorts may represent potential sources for the emergence of a successful S. aureus strain highly adapted to the hospital environment.


Introduction
Staphylococcus aureus can interact with its host as a commensal member of the microbiota [1] or as an opportunistic pathogen causing a wide range of community and hospitalassociated infections [2][3][4][5][6][7]. In humans, this bacterium can colonize multiple body sites, but the nose (anterior nares) is the frequent ecological niche of S. aureus carriage. Indeed, around 20-30% of the human population can harbor persistently and asymptomatically this bacterium in this niche [1]. Importantly, persistent nasal colonization by S. aureus may increase the risk for subsequent infections and this scenario is even more complicated in hospitalized and immunocompromised individuals who can develop invasive infections with high morbidity and mortality rates [8,9].
Nasal colonization by S. aureus may begin within the first days of human life and the horizontal transmission from colonized mother seems to be the major source of S. aureus carriage in newborns [10]. After birth, hands are the main vector for transmitting S. aureus from the surfaces to the nose [11,12]. Particularly in hospitals, healthcare workers and patients who are nasal carriers may be the source for the transmission and spread of S. aureus in these settings [13][14][15][16][17][18]. Similarly, students who have continuous hospital exposure during their internship may represent an important source for transmitting S. aureus within the hospital environments [14,19]. Crucially, high rates of colonization by methicillin-resistant S. aureus (MRSA) have been reported among healthcare workers [16,17], and increasing rates of MRSA colonization among students with continuous hospital exposure have also been reported [19]. Besides their role in S. aureus transmission in hospital settings, healthcare workers and students may also serve as the vector for cross-transmission of MRSA between the community and hospital populations [14,18].
Altogether these data indicate the importance of screening S. aureus nasal carriers in both healthcare workers and students as a component of the control of S. aureus infections in hospitals. In this regard, few studies on the prevalence of S. aureus colonization in healthcare workers and students attending the Brazilian hospital environments were reported in the literature.
e University Hospital of Universidade Estadual de Londrina (UH-UEL) is a public teaching hospital and the reference in high complexity cases in the north of Paraná, Brazil. e hospital serves as an internship field for undergraduate students of Medicine, Pharmacy, Nursing, and Physiotherapy and offers postgraduate programs in different biomedical skills. In a retrospective study performed during 2001-2015, MRSA accounted for 43.7% of bacteremia in hospitalized patients [5]. However, no study on nasal colonization by S. aureus of healthcare workers and students attending the hospital was performed so far. erefore, the aims of this study were to estimate the prevalence of and the potential risk for S. aureus nasal carriage among healthcare workers and students of the UH-UEL and to characterize the isolates by phenotypic and molecular methods.

Study Design.
A cross-sectional study was carried out with the community of the UH-UEL from December 2017 to May 2018. e sample size was calculated using the software EPI Info ™ 7 (Atlanta, USA) and was based on the population (students and professionals) attending the UH-UEL, which amounts about 2000 individuals during the period of the study. Assuming a prevalence of S. aureus colonization of about 40% [20] and a confidence interval of 95%, at least 324 individuals were included in this study. e collection of data occurred at the time of nasal swab sampling through a standardized questionnaire that contained demographic variables, antibiotic use and/or hospitalization in the last three months, and direct contact with hospitalized patients. e study protocol was approved by the Ethics Committee of the UEL (CAAE 79663417.2.0000.5231-opinion number 2.421.361-CEP-UEL).

Collection and
Processing of Nasal Swabs. One nasal sample was collected from each participant using the Stuart collection device (COPAN Diagnostic, Italy). e swab was introduced into a nostril and gently rotated three times, transported to the laboratory, and processed up to 2 h after collection. Nasal samples were inoculated into Tryptone Soya Broth (TSB, Oxoid, Brazil) supplemented with 6.5% sodium chloride and incubated at 35°C for 24 h. After, the sample was subcultured on Mannitol Salt Agar (Oxoid, Brazil) at 35°C for 24 h. Suggestive colonies of staphylococci were subjected to standard phenotypic identification [21]. Species identification was confirmed by a multiplex-PCR targeting the nuc gene (encoding thermonuclease) [22]. Bacteria were stored at −80°C in TSB containing 30% glycerol.

Antimicrobial Susceptibility Testing.
e antibacterial susceptibility profiles for 12 antimicrobials were evaluated by the disk diffusion method. e minimal inhibitory concentration (MIC) for oxacillin was determined by the broth microdilution assay. e inducible clindamycin resistance was determined by the double disk diffusion method. All methods were performed as recommended by the Clinical and Laboratory Standards Institute [23]. e susceptibility breakpoints were those recommended by CLSI [23] except tigecycline that was interpreted according to the European Committee on Antimicrobial Susceptibility Testing [24]. Enterococcus faecalis ATCC 29212 and S. aureus ATCC 25923 were used as quality controls. Cefoxitin disk and oxacillin MIC were used to define MRSA.

DNA Extraction.
A single colony of each bacterium was cultured in 3 mL TSB at 35°C for 24 h. e cells were harvested by centrifugation (10,000 × g for 5 min), washed once with sterile 0.15 M phosphate-buffered saline pH 7.2, and resuspended in 300 μL sterile solution (10 mM Tris-HCl, 1 mM EDTA pH8.0, and 1.0 mg/mL lysozyme). Genomic DNA was extracted as described previously [26], and 2 μL were used in all amplification reactions.
e genetic relatedness of all MRSA was analyzed by repetitive element sequence based-PCR (REP-PCR) [30]. Fingerprintings that had more than one band differing in size were considered different REP-PCR types [31]. Banding patterns were categorized using the UPGMA algorithm and Jaccard coefficient [32] of the Bionumerics v.6.5 software (Applied Mathematics, Kortrijk, Belgium), with the band tolerance set at 3% and the threshold cutoff value set at 85% [3].

Statistical Analyses.
Analyses of contingency tables (χ 2 test) were employed to check the associations between categorical variables and diagnostic groups.
e Kolmogorov-Smirnov test was used to assess the normality of distribution of age. Categorical variables were expressed as absolute number (n) and percentage (%) and continuous variables were expressed as mean ± standard deviation. e association between explanatory variables and colonization was evaluated using three different models of automatic stepwise binary logistic regression analysis controlled for covariates that may confound the association of interest. All statistical analyses were performed using IBM SPSS windows version 20. Tests were 2-tailed and an alpha level of 0.05 indicated statistically significant results. Although one single nasal sampling was performed on each participant, which does not allow the identification of persistent carriers [33], a high prevalence of S. aureus carriers was detected in this study. e overall prevalence of S. aureus nasal carriage was 42.9% (139/324); among them, 28.8% (40/139) were identified as MRSA carriers; that is, the isolate harbored the mecA gene, representing 12.3% of all participants ( Figure 1).

S. aureus Nasal Carriage Was Significantly Higher in
Males and Students. Colonization by S. aureus depends on bacterial ability to survive and adapt in the host's niches. ese include overcoming innate and adaptive defenses of the host along with coexisting with other members of the microbiota that may promote or inhibit its growth [1]. Moreover, several factors including younger age, being male, contact with healthcare workers, professional occupation, and geographical region have been correlated with S. aureus nasal colonization [13,[34][35][36].
In this study, we analyzed six descriptive characteristics of the participants (gender, age, occupation, antibiotic use and/or hospitalization in the last three months, and direct contact with hospitalized patients) and their potential risk for S. aureus nasal carriage (Table 1). In order to evaluate the independence of associations, we have carried out three different automatic stepwise logistic regression analyses. In model #1 we added gender and occupation, both variables were positively associated with colonization with S. aureus (p � 0.004 and p � 0.019, respectively) (χ 2 � 12.337, df � 2, p � 0.002). e second model (#2) is the model #1 added to the other explanatory variables evaluated in this study. Gender (male) remained associated with colonization (p � 0.003), whereas age was negatively associated with colonization (p � 0.021) (χ2 � 12.915, df � 2, p � 0.002). Finally, in model #3, when evaluating the interaction between gender (male) and occupation (student), only students of male gender were independently associated with colonization by S. aureus (p � 0.001) (χ2 � 1.733, df � 1, p � 0.001). Furthermore, model #3 contributed to correctly assessing 61.8% of the cases, with high sensitivity (90.2%); however, it presented low specificity (23.9%). e other explanatory variables were not associated with S. aureus nasal carriage (p > 0.05) ( Table 2). Accordingly, male participants had 2.107 (95% CI: 1.268-3.502; p � 0.004) times the chance of being a carrier of S. aureus compared to female participants. Additionally, students had 1.819 (95% CI: 1.105-2.996, p � 0.019) times the chance of being a carrier of S. aureus compared to healthcare workers ( Table 2). None of the analyzed variables were associated with MRSA nasal carriage (Table 1).
Few studies analyzing together both cohorts and healthcare workers and students attending a hospital are described in the literature. Hogan et al. [14] reported a low prevalence of S. aureus nasal colonization in healthcare workers (10.4%) and in nonmedical students (11.4%) of the University of Madagascar. In contrast to our results, the carriage was higher in female participants and increased with age. Hussein et al. [15] reported a prevalence of 22.5% and 18.7% of S. aureus nasal carriage in healthcare workers, and students of a preparatory school that had no contact with a hospital, respectively, in Iraq. None of the variables analyzed (gender, age, and working hours per day or year) were associated with S. aureus colonization. Finally, El-Mahdy et al. [18] reported a prevalence of 26.0% of S. aureus nasal carriers among both healthcare workers and clinical students in Saudi Arabia. ese authors did not evaluate the risk factors associated with colonization.
In Brazil [13]. In another study, a prevalence of 21.7% (30/138) was observed in nursing students of the Federal University of Piauí, Piauí, and, among the carriers, 23.3% (7/30) were colonized with MRSA [41]. Similar to our results, a high prevalence (102/250, 40.8%) of colonization by S. aureus was found among undergraduate biomedical students of the State University of Maringá, in the northwest of Paraná. MRSA was identified in 5.9% (6/102) of colonized students [20].

S. aureus Isolated from Healthcare Workers and Students
Exhibits Different Antimicrobial Susceptibility Patterns. In our hospital, vancomycin and linezolid are the most used antimicrobials for the treatment of MRSA infections, whereas the others analyzed in this study are used in MSSA infections. e majority of S. aureus isolates exhibited susceptibility to most antimicrobials, and this result is consistent with studies reported previously [14], including in Brazil [13,20,41]. According to the phenotypic results, six (4.3%) out of 139 isolates were susceptible to all antibacterial agents tested in this study, and all were susceptible to vancomycin. e majority of the isolates were also susceptible to chloramphenicol (135/139,  However, the rates of clindamycin/erythromycin resistance may vary according to the geographical region. Low resistance rates were observed in studies from Colombia [39] and Italy [40]. Conversely and consistent with our findings, high rates of resistance to both antimicrobials were detected in Iran [38], China [36], and Brazil [20,41]. Finally, in some countries, high resistance rates were detected only for erythromycin, including Saudi Arabia [18] and Ethiopia [17]. Herein, S. aureus isolates were classified into 38 groups according to the antimicrobial resistance profile (antibiotype) ( Table 4). Although a high number of antibiotypes was detected, the majority consisted of a single isolate, and the most frequent profile was coresistance to penicillin, erythromycin, and clindamycin (43/139, 30.9%), followed by resistance to penicillin (30/139, 21.6%). Importantly, 24.5% (34/139) of these isolates were classified as multidrug-resistant, that is, being resistant to three or more antimicrobial classes [42].
Another important finding of our study is that a high proportion of MRSA isolates (18/40, 45.0%) was not detected by phenotypic methods used for identifying methicillin resistance. Indeed, MRSA isolates (n � 40) exhibited different patterns of susceptibility to cefoxitin/oxacillin:  (Figures 1 and 2).
Currently, OS-MRSA isolated from both infection [7,43,44] and nasal colonization [18] has been increasingly reported worldwide. e presence of OS-MRSA carriers and the potential risk of its transmission within the hospital and community are of concern.
is bacterium can be misidentified as MSSA by phenotypic methods routinely used in  Table 5).
Not surprisingly, all S. aureus harbored the coa gene that encodes coagulase, a secreted factor that induces intravascular coagulation, allowing bacterial aggregation in blood and hence promotes its survival and dissemination [45]. Moreover, coagulase activity has been used to differentiate between S. aureus (coagulase positive) and those staphylococci historically classified as less pathogenic (coagulase negative) [46]. e gene icaA is a component of the icaABDC operon that encodes the enzymes involved in the biosynthesis of polysaccharide intercellular adhesin (PIA). is adhesin promotes intercellular aggregation during the maturation stage of PIA-dependent biofilm formation by S. aureus [47] and is essential for bacterial virulence in murine model of systemic infection [48].
Notably, the genes lukS-PV/lukF-PV and tst were, respectively, detected exclusively and predominantly in MSSA isolates of our study, and, in general, these isolates exhibited resistance to fewer antimicrobials. Accordingly, the lukS-PV/lukF-PV-positive isolates displayed the antibiotypes 3, 6, 7, 12, and 14; two tst-positive isolates were susceptible to all antimicrobials and the others displayed antibiotypes 1, 3, e presence of PVL encoding genes is highly variable and can be related to S. aureus strain types. Indeed, PVL encoding genes have been primarily associated with community-acquired methicillin-resistant S. aureus (CA-MRSA) [49]. On the other hand, tst gene has been frequently associated with MSSA strains [50]. However, both genes have also been detected in S. aureus exhibiting different characteristics than those of these strains [3,6,14,39,49].
PVL, a bicomponent β-pore-forming toxin, targets the cell membranes of polymorphonuclear leukocytes, monocytes, and macrophages and causes tissue necrosis [51]. Invasive infections caused by PVL-positive S. aureus strains have been associated with poor outcome and high mortality rates, regardless of the methicillin-resistance [52,53]. Importantly, the PVL encoding genes are carried by lysogenic bacteriophages that can be induced by the treatment with various antibacterial agents, including ciprofloxacin, imipenem, trimethoprim and/or sulfamethoxazole, and tobramycin [49], thus, facilitating the horizontal transmission among different isolates [51]. Furthermore, it has been known that sub-MIC of beta-lactam antibiotics, which can be found within biofilms, enhances the production of PVL in vitro and in vivo [54]. Although the clinical significance of this effect is still unknown, these data highlight the importance of the  Antimicrobial susceptibility was determined by disk diffusion, except oxacillin and vancomycin that were determined by broth microdilution assay [23] and the agar-screen test [25], respectively. Disk diffusion and microdilution results were interpreted as recommended by CLSI [23]  continuous monitoring of the antimicrobial susceptibility and virulence potential of S. aureus for the proper management of the infections caused by this bacterium. TSST-1 is a potent exotoxin of the superantigens family that activates T lymphocytes, resulting in systemic inflammation and shock due to the overproduction of cytokines [55]. Infections caused by tst-positive S. aureus are also potentially fatal [53] and remain an important cause of morbidity among young people, especially in young women [50]. Although tst gene is located in a pathogenicity island, it can also be transferred among different isolates by a helper bacteriophage [56]. e limitation of this analysis is that the presence of virulence-encoding genes was evaluated only by PCR, which does not reflect the expression of the phenotype. However, nasal colonization by S. aureus harboring PVL and TSST-1 encoding genes in healthcare workers and students attending the hospital settings may represent a potential source for the emergence of a successful S. aureus strain highly adapted in hospital environment due to the silent spread of these virulence markers to multidrug-resistant strains and hence severe and difficult to treat infection in patients. Moreover, both populations may serve as agents of cross-transmission of this strain between the hospital and community individuals.

MRSA Isolates Harboring Different SCCmec and REP-PCR Types Are Grouped into Four Main
Clusters. MRSA isolates were distributed into six SCCmec types and the type I    (25/40, 62.5%) was the most frequent, followed by type IV (9/40, 22.5%). One isolate each (1/40, 2.5%) harbored the SCCmec types II, III, V, and VI; two isolates (2/40, 5.0%) were classified as NT (Figure 2). e SCCmec types identified herein were also detected in MRSA isolated from different staphylococcal infections in our hospital, where SCCmec type II was the most prevalent [3,6,7]. Visual observation of bands generated by REP-PCR typing showed a total of 21 banding patterns (named a to u), indicating a relatively high genetic diversity among MRSA isolates. REP-PCR banding patterns were automatically sized and all isolates with the same type were recognized to be identical. e level of similarity between the REP-PCR fingerprinting of the MRSA isolates ranged from 44.1 to 100%. Clusters analysis, by using a cutoff value of 85% [3], revealed that the majority of MRSA isolates were distributed into four groups (named as A to D clusters). Seven isolates exhibited unique fingerprinting patterns. Clusters A and D harbored two isolates each of the different REP-PCR and SCCmec types. Cluster B consisted of 17 isolates, comprising five REP-PCR (g, h, i, j, and k) and six SCCmec (I, II, IV, V, VI, and NT) different types, which were grouped in three subclusters (B1, B2, and B3). Finally, cluster C consisted of 12 isolates, comprising five REP-PCR (l, m, n, o, and p) and two SCCmec (I and IV) different types, which were grouped in two subclusters (C1 and C2) ( Figure 2).
ese results may explain the high prevalence of OS-MRSA nasal carriers in our study.

Conclusion
is study provided an overview of S. aureus nasal carriage among healthcare workers and students attending our hospital. We observed a high prevalence of S. aureus exhibiting different antimicrobial resistance profiles, including multidrug-resistant and OS-MRSA isolates; higher prevalence of MRSA harboring the SCCmec types I and IV, which is in contrast to our previous studies reporting the prevalence of SCCmec type II among MRSA isolated from different infections [3,6]; and the presence of PVL and TSST-1 encoding genes among these isolates. ese data alert us for continuous surveillance of potential reservoirs of S. aureus and improvement in the infections control measures for detecting the various methicillin susceptibility phenotypes and virulence potential of S. aureus.
Data Availability e data generated and analyzed in the current study are included in the article. Data are also available from the corresponding author upon reasonable request.