Antibiotic resistance pattern and spa types of Staphylococcus aureus strains isolated from food business and hospital kitchen employees in Çanakkale, Turkey

Background/aim The aim of this study was to determine antibiotic resistance profiles and spa types of Staphylococcus aureus strains isolated from food business employees in Çanakkale, Turkey. Materials and methods S. aureus isolates were collected from hand and nasal swabs of 300 individuals working in 17 food businesses and 9 hospital kitchens. All S. aureus isolates obtained from each carrier were typed by staphylococcal protein A (spa) typing method.Staphylococcal cassette chromosome mec (SCCmec) and multilocus sequence typing (MLST) of MRSA were performed by sequencing method. Results Of the 300 individuals, 125 (41.6%) were found to be carriers of S. aureus, 215 isolates of which were obtained in total. Three (1.4%) of 215 isolates were identified as MRSA. Sixty spa types were identified among the 121 MSSA isolates, the most common being t084 (9%). A novel spa type was discerned and added to the Ridom SpaServer database as t14963. The MLST type of the MRSA strains identified as spa type t786 was ST88 and as spa type t223 was ST22. All MRSA were determined to be SCCmec type IVa. Conclusion spa typing can be performed to screen for transmission of S. aureus. t786, ST88, and SCCmec IVa MRSA strains were identified for the first time in Turkey.


Introduction
Staphylococcus aureus is an important organism in public health, as a cause of nosocomial infections and food poisoning incidents. Owing to its potential resistance to multiple antibiotics and production of virulence factors, S. aureus represents an important source of both nosocomial and community-acquired infections (1,2). Molecular typing has become increasingly important both for controlling the spread of S. aureus and determining the origins of epidemics associated with virulence factors (3). Staphylococcal protein A (spa) typing, which relies only on the assessment of repeats at the x region of spa, exhibits excellent discriminatory power and shares with multilocus sequence typing (MLST) the advantages of unambiguous typing results that can be compared between laboratories and over time. spa typing is also both easier and less costly to perform than MLST or PFGE (4,5). The x region exhibits high variability and may be useful for the rapid typing of MRSA, particularly in hospital environments. spa typing has been found to be a fast and practical method for identifying the causative organisms of epidemics, since it involves a single locus (5,6). A strong correlation has been demonstrated between the results of other techniques and the clonal groups obtained with spa typing. The importance of this method in epidemiological studies and the determination of clonal associations is gradually increasing (7,8).
Software such as Ridom StaphType enables the description of spa repeats and types, using consistent spa code terminology (6,9). Alternative free programs have also been developed to evaluate spa region sequencing data. The Java-based program DNAGear, developed on the NetBeans platform, is one example (5).
The aim of the present study was to determine the antibiotic resistance profiles and spa types of S. aureus strains isolated from catering employees in hospitals and food businesses (restaurants and catering industries) located in Çanakkale, Turkey, and to determine transmission events between hospitals and food businesses.

Sample collection and strain isolation
Between April and December 2014, three samples were collected from each of 300 food business employees (chefs, sous-chefs, waiting staff, and kitchen cleaners) by swabbing the right and left hands (bare fingers, thumbs, and palms) and nasal mucosa. Individuals were a random sample of workers in their institutions. Of these, 228 were employed by food businesses (restaurants and catering industries), and 72 worked in hospitals in the city center and districts of Çanakkale. The swabs were used to inoculate 5 mL of sterile Brain Heart Infusion Broth (Oxoid). Inocula were then streaked onto Baird-Parker Agar medium (Merck) containing egg yolk emulsion. After incubation at 37 °C for 24 h, the purity of the resulting gray-black colonies surrounded by a 1-1.5 mm clear zone was verified before their use in further analyses. S. aureus isolates were identified morphologically by Gram staining and catalase and coagulase testing. The study was conducted using a protocol approved by the Local Ethics Committee of Clinical Research (Decision No. 050.99-214), and written informed consent was obtained from all participants.

DNA extraction
Several colonies obtained from a fresh culture were mixed thoroughly in tubes containing 50 µL of buffer solution from a stock comprising 925 µL of H 2 O, 25 µL of 10% SDS (Sigma), and 50 µL of 2 M NaOH (Merck). The tubes were then incubated at 100 °C for 10 min in a TDB-120 dry heating block (Biosan) to lyse cells. Tris-EDTA buffer (50 µL; Sigma) was subsequently added and the tubes were centrifuged at 13,000 rpm for 10 min. The supernatant (40 µL) was collected from each centrifuged sample and added to 150 µL of Tris-EDTA buffer.

spa typing
The polymorphic x region was amplified as previously described using primers spa 1113F (5′-TAAAGACGATCCTTCGGTGAGC-3′) and 1514R (5′-CAGCAGTAGTGCCGTTTGCTT-3′) ( Figure). for each sample. Isolate spa types were determined with the program DNAGear, which retrieves the updated spa type and repeat data automatically from the Ridom SpaServer database (5). The discriminatory power of the spa typing method with which the present data were obtained was calculated using an online application designed according to the formula proposed by Hunter (http://insilico.ehu.es/ mini_tools/discriminatory_power) (12).

SCCmec and MLST typing
SCCmec types were determined by PCR analysis of the cassette chromosome recombinase (ccr) and mec gene complexes, as previously described (13,14). MLST analysis was performed by amplifying 7 gene regions for each clone as described online (http://saureus.mlst.net/misc/info. asp) (15).

Statistical analysis
Descriptive statistical analysis of the study data was performed with SPSS 19.0 (IBM Corp.).

Strain isolation and antimicrobial susceptibility
S. aureus bacteria were isolated from both right and left hands and the nasal mucosa of 29 employees (n = 87). Additionally, these strains were isolated from the nasal mucosa and the right hands of 9 people (n = 18), the nasal mucosa and the left hands of 11 people (n = 22), the right and left hands of 12 people (n = 24), the right hands of 15 people (n = 15), the left hands of 8 people (n = 8), and the nasal mucosa of 41 (n = 41) people. Of the 300 participants in this study, 125 (41.6%) were found to be carriers of S. aureus, 215 isolates of which were obtained in total. Of the 215 S. aureus isolates, 90 were isolated from the nasal mucosa, 65 were from right hands, and 60 were from left hands. Antibiotic resistance among these isolates was as follows: 19 (8.8%) were resistant to erythromycin, 15 (6.9%) to tetracycline, and 3 (1.4%) to cefoxitin. The cefoxitin-resistant isolates were obtained from nasal mucosa. These isolates were found to be susceptible to other antibiotics tested.

Determination of methicillin resistance by PCR
Three MRSA strains were isolated from the nasal mucosa of 3 food employees. Two individuals from whom MRSA strains were isolated were employees of food businesses and one worked in a hospital kitchen. The MRSA carrier rate among the participants of this study was 1%.

spa typing of S. aureus isolates
An isolate was taken from each carrier for spa typing. If one individual carried S. aureus in more than one body region, the isolates isolated from the nasal mucosa, right hand, and left hand were selected for genotyping, respectively. Therefore, only one isolate of every person was taken for genotyping studies. Of the 125 S. aureus strains subjected to spa typing, 90 (72%) were isolated from nasal mucosa, 27 (21.6%) from right hands, and 8 (6.4%) left hands. Thirtytwo (25.6%) isolates were obtained from hospital kitchen workers and 93 (74.4%) from food business employees. The spa gene region of one isolate could not be amplified with the primers used in MSSA isolates.
Sixty spa types were identified among the 121 MSSA isolates, of which 11 (9.0%) were t084, 9 (7.4%) t008, 8 (6.6%) t005, 6 (4.9%) each of t012 and t267, 5 (4.1%) t091, 4 (3.3%) each of t010 and t571, and 3 (2.5%) each of t223, t189, t310, t346, and t774. Furthermore, types t065, t158, t160, t304, t537, and t3266 were detected in two isolates each. Finally, 41 spa types were represented by single isolate. The distribution of spa types according to food businesses and hospitals is given in Table 1. The distribution of 121 MSSA isolates between food businesses (restaurants and catering industries) and hospitals according to spa types is shown in Tables 2 and 3. The spa types of the MRSA (n = 3) isolates obtained from food business employees were t786 and t223, and that taken from the hospital kitchen worker was of type t223. A total of 5 isolates were identified as spa type t223. Two of these were MRSA and 3 were MSSA.

SCCmec and MLST typing of MRSA isolates
SCCmec analysis identified the 3 MRSA isolates as type IVa, suggesting that these strains were communityacquired MRSA (CA-MRSA). The MLST type of the MRSA strains identified as spa type t786 was ST88 and as spa type t223 was ST22 (Table 4).

Discussion
Recent studies have revealed that methicillin resistance is gradually increasing among CA-MRSA. A previous examination of 526 S. aureus isolates obtained from individuals employed in food businesses revealed 28 (5.3%) to be MRSA in a study performed in Antalya, Turkey, in 2013 (16). A similar investigation identified methicillin resistance in 17.1% of 169 S. aureus isolates recovered from nasal swabs of food employees in Bursa, Turkey, in 2010 (17). Another study of the carriage of this bacterium by individuals working in comparable roles reported that 0.4% of 73 isolates were MRSA in Alanya, Turkey (18). The prevalence of S. aureus methicillin resistance in our study was 1.4%, and 1% of the food employees tested carried MRSA. These values are lower than those reported by the majority of other investigations.
In the largest epidemiological survey ever done in Turkey to date, 397 MRSA strains collected from 12hospitals in 11 cities in different geographical regions of Turkey between 2006 and 2008 and 91 MRSA strains collected from 4 hospitals in 2011 were investigated in terms of spa types and clonalities according to MLST and PFGE methods. This study showed the presence of a hospital Turkish clone TR09 (ST239-SCCmecIII-t030) and a community clone TR10 (ST737-SCCmecIV-t005) largely disseminated in Turkey. The most common MRSA clone in Turkey was found to be ST239 (91.4%)-SCCmec type III (91.4%), t030 (85.1%) (15). In a previous report, 54 MRSA strains obtained from 8 university hospitals, geographically distributed over the 6 main regions of Turkey, were shown to be of 4 spa types (t030, t459, t1459, and t189). t030 predominated, accounting for 89% of the isolates. All of them were ST239, SCCmec III (19). It was determined that all of the 48 MRSA isolates identified in the Eastern Black Sea Region except for 2 spa types were t030. The MLST types of strains representing the 3 different spa types were ST239 (3). In another study conducted on 9 MRSA strains isolated from blood cultures in İzmir, Turkey, it was shown  that the dominant strain was t030 spa and it was a member of the clone ST239-IIIA (20). The most common spa and MLST types of 12 MRSA strains isolated from mastitic cow milk and nasal swabs of related agricultural workers in the Aydın region were t030, ST239. Ten of them were identified as SCCmec-III (21). In the latest work reported from Turkey, clinical MRSA isolates were collected from 6 university hospitals. Of the 270 MRSA isolates, 91% were found as SCCmec III and 81.1% as t030. The majority of MRSA isolates reported in Turkey belonged to ST239, t30, SCCmec III (22). We identified 2 of 3 MRSA strains as t223, ST22, SCCmec IVa and one as t786, ST88, SCCmec IVa. spa type t223, which was previously determined in Aydın, has also been determined in our work (15). ST22  was detected in the genotyping study of 102 MRSA strains at a university hospital in İstanbul (23). spa type t223 MRSA isolates have been reported in many studies. In Jordan, 56 MRSA strains obtained from nasal swabs were identified as spa type t223 (39%). The majority of MRSA isolates harbored SCCmec IVa (75%) (24). spa type t223 was found to be the second most common type in Jordanian adults (25). The ST22-MRSA-IVa clone (spa t223) was dominant in the Gaza Strip and accounted for 64% of the MRSA in healthy children and their parents (26). In the Middle East, a study was conducted to investigate the molecular epidemiology of S. aureus outpatients attending primary healthcare centers in Egypt and Saudi Arabia. Only 2 spa types, t008 and t223, coexisted in both countries (27). In the Russian Federation, 9 of 13 CA-MRSA isolates belonged to ST22 (spa type t223, SCCmec type IVa) and were similar to the EMRSA-15/ Middle Eastern variant (Gaza strain) (28). In Italy, 8 of 10 MRSA isolates obtained from nasal swabs of healthy children were ST22-MRSA-IVa, spa t223 (29). t223, one of the 10 most frequent spa types, originated from Sweden, Scandinavia, the rest of Europe, the Middle East, and Africa (30). spa type t786 MRSA has also been reported in a number of countries. A total of 1462 inpatients and healthcare workers were screened for MRSA nasal carriage in Sao Tome and Principe and Angola. The majority of the isolates (51.7%) belonged to spa types t186/t325/t786/ t1814/t1951, ST88, and SCCmec type IVa, and such isolates were identified in both African countries (31). t786 (2) was reported in 20 MRSA isolates obtained from nasal swabs of outpatients attending primary healthcare centers in Egypt (27). In addition, t786 was detected in 381 MRSA strains obtained from patients admitted to 36 German hospitals (32). spa type t786 was also found in the Netherlands and Switzerland (33,34). In the majority of reported studies, it was found that the isolates of t786 were ST88, SCCmec type IVa. This strain type (t786, ST88) that we found has not been reported previously in Turkey.
In the current work, the same spa type (t223) was identified in the MRSA strains obtained from a food business employee and a hospital worker. MRSA may spread through a community via food business workers, and dissemination of a strain may occur from the wider community to a hospital, as well as in the opposite direction. Moreover, there is no relation between these MRSA carriers identified with the same spa type (t223) and they may have been colonized by different sources.
Over a time span of 6 months (May to October 2006), a total of 1490 staphylococcal isolates were sent to the National Reference Center in Germany for staphylococci for further characterization and typing. Among the 283 MSSA isolations, the most common spa types were t008 (14.5%) and t091 (4.9%) (7). In Hong Kong, food handlers from 6 catering establishments were nasally sampled for S. aureus. Of 434 food handlers, 99 (22.8%) were colonized with S. aureus. Forty-one spa types were determined within 94 MSSA isolates. spa typing revealed 18% of isolates as t189, 8.5% as t127, and 7.4% as t091 (35). A randomly selected, healthy Norwegian community population was studied for nasal carriage of S. aureus to describe frequency and molecular diversity. No MRSA isolates were found. The most frequent spa type t084 was detected in 7.4% (7/95) of the individuals. In our study, the most common spa type was t084 (36). In our study, the most common spa types of 121 MSSA isolates were t084 (9%), t008 (7.4%), and t005 (6.6%).
All MSSA isolates (n = 3) obtained from 3 individuals working in one of the district hospitals (H8) were of the same spa type (t223). Moreover, all MSSA isolates (n = 4) obtained from 4 individuals working in one of the hospitals in the province center (H9) were of the same spa type (t010). Here, MSSA isolates obtained from 2 hospitals were of the same spa type (t223 at H8 and t010 at H9), which was considered significant. This shows that the same clone may have spread to other people through a single person. Furthermore, 4 of the 7 isolates obtained from hospital H6 also shared the same spa type (t571). Interestingly, spa type t571, which we identified in 4 MSSA isolates, has been frequently found in many parts of the world among ST398 MSSA isolates, including isolates causing severe human infections. The ST398 genetic background is often associated with livestock-associated MRSA (37).
Here, 60 different spa types were detected among the MSSA isolates circulating in Çanakkale, Turkey. This draws attention to the fact that spa typing is also of use for MSSA strains. spa typing of MSSA isolates may contribute to identifying the source of any future local epidemics.
In conclusion, we have shown that rates of antibiotic resistance among S. aureus isolates obtained from food business employees and the carriage of MRSA by such individuals are low. This study contributed to the epidemiological data because of the detection t786, ST88, and SCCmec IVa MRSA strains, which have not been previously reported in Turkey.