Antimicrobial susceptibility and resistance mechanisms of methicillin resistant Staphylococcus aureus isolated from 12 Hospitals in Turkey

Introduction Methicillin-resistant Staphylococcus aureus (MRSA) is one of the most important nosocomial pathogens and is also emerging in Turkish hospitals. The aim of this study was to determine the antimicrobial susceptibility profiles of MRSA isolated from Turkish hospitals. Materials and methods A total of 397 MRSA strains isolated from 12 hospitals in Turkey were included to present study. Antimicrobial susceptibilities were tested using agar dilution method. Presence of ermA, ermB, ermC, msrA, tetM, tetK, linA and aac-aph genes were studied by PCR. Results All strains were susceptible to vancomycin and linezolid. The susceptibility rates for fusidic acid, lincomycin, erythromycin, tetracyclin, gentamycin, kanamycin, and, ciprofloxacin were 91.9%, 41.1%, 27.2%, 11.8%, 8.5%, 8.3% and 6.8%, respectively. Lincomycin inactivation was positive for 3 isolates. Of 225 erythromycin resistant isolates 48 had ermA, 20 had ermC, and 128 had ermA-C. PCR was negative for 15 strains. Of 3 isolates with lincomycin inactivation one had linA and msrA. Of 358 gentamycin resistant isolates 334 had aac-aph and 24 were negatives. Among 350 tetracyclin resistant isolates 314 had tetM. Of 36 tetM negative isolates 10 had tetK. Conclusion MRSA isolates from Turkish hospitals were multiresistant to antimicrobials. Quinolone and gentamycin resistance levels were high and macrolide and lincosamide resistance were relatively low. Susceptibility rates for fusidic asid were high. Linezolide and vancomycin resistance are not emerged. The most common resistance genes were ermA, tetM and aac-aph. Evolution of antimicrobial susceptibilities and resistance genes profiles of MRSA isolates should be surveyed at regional and national level for accurate treatment of patients and to control dissemination of resistance genes.


Introduction
Staphylococci are important infection agents that cause hospital and community acquired infections. These bacteria have ability to adapt themselves to difficult conditions and successful clones have capacity of epidemic and pandemic dissemination [1]. Increasing resistance problem in staphylococci became an important public health problem. In 1944 when penicillin became available for use the susceptibility rate of Staphylococcus aureus to penicillin was >94% which became <5% recently [2]. Methicillin resistance appeared and started to disseminate from 1980 and became one of the major problem in hospital infections. Methicillin resistance is due to acquisition of a transpeptidase, PBP2a, involved in cell wall synthesis that has low affinity for beta lactam antibiotics which rends bacteria resistant to all beta lactam antibiotics. Treatment of infections due to methicillin resistant S. aureus (MRSA) causes problems due to restricted number of choices [1]. Especially from 2003, when vancomycin resistant S. aureus emerged it became urgent to search new treatment possibilities for these bacteria [3]. In addition emergence and dissemination of community MRSA isolates forced to evaluate empiric treatment options in consideration with changing resistance profiles of these bacteria. MRSA strains do not affect only human but also infect farm animals and pets [4]. Although development of new antibiotics reduced dramatically recently, some antibiotics like daptomycin, linezolid and tigecyclin could be commercialized lately [1].
In the present study susceptibilities of 397 MRSA isolated from 12 centers in Turkey to linezolid, fusidic acid, kanamycin, gentamycin, erythromycin, lincomycin, tetracyclin, vancomycin and ciprofloxacin were tested by agar dilution method and presence known resistance genes were verified by PCR using specific primers.

Determination of antimicrobial susceptibilities Agar dilution method
Antibiotics tested were linezolid, fusidic acid, kanamycin, gentamycin, erythromycin, lincomycin, tetracyclin, vancomycin and ciprofloxacin. Erythromycin and fusidic acid were from Koçak Farma (Tekirdağ, Türkiye), kanamycin, tetracyclin and vancomycin were purchased from Sigma, and commercial injectable preparations were used for the remaining antimicrobials. Agar dilution method was used as described previously [5]. Shortly plates were prepared with serial dilution from 64 or 128 mg/L antibiotic concentrations. Inoculum with 5X10 4 bacteria was placed onto agar using multipoint inoculator. After 16-20 h incubation at 37°C the lowest concentration that inhibits bacterial growth was accepted as MIC. Reference strain S. aureus RN4220 was included to each run.

Gots test
All lincomycin resistant isolates were tested by Gots' test for presence of resistance by antibiotic inactivation. For this purpose to 19 ml agar at 50-60°C 19 ml BHI (Brain Heart infusion) agar 0,5 mg/L clindamycin and 1 ml overnight broth of Micrococcus luteus ATCC9341 were added, mixed and poured to petri dish and left for solidification. The MRSA isolates were inoculated as small round onto agar. On one plate approximately 20 MRSA isolates were inoculated. After 24 h incubation at 37°C plates were left 24 h at room temperature. Growth of indicator bacteria in the round of tested bacteria was accepted as positive which showed presence of resistance mechanism by inactivation [6].
Determination of resistance mechanisms DNA extraction DNA extraction was done using Instagen Matrix (BioRad) as recommended by manufacturer. Shortly 1-2 colonies were homogenized in 1 ml of distillated water and centrifuged at 10000 rpm for 1 minute. Supernatant were discarded and pellet was homogenized with 100 μl of instagen matrix. After incubation at 55°C during 15-30 min the mixture was vortexed and incubated at 95°C during 8 min. Lysate were centrifuged and 2 μl of supernatant were used as DNA for PCR reactions.

Susceptibilities to antibiotics
MICs and resistance was evaluated using CLSI criteria [12]. All 397 MRSA isolates tested were found to be susceptible to vancomycin and linezolid. Only 8 of 397 MRSA isolates were susceptible to all antibiotics tested. In Table 2 Table 3). Distribution of resistance levels for the antibiotics for each centre is shown at Table 4.

Resistance mechanisms
Of 225 erythromycin resistant MRSA isolates 48 carried ermA, 20 carried ermC, 1 carried both ermA and ermB, 1 carried both ermB and ermC, 128 carried both ermA and ermC, 2 carried ermA, ermB and ermC, 2 carried msrA, 2 carried msrA and ermA, 1 had msrA and ermB, 4 had msrA, ermA and ermC, 1 had msrA and ermC genes. A total of 15 isolates were negatives for all erythromycin resistance genes tested. Among MRSA isolates 64 were intermediate resistant to erythromycin. Of these isolates 36 were positive for ermA, 1 isolate had both ermA and ermC, and 1 isolate was positive for msrA. All remaining 26 isolates were negatives for the genes tested. Among 168 lincomycin resistant isolates 9 had ermA, 17 had ermC, 1 had both ermA and ermB, 1 had both ermB and ermC, 124 had both ermA and ermC, 4 had msrA, ermA and ermC, 2 had ermA, ermB and ermC, 1 had linA and msrA, 1 had ermC and msrA genes found and 8 isolates were negative by PCR for the genes tested. All lincomycin resistant isolates were tested for clindamycin inactivation by Gots' test and 3 isolates were found to be positive for inactivation. Of these 3 isolates one carried linA gene responsible for lincosamide inactivation and also msrA gene, but remaining 2 isolates were negatives for both linA and linB genes.
A total of 358 isolates were resistant to gentamycin and 334 of these isolates were positive for aac-aph gene and remaining 24 isolates were negative for this gene.
It was found that 350 of 397 isolates were resistant to tetracyclin. Of these 350 isolates 314 carried tetM gene and 36 did not carry this gene. Among tetM negative 36 isolates 10 had tetK gene and remaining 26 isolates were negative both tetM and tetK. Distribution of resistance genes among resistant isolates are shown in Table 5. Among macrolide resistance isolates the most common gene combination was ermA-ermC. Among tetracyclin and gentamycin resistant isolates the most common resistant genes were tetM and aac-aph, respectively.

Discussion
Antibiotic resistance became an important public health problem in Turkey as it is in whole world. Restriction of beta lactam use in MRSA isolates required use of other types of antibiotics for the treatment of infections due to MRSA isolates so survey of susceptibilities of MRSA isolates for antibiotics other than beta lactams became very important.
Our study is the largest study done in Turkey which evaluates both phenotypic and genotypic aspect   [17]. Ardıç et al. also found among 28 erythromycin resistant MRSA that presence of both ermA and ermC was the most frequent genotype [18].
In our study the most frequent mechanism of macrolide resistance among MRSA isolates found to be presence of methylase. Presence of methylase may confer inducible lincomycin resistance which should be taken in consideration for treatment design. The dominant genes among tetracyclin and aminoglycoside resistant isolates were tetM and aac-aph, respectively.
The dissemination of resistance was also analysed at regional level. İsolates from Istanbul had lower tetracycline resistance than other regions. MRSA isolates from Van had higher macrolide resistance rates than other regions. Ciprofloxacin resistance rates were very high in all centers   [22]. A study from Iran, neighbour state of Turkey, showed that isolates from Tehran aac-aph gene was the most common gene among gentamycin resistant S. aureus (83%) [23]. Tetracyclin resistance gene tetM was 90% positive among tetracyclin resistant isolates which were only 49% among resistant isolates from Malesia [20].

Conclusion
Our study is one of the largest epidemiological study done in Turkey. These multi-centre data of resistance level and mechanism of resistance of MRSA isolates will be important for future surveillance studies to determine the evolution of resistance levels and mechanisms at national and regional level. Also our results and follow up studies may constitute a database for empirical treatment of infections due to MRSA. Our multicentre study showed that isolates from 12 centres from Turkey had multiple resistances. Quinolone and gentamycin resistance found to be very high. Fusidic acid resistance was low and erythromycin and lincomycin susceptibility found to be relatively high. This study indicated that resistance to linezolid and vancomycin resistance is not emerged among MRSA isolates from Turkish hospitals.