Abstract. Introduction. Editor: G. Lina

Transcription

1 ORIGINAL ARTICLE BACTERIOLOGY Impact of psm-mec in the mobile genetic element on the clinical characteristics and outcome of SCCmec-II methicillin-resistant Staphylococcus aureus bacteraemia in Japan T. Aoyagi 1, *, C. Kaito 2, *, K. Sekimizu 2, Y. Omae 2, Y. Saito 2, H. Mao 2, S. Inomata 1, M. Hatta 1, S. Endo 1, H. Kanamori 1,Y.Gu 3, K. Tokuda 1, H. Yano 1, M. Kitagawa 1 and M. Kaku 1 1) Department of Infection Control and Laboratory Diagnostics, Internal Medicine, Tohoku University Graduate School of Medicine, Sendai, 2) Graduate School of Pharmaceutical Sciences, The University of Tokyo, Tokyo and 3) Department of Regional Cooperation for Infectious Diseases, Tohoku University Graduate School of Medicine, Sendai, Japan Abstract Over-expression of alpha-phenol-soluble modulins (PSMs) results in high virulence of community-associated methicillin-resistant Staphylococcus aureus (MRSA). The psm-mec gene, located in the mobile genetic element SCCmec-II, suppresses PSMas production. Fifty-two patients with MRSA bacteraemia were enrolled. MRSA isolates were evaluated with regard to the psm-mec gene sequence, bacterial virulence, and the minimum inhibitory concentration (MIC) of vancomycin and teicoplanin. Fifty-one MRSA isolates were classified as SCCmec-II, and 10 had one point mutation in the psm-mec promoter. We compared clinical characteristics and outcomes between mutant MRSA and wild-type MRSA. Production of PSMa3 in mutant MRSA was significantly increased, but biofilm formation was suppressed. Wild-type MRSA caused more catheter-related bloodstream infections (30/41 vs. 3/10, p ), whereas mutant MRSA formed more deep abscesses (4/10 vs. 3/41, p 0.035). Bacteraemia caused by mutant MRSA was associated with reduced 30-day mortality (1/10 vs. 13/41, p 0.25), although this difference was not significant. The MIC 90 of teicoplanin was higher for wild-type MRSA (1.5 mg/l vs. 1 mg/l), but the MIC of vancomycin was not different between the two groups. The 30-day mortality of MRSA with a high MIC of teicoplanin ( 1.5 mg/l) was higher than that of strains with a lower MIC ( 0.75 mg/l) (6/10 vs. 6/33, p 0.017). Mutation of the psm-mec promoter contributes to virulence of SCCmec-II MRSA, and the product of psm-mec may determine the clinical characteristics of bacteraemia caused by SCCmec-II MRSA, but it does not affect mortality. Keywords: clinical characteristics, MRSA, phenol-soluble modulins, psm-mec mutation, teicoplanin, vancomycin Original Submission: 31 October 2013; Revised Submission: 20 January 2014; Accepted: 27 January 2014 Editor: G. Lina Article published online: 30 January 2014 Clin Microbiol Infect 2014; 20: / Corresponding author: Tetsuji Aoyagi, Department of Infection Control and Laboratory Diagnostics, Tohoku University Graduate School of Medicine, 1-1 Seiryoumachi, Aobaku, Sendai , Japan tetsujiaoyagi@med.tohoku.ac.jp *These authors contributed equally to this work. Introduction Methicillin-resistant Staphylococcus aureus (MRSA) has rapidly spread in healthcare settings worldwide [1]. Community-associated (CA) MRSA, the USA300 (SCCmec-IV) and USA400 (SCCmec-V) strains, has been observed in especially North America over the last decade [2]. The virulence of CA-MRSA is higher than that of hospital-associated (HA)-MRSA, possibly because of the production of Panton-Valentine leukocidine (PVL) [3]. However, there is controversy as to whether the role of PVL affects disease severity and clinical outcome in MRSA infection [4 6]. Recently, Wang et al. [7] found novel cytolytic peptides to be the alpha-type phenol-soluble modulins (PSMs), which are encoded in an operon present in all sequenced S. aureus strains. PSMas consist of amino acids and contribute to evasion of the innate immune system by MRSA [7]. Production Clinical Microbiology and Infection ª2014 European Society of Clinical Microbiology and Infectious Diseases

2 CMI T. Aoyagi et al. psm-mec determines MRSA clinical feature 913 of PSMas is elevated in the most prevalent CA-MRSA strains, and PSMas contribute to CA-MRSA virulence in vivo [7]. Recent reports suggest that the transcription and translation products of psm-mec, which is located in the mobile genetic elements SCCmec-II and -III in HA-MRSA, suppress the production of PSMas [8,9]. Limited surveillance data on the genetic characteristics of MRSA are available in Japan; the NY/Japan clone (SCCmec-II) was predominantly clinically isolated. SCCmec-IV MRSA was detected in only 4 20% of samples, and these strains did not contain the luks/f-pv gene [10,11]. Thus, the characteristics of CA-MRSA strains in Japan are different from those of strains reported in other countries. Mutations in the psm-mec promoter found in HA-MRSA (SCCmec-II) attenuated the transcription of psm-mec. These strains showed increased expression of PSMas to levels observed in the USA300 and USA400 strains [9]. We hypothesized that PSMas may be important in determining the manifestation and severity of invasive MRSA infections. Our aim in this study was to examine whether SCCmec-II MRSA strains with mutations in the psm-mec promoter, which produce high amounts of PSMas, are associated with clinical characteristics and outcome in bacteraemic patients. Patients and Methods Patients and study design In total, 72 patients with MRSA bacteraemia were recruited retrospectively from March 2009 to December 2011 at Tohoku University Hospital, a 1300-bed tertiary-care teaching hospital in Japan. Bacteraemia was defined based on positive blood culture with systemic manifestation of infection. We excluded patients who (i) presented other pathogens in the blood culture during one MRSA bacteraemia episode; (ii) were dead within 24 h after the blood culture was obtained; and (iii) were under 18 years of age. Thus, 52 patients were ultimately enrolled. The following data were collected from each patient s electric medical record: age, sex, hospitalization data, housing, co-morbidities, infection site, risk factors for MRSA infection, including recent surgery, wounds and insertion of catheters, and site of MRSA infection. Complications of MRSA bacteraemia were evaluated using radiological records such as computed tomography in addition to surgical pathology and additional culture results. The outcomes studied included 30-day mortality and length of hospital stay (LOHS). Data collection was approved by the Ethics Committees of Tohoku University. Microbiological molecular analysis Methicillin-resistant Staphylococcus aureus was confirmed by meca gene carriage, and the presence of the luks/f-pv gene was investigated by polymerase chain reaction (PCR) [12]. Type of SCCmec was determined by using SCCmec-II specific primers [12] or the class of the mec and the ccr gene complex [13]. DNA fragments including the psm-mec were sequenced by using primers as previously reported [8]. Type of polymorphic region of the protein A gene (spa) was determined by sequenced spa PCR [14]. We present the primers used in this study in Table S1. Measurement of PSMs The amount of PSMas was measured as previously described [9]. Briefly, MRSA culture supernatants were evaporated and the remaining solid was dissolved in 40% acetonitrile. The supernatants remaining after subsequent centrifugation were evaporated, and the evaporated products were dissolved in water and subjected to reversed-phase high-performance liquid chromatography (HPLC). d-hemolysin (Hld) and PSMa1 were not separated in this system. Biofilm formation assay Methicillin-resistant Staphylococcus aureus was cultured in tryptic soy broth containing 0.25% glucose in 96-well polystyrene plates for 3 days at 37 C. Cells attached to the plate were stained with safranin; staining was measured according to the absorbance at 490 nm. Expression of AgrA Protein of AgrA was obtained through sodium dodecyl sulphate (SDS)-polyacrylamide gel. Expression of AgrA was performed by western blotting assay as previously described [9]. The band intensity was measured by densitometry scanning (Image J 1.45 s, NIH). Antimicrobial susceptibility The minimal inhibitory concentration (MIC) of vancomycin and teicoplanin was determined by the E-test (range mg/l; biomerieux, Lyon, France). The MIC breakpoint for vancomycin and teicoplanin resistance was >2 mg/l according to EUCAST [15]. Statistical analysis Descriptive statistics, such as means, standard deviations, frequencies and percentages, were collected. The chi-squared test, Fisher s exact test and unpaired t-test were conducted using Graphpad prism 5 (GraphPad Software, La Jolla, CA, USA). p < 0.05 was considered statistically significant.

3 914 Clinical Microbiology and Infection, Volume 20 Number 9, September 2014 CMI Results Molecular characteristics of MRSA isolated from blood cultures Figure 1 shows the gene sequencing data; 10 of 52 MRSA isolates (19.2%) had a point mutation (-7T>C) in the psm-mec promoter (Table 1). Fifty-one out of 52 (98.1%) MRSA isolates were classified as SCCmec-II, but one MRSA isolate without psm-mec gene was classified as SCCmec-IV. None of the MRSA isolates carried the LukS/F-PV gene. Forty out of 51 (78.4%) SCCmec-II MRSA isolates were classified as spa type t002: 31 (76%) of intact psm-mec MRSA; nine (90%) of a mutated psm-mec. These data indicated that MRSA strains with a mutated psm-mec promoter and intact psm-mec have a closely related genetic background. Production of PSMas, biofilm formation and AgrA expression by MRSA isolates PSMa3 production by MRSA isolates with a mutated psm-mec promoter was five-fold higher than that by MRSA isolates with intact psm-mec (p 0.023) and was equal to that by the SCCmec-IV MRSA isolate that did not contain psm-mec (Fig. 2b). However, there was no difference in the expression of PSMa1 + hld between mutant and wild-type MRSA (Fig. 2a). Mutant MRSA formed significantly less biofilm than wild-type strains (p ), and the MRSA that did not contain psm-mec also showed decreased biofilm formation (Fig. 2c). The accessory gene regulator (agr) system represents a prototype of quorum-sensing regulators, which has been recognized as a pivotal regulator of virulence factor expression [16]. AgrA binds to the psm promoter regions directly and regulates PSMas production [17]. However, psm-mec-rna inhibits translation of the agra gene by specific binding to agra mrna [9]. In our study, the AgrA expression of both no psm-mec and mutant MRSA was higher than that of wild-type MRSA (Fig. 2d). Comparison of clinical characteristics and outcome between mutant and wild-type MRSA Among the 51 SCCmec-II MRSA isolates, we did not detect any difference in age, sex, underlying disease or hospital settings between mutant and wild-type MRSA strains. Hospital-associated community onset (HACO) MRSA bacteraemia occurred more frequently in patients infected by mutant MRSA than in those infected by wild-type strains (4/10 vs. 4/41, p 0.038; Table 2). In total, 33 (63.5%) patients with MRSA bacteraemia were diagnosed as having catheter-related blood stream infection (CRBSI), and intact MRSA isolates were more frequently found in patients with CRBSI (30/41 vs. 3/10, p ). However, mutant MRSA was obtained more frequently in cases of pneumonia with MRSA bacteraemia (2/10 vs. 0/41, p 0.035). Additionally, 7/51 (13.5%) patients had the complication of abscess formation, and mutant MRSA was associated with abscess formation more frequently than wild-type MRSA (4/10 vs. 3/41, p 0.035; Table 2). In total, 14/51 (27.5%) patients were dead within 30 days of MRSA bacteraemia. The 30-day mortality rate was lower in MRSA bacteraemia caused by mutant MRSA than in bacteraemia caused by the wild-type strains, but the difference was not significant (1/10 vs. 13/41, p 0.25). However, the patients LOHS for bacteraemia caused by mutant MRSA was longer than that for bacteraemia caused by wild-type strains (57 days vs. 37 days, p ), although this difference also did not reach statistical significance. Mutation in the psm-mec promoter (-7T>C) psm-mec ORF TTGTTTGATATTATACTTAATGTATCTTAAATAGAAAGAGGGTATGCATATGGATTTCACTGGTGTTATTACAAGCATTATTGATTTAAT AACAAACTATAATATGAATTACATAGAATTTATCTTTCTCCCATACGTATACCTAAAGTGACCACAATAATGTTCGTAATAACTAAATTA CAAGACTTGCATTCAGGCTTTCGGTTAATTTTTTCAACTA GTTCTGAACGTAAGTCCGAAAGCCAATTAAAAAAGTTGAT -5 FIG. 1. Nucleotide sequence of the psm-mec open reading frame (ORF) in SCCmec-II. The nucleotide sequence of the psm-mec ORF and its promoter is shown [8, 30]. The psm-mec ORF (grey) is encoded from left to right. Black bold nucleotide A indicates the transcription start site (+1) of the psm-mec ORF. Underlined nucleotides (TATACT) indicate the -10 region of the psm-mec promoter; a -7T>C point mutation in the promoter region was found in this study.

4 CMI T. Aoyagi et al. psm-mec determines MRSA clinical feature 915 TABLE 1. Molecular characteristics of 52 MRSA strains from blood cultures psm-mec gene Number of isolates (n = 52) SCCmec type spa type II IV t002 t242 t437 t045 t439 t539 t688 t1094 t8602 NT LukS/F-PV Absence The mutation in the psm-mec promotor (-7T>C) Intact NT, non-typable. (a) (X10 6 AU) 80 PSMα1+Hld (b) (X10 6 AU) 20 PSMα3 Production of PSMα1+Hld P = P = 0.023* 15 Production of PSMα psm-mec absent MRSA Mutated psm-mec Wild-type psm-mec 0 psm-mec absent MRSA Mutated psm-mec Wild-type psm-mec (c) Biofilm formation (OD490) psm-mec absent MRSA Biofilm formation Mutated psm-mec P = * Wild-type psm-mec (d) Relative expression of AgrA per wild-type psm-mec P = 0.024* psm-mec absent MRSA P = * Mutated psm-mec FIG. 2. Mutation of the psm-mec promoter in SCCmec-II MRSA strains increases the amount of extracellular PSMs and decreases biofilm formation. The amount of PSMa1 + Hld (a) and PSMa3 (b) in each clinically isolated MRSA strain (psm-mec-absent MRSA (n = 1), mutated psm-mec promoter MRSA (n = 10), wild-type psm-mec (n = 41)) was measured by HPLC. (c) Biofilm formation of each MRSA strain onto polystyrene microplates was examined. (a c) Mean standard deviation (SD) from three independent experiments is presented. Statistical analysis was performed using the Mann Whitney test. (d) Cell extra production of MRSA culture was subjected to western blotting by anti-agra IgG and band intensities of AgrA were measured. Data analysis was presented to calculate the relative ratio against AgrA expression of wild-type psm-mec. Means and SD from three independent experiments are presented. Statistical analysis was performed using the Student t-test. Difference between MIC values of vancomycin and teicoplanin in SCCmec-II MRSA Methicillin-resistant Staphylococcus aureus bacteraemia and pneumonia caused by high-mic glycopeptides (vancomycin and teicoplanin) are associated with an unfavourable outcome [18 20]. We did not detect any difference in the MIC 50 and MIC 90 values of vancomycin between mutant and wild-type MRSA strains (MIC 50, 2 mg/l vs. 2 mg/l; MIC 90, 1.5 mg/l vs. 1.5 mg/l). However, the MIC 50 and MIC 90 values of teicoplanin in MRSA were lower than those in wild-type strains (MIC 50, 0.75 mg/l vs. 1 mg/l; MIC 90, 1 mg/l vs. 1.5 mg/l; Fig. 3a). In our study, the 30-day mortality rate did not depend on the MIC of vancomycin; however, MRSA isolates with a high

5 916 Clinical Microbiology and Infection, Volume 20 Number 9, September 2014 CMI TABLE 2. Comparison of clinical characteristics and outcome between SCCmec-II MRSA isolates with and without a mutated psm-mec promoter Mutant MRSA a (n = 10) Wild-type MRSA b (n = 41) p value Age (median, 25 75%) 71 ( ) 72 ( ) >65 year 7 (64%) 31 (76%) Male 8 (73%) 26 (63%) Underlying condition, No. (%) Cardiovascular disease 4 (40%) 13 (32%) Malignancy (solid cancer) 2 (20%) 8 (20%) Central nervous system 1 (10%) 9 (22%) disease Chronic renal failure 2 (20%) 7 (17%) Diabetes mellitus 3 (30%) 6 (15%) Autoimmune disease 2 (20%) 5 (12%) Corticosteroid use 2 (20%) 5 (12%) Digestive disease 0 (0%) 6 (15%) Chronic respiratory disease 1 (10%) 3 (7%) Chronic liver disease 0 (0%) 3 (7%) Charlson co-morbity index 2.60 (2.17) 2.31 (1.88) (mean, SD) Hospital setting, No. (%) Emergency room 1 (10%) 10 (24%) Medical wards 5 (50%) 16 (39%) Surgical wards 4 (40%) 15 (37%) 1.00 Site of acquisition, No. (%) Community-associated 0 (0%) 0 (0%) infection c Healthcare-associated 4 (40%) 4 (10%) 0.041* community onset d Hospital onset e 6 (60%) 37 (90%) Primary site of infection CRBSI 3 (30%) 30 (73%) * CRBSI + multiple sites 1 (10%) 3 (10%) Skin and soft tissue 1 (10%) 3 (7%) 1.00 infection Surgical wound infection 0 (0%) 2 (5%) 1.00 Pneumonia 2 (20%) 0 (0%) 0.035* IE 0 (0%) 1 (2%) 1.00 Vascular infection other 1 (10%) 0 (0%) than IE Intra-abdominal infection 1 (10%) 1 (2%) Bone and joint 1 (10%) 1 (2%) UTI 1 (10%) 0 (0%) Sepsis (origin unknown) 0 (0%) 3 (7%) 1.00 Complication with MRSA bacteraemia Abscess formation 4 (40%) 3 (7%) 0.035* Deep vein thrombosis 0 (0%) 2 (5%) 1.00 Treatment Vancomycin 6 (60%) 18 (44%) 0.49 Teicoplanin 2 (20%) 15 (37%) 0.46 Other agents of 2 (20%) 8 (20%) 1.00 anti-mrsa drugs f Outcomes 30-day mortality 1 (10%) 13 (32%) Length of hospital stay (median, 25 75%) 57 ( ) 37 ( ) MRSA, methicillin-resistant Staphylococcus aureus; CRBSI, catheter-related blood stream infection; IE, infectious endocarditis; UTI, urinary tract infection. a Mutant MRSA: MRSA with a point mutation (-7T>C) in the psm-mec promoter. b Wild-type MRSA: MRSA intact in the psm-mec promoter. c Community-associated infection: if the culture was obtained from an outpatient or obtained 48 h after admission from a patient without documentation of a healthcare risk factor. d Healthcare-associated community onset (HACO): if the culture was obtained from an outpatient or obtained 48 h after admission from a patient with a known healthcare risk factor. e Hospital onset: if the first positive blood culture was obtained 48 h after admission. f Other agents of anti-mrsa drugs were included in Linezolid and Arbekacin. Data indicate the number (%) of patients, median SD and median (25 75 percentile). *p < MIC of teicoplanin ( 1.5 mg/l) (i.e. isolates with an intact psm-mec gene) were associated with significantly higher mortality than isolates with low teicoplanin MIC ( 0.75 mg/ L) (6/10 vs. 6/33, p 0.017). Discussion CA-MRSA is suggested to secrete a high amount of toxins, including PSMa, PVL and Hld. These toxins are involved in evading and destroying host defences, and have been associated with skin infections, necrotizing pneumonia and abscess formation in animal models [21]. However, the molecular mechanism underlying the high virulence of CA-MRSA is not fully understood in humans. PVL is not associated with increased mortality from MRSA bacteraemia, and most cases of MRSA infection are complicated with bacteraemia associated with skin and soft tissue infection [4,6]. In some countries, PVL-positive strains are not present in many cases of MRSA bacteraemia [22,23]. In this study, we did not observe PVL-positive MRSA strains, and 98% of the MRSA strains were classified as SCCmec-II. These data suggest that factors other than PVL contribute to the pathogenesis of MRSA in Japan. PSMas and Hld had lytic activity against neutrophils [7]. However, it remains unclear whether these cytolytic toxins contribute to the clinical characteristics and outcome of human MRSA infections, as only one case report has described the potential association of an SCCmec-II MRSA strain with high PSMas and Hld with severe infection in Japan [24]. In this study, 20% of MRSA isolates had a mutation (-7T>C) in the psm-mec promoter, and these strains produced a significantly higher amount of PSMa3 than the wild-type MRSA strains. PSMa3 has higher proinflammatory and lysis activity compared with other PSMas and Hld [7]. Thus, a high amount of PSMa3 production may determine the characteristics of MRSA infection. On the other hand, wild-type MRSA isolates had significantly higher biofilm formation than the mutant strains. These results suggest that the wild-type psm-mec in SCCmec-II MRSA inhibits the virulence properties that lead to invasive infections such as pneumonia and abscess formation while promoting the virulence properties associated with attachment to intravenous catheters that lead to CRBSI. CRBSI was more frequently associated with HA-MRSA infection, but CA-MRSA was associated with a higher proportion of pneumonia and deep-site abscesses with bacteraemia [25 27]. In our study, mutant MRSA caused infections other than CRBSI and was frequently associated with pneumonia and abscess formation. However, CA-MRSA was previously associated with lower mortality or better outcomes than HA-MRSA [25,26,28], possibly because the patients with CA-MRSA bacteraemia in these studies were younger and had fewer co-morbidities. In our study, there was no difference in age or co-morbidities

6 CMI T. Aoyagi et al. psm-mec determines MRSA clinical feature 917 (a) Vancomycin MIC Teicoplanin MIC Number of patients Number of patients Mutated psm-mec Wild-type psm-mec 0 <= >=2 (mg/l) 0 <= >=2 (mg/l) (b) 30-day mortality and vancomycin MIC 30-day mortality and teicoplanin MIC Non-survivor Survivor 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% <= >=2 0% <= >=2 (mg/l) (mg/l) FIG. 3. MIC values of vancomycin and teicoplanin in SCCmec-II MRSA strains. Vancomycin and teicoplanin MIC values for SCCmec-II MRSA strains (mutated psm-mec (n = 10), wild-type psm-mec (n = 41)) were evaluated by the E-test. between patients with mutant MRSA and those with wild-type MRSA, although the 30-day mortality rate for bacteraemia caused by mutant MRSA was lower than that for bacteremia caused by wild-type MRSA. The average LOHS of bacteraemic patients with mutant MRSA was longer than that of patients with wild-type MRSA. Taken together, these findings suggested that the product of psm-mec may be involved in clinical characteristics but does not contribute to mortality. In HACO MRSA isolates, SCCmec-II and -III strains were found to be present together with SCCmec -IV and -V strains [27,28]. In our study, eight of nine HACO MRSA strains were classified as SCCmec-II, and 40% of the mutant MRSA strains were isolated within 48 h of admission. SCCmec-II mutant MRSA strains may cause MRSA bacteraemia in patients with a history of exposure to community-based medical care. Methicillin-resistant Staphylococcus aureus bacteraemia and higher vancomycin MIC values determined by the E-test have a higher likelihood of mortality and treatment failure [18]. In addition, CA-MRSA bacteraemia has a better outcome than HA-MRSA bacteraemia because of the difference in vancomycin MIC [29]. Teicoplanin is a glycopeptide alternative to vancomycin and has been widely used in Europe and East Asia. In previous studies, a higher MIC of teicoplanin (>1.5 mg/l), as determined by the E-test, predicted unfavourable outcome and higher mortality rate in bacteraemia [19] and pneumonia [20], but these reports did not evaluate the genetic background of the MRSA strains. In our study, there was no difference in the MIC 50 and MIC 90 of vancomycin for the two types of SCCmec-II MRSA, but the MIC 50 and MIC 90 of teicoplanin for wild-type MRSA were higher than those of the mutant MRSA. In addition, the mortality rate was not associated with the MIC of vancomycin, but the mortality of SCCmec-II MRSA with a higher MIC of teicoplanin ( 1.5 mg/l) was significantly higher than that of SCCmec-II MRSA with a lower MIC of teicoplanin ( 0.75 mg/l). These data indicate that the MIC of teicoplanin may be a prognostic factor for bacteraemia caused by SCCmec-II MRSA, and that the lower 30-day mortality of bacteraemia associated with mutant MRSA may be related to the lower MIC of teicoplanin. This observational study has some limitations. First, this was a single-centre study conducted in Japan with a small sample

7 918 Clinical Microbiology and Infection, Volume 20 Number 9, September 2014 CMI size; therefore it may be possible that we could not find a significant difference in clinical outcome between patients with mutant MRSA and those with wild-type MRSA. Moreover, we could not evaluate differences in clinical characteristics and bacteraemia outcome between SCCmec-IV MRSA and mutant SCCmec-II MRSA because the SCCmec-IV MRSA strain was only isolated from a few samples. Second, Queck et al. found that psm-mec has a positive effect on virulence of the HA-MRSA strain MSA890, where the amount of the translational product of psm-mec was higher than that of other PSMs. However, other strains in which the amount of the translational product of psm-mec was lower than that of other PSMs did not recapitulate this effect in vitro and in vivo [30]. To determine the extent to which our findings can be generalized, further study is needed. Third, the clinical data in this observational study were retrospectively collected and are therefore subject to information bias. However, given that the two groups were compared according to molecular typing and antimicrobial susceptibility, this bias was likely to be inconsequential. In conclusion, high production of PSMa3 by MRSA strains containing mutated psm-mec was not associated with the treatment outcome of MRSA bacteraemia but affected its clinical characteristics. The difference in clinical impact between these two different types of MRSA strains was independent of teicoplanin MIC. Furthermore, the emergence of the MRSA strain with mutated psm-mec and a high MIC of teicoplanin indicates a need for vigorous strategies for clinical management and infection control in both hospital and community settings in Japan. Acknowledgements The authors thank Makoto Katsumi, Mina Kawauchi and Mitsuaki Nagasawa for their technical assistance. This study was supported by Grants-in-Aid for Scientific Research ( , ). Transparency Declaration All authors: no reported conflicts. Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Primer used in this study. References 1. Barrett FF, McGehee RF Jr, Finland M. Methicillin-resistant Staphylococcus aureus at Boston City Hospital. Bacteriologic and epidemiologic observations. N Engl J Med 1968; 279: Zetola N, Francis JS, Nuermberger EL, Bishai WR. Community-acquired meticillin-resistant Staphylococcus aureus: an emerging threat. Lancet Infect Dis 2005; 5: Vandenesch F, Naimi T, Enright MC et al. Community-acquired methicillin-resistant Staphylococcus aureus carrying Panton-Valentine leukocidin genes: worldwide emergence. Emerg Infect Dis 2003; 9: Voyich JM, Otto M, Mathema B et al. Is Panton-Valentine leukocidin the major virulence determinant in community-associated methicillin-resistant Staphylococcus aureus disease? J Infect Dis 2006; 194: Seybold U, Kourbatova EV, Johnson JG et al. Emergence of community-associated methicillin-resistant Staphylococcus aureus USA300 genotype as a major cause of health care-associated blood stream infections. Clin Infect Dis 2006; 42: Sharma-Kuinkel BK, Ahn SH, Rude TH et al. Presence of genes encoding panton-valentine leukocidin is not the primary determinant of outcome in patients with hospital-acquired pneumonia due to Staphylococcus aureus. J Clin Microbiol 2012; 50: Wang R, Braughton KR, Kretschmer D et al. Identification of novel cytolytic peptides as key virulence determinants for community-associated MRSA. Nat Med 2007; 13: Kaito C, Saito Y, Nagano G et al. Transcription and translation products of the cytolysin gene psm-mec on the mobile genetic element SCCmec regulate Staphylococcus aureus virulence. PLoS Pathog 2011; 7: e Kaito C, Saito Y, Ikuo M et al. Mobile genetic element SCCmec-encoded psm-mec RNA suppresses translation of agra and attenuates MRSA virulence. PLoS Pathog 2013; 9: e Takizawa Y, Taneike I, Nakagawa S et al. A Panton-Valentine leucocidin (PVL)-positive community-acquired methicillin-resistant Staphylococcus aureus (MRSA) strain, another such strain carrying a multiple-drug resistance plasmid, and other more-typical PVL-negative MRSA strains found in japan. J Clin Microbiol 2005; 43: Motoshima M, Yanagihara K, Morinaga Y et al. Genetic diagnosis of community-acquired MRSA: a multiplex real-time PCR method for Staphylococcal cassette chromosome mec typing and detecting toxin genes. Tohoku J Exp Med 2010; 220: Makgotlho PE, Kock MM, Hoosen A et al. Molecular identification and genotyping of mrsa isolates. FEMS Immunol Med Microbiol 2009; 57: Zhang KY, McClure JA, Elsayed S, Louie T, Conly JM. Novel multiplex PCR assay for characterization and concomitant subtyping of staphylococcal cassette chromosome mec types I to V in methicillin-resistant Staphylococcus aureus. J Clin Microbiol 2005; 43: Shopsin B, Gomez M, Montgomery SO et al. Evaluation of protein A gene polymorphic region DNA sequencing for typing of Staphylococcus aureus strains. J Clin Microbiol 1999; 37: European Committee on Antimicrobial Susceptibility Testing Breakpoint tables for interpretation of MICs and zone diameters Version 1.3, January 5, CAST_files/Disk_test_documents/EUCAST_breakpoints_v1.3_pdf.pdf accesed 30 August, Novick RP, Geisinger E. Quorum sensing in staphylococci. Annu Rev Genet 2008; 42: Queck SY, Jameson-Lee M, Villaruz AE et al. RNAIII-independent target gene control by the agr quorum-sensing system: insight into the evolution of virulence regulation in Staphylococcus aureus. Mol Cell 2008; 32:

8 CMI T. Aoyagi et al. psm-mec determines MRSA clinical feature van Hal SJ, Lodise TP, Paterson DL. The clinical significance of vancomycin minimum inhibitory concentration in Staphylococcus aureus infections: a systematic review and meta-analysis. Clin Infect Dis 2012; 54: Chang HJ, Hsu PC, Yang CC et al. Influence of teicoplanin MICs on treatment outcomes among patients with teicoplanin-treated methicillin-resistant Staphylococcus aureus bacteraemia: a hospital-based retrospective study. J Antimicrob Chemother 2012; 67: Chen KY, Chang HJ, Hsu PC et al. Relationship of teicoplanin MICs to treatment failure in teicoplanin-treated patients with methicillin-resistant Staphylococcus aureus pneumonia. J Microbiol Immunol Infect 2013; 46: Gordon RJ, Lowy FD. Pathogenesis of methicillin-resistant Staphylococcus aureus infection. Clin Infect Dis 2008; 46(Suppl 5): S350 S Ellington MJ, Perry C, Ganner M et al. Clinical and molecular epidemiology of ciprofloxacin-susceptible MRSA encoding PVL in England and Wales. Eur J Clin Microbiol Infect Dis 2009; 28: Kim JS, Park JS, Song W et al. Panton-Valentine leukocidin positive Staphylococcus aureus isolated from blood in Korea. Korean J Lab Med 2007; 27: Khokhlova O, Tomita Y, Hung WC et al. Elderly infection in the community due to ST5/SCCmecII methicillin-resistant Staphylococcus aureus (the New York/Japan clone) in Japan: panton-valentine leukocidin-negative necrotizing pneumonia. J Microbiol Immunol Infect. 2012; doi: /j.jmii [Epub ahead of print]. 25. Lalani T, Federspiel JJ, Boucher HW et al. Associations between the genotypes of Staphylococcus aureus bloodstream isolates and clinical characteristics and outcomes of bacteremic patients. J Clin Microbiol 2008; 46: Chen SY, Wang JT, Chen TH et al. Impact of traditional hospital strain of methicillin-resistant Staphylococcus aureus (MRSA) and community strain of MRSA on mortality in patients with community-onset S aureus bacteremia. Medicine (Baltimore) 2010; 89: Lee SC, Lee CW, Shih HJ, Chiou MJ, See LC, Siu LK. Clinical features and risk factors of mortality for bacteremia due to community-onset healthcare-associated methicillin-resistant S. aureus. Diagn Microbiol Infect Dis 2013; 76: Lessa FC, Mu Y, Ray SM et al. Impact of USA300 methicillin-resistant Staphylococcus aureus on clinical outcomes of patients with pneumonia or central line-associated bloodstream infections. Clin Infect Dis 2012; 55: Chen SY, Liao CH, Wang JL et al. Methicillin-resistant Staphylococcus aureus (MRSA) staphylococcal cassette chromosome mec genotype effects outcomes of patients with healthcare-associated MRSA bacteremia independently of vancomycin minimum inhibitory concentration. Clin Infect Dis 2012; 55: Queck SY, Khan BA, Wang R et al. Mobile genetic element-encoded cytolysin connects virulence to methicillin resistance in MRSA. PLoS Pathog 2009; 5: e