a comparison of virulence patterns and in vivo fitness ... · between hospital- and...

ARTICLE

A comparison of virulence patterns and in vivo fitnessbetween hospital- and community-acquired methicillin-resistantStaphylococcus aureus related to the USA400 clone

M. A. Guimarães & M. S. Ramundo & M. A. Américo & M. C. de Mattos & R. R. Souza &

E. S. Ramos-Júnior & L. R. Coelho & A. Morrot & P. A. Melo & S. E. L. Fracalanzza &

F. A. Ferreira & A. M. S. Figueiredo

Received: 9 August 2014 /Accepted: 16 September 2014 /Published online: 14 October 2014# Springer-Verlag Berlin Heidelberg 2014

Abstract Methicillin-resistant Staphylococcus aureus(MRSA) isolates genetically related to the CA-MRSA cloneMW2/USA400 (ST1-SCCmecIV lineage) from the UnitedStates have emerged in hospitals in Rio de Janeiro and areassociated with nosocomial bloodstream infections. To under-stand the virulence mechanisms involved in the adaptability ofST1 isolates as a hospital pathogen in Rio de Janeiro, wecompared the virulence traits and fitness properties of theBrazilian isolates with those displayed by the CA-MRSAisolates from the United States. Similar to the USA400 fromthe United States, all the Brazilian isolates tested carried thegenes encoding SEH and LukDE. In contrast, none of theBrazilian isolates carried the lukSFPVL, sea, sec, and sekgenes. Competition experiments in mice demonstrated a

significant increase in the fitness for the CA-MRSA isolatesMW2 and USA400-0051 from the United States compared toother isolates. In the foreign body animal model, 83 % moreNorth-American bacterial cells were recovered compared tothe Brazilian ST1 isolates. Differences in gene expression ofimportant virulence factors were detected. Transcription ofrnaIII and psmα3was increased about two-fold in the isolatesfrom the United States, and sasG about two-fold in theBrazilian isolates. Thus, it is possible that the virulence atten-uation observed among the Brazilian hospital isolates, associ-ated with the acquisition of multiple resistant determinants,are consequences of microevolutionary events that contribut-ed to the necessary fitness adjustment of this lineage, allowinga typically community-acquired MRSA (MW2/USA400) toemerge as a successful hospital pathogen (Brazilian ST1-SCCmecIV).

Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) hasbeen considered one of the most important hospital pathogensworldwide [1]. The mecA gene, the determinant of methicillinresistance in S. aureus, is carried by the staphylococcal cas-sette chromosome mec (SCCmec), and 11 types of SCCmechave been described to date (www.sccmec.org; last accessedin June 19, 2014). The wide spectrum of diseases caused by S.aureus has been linked to an incredible number of virulencefactors, which promote adherence to and internalization intothe host cells, evasion of the immune response, biofilmformation, and induction of toxic/harmful effects to the hosttissues [2–8]. Hospital-associated MRSA (HA-MRSA) arecausative agents of catheter-related bloodstream infection(CR-BSI), ventilator-associated pneumonia (VAP), surgical

M. A. Guimarães :M. S. Ramundo :M. A. Américo :M.C. deMattos :R. R. Souza : S. E. L. Fracalanzza : F. A. Ferreira :A. M. S. Figueiredo (*)Departamento de Microbiologia Médica, Instituto de MicrobiologiaPaulo de Góes, Universidade Federal do Rio de Janeiro, Rio deJaneiro, Brazile-mail: [email protected]

E. S. Ramos-JúniorLaboratório de Imunologia Molecular, Instituto de Biofísica,Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil

L. R. CoelhoCampus Macaé Prof. Aluísio Teixeira, Universidade Federal do Riode Janeiro, Rio de Janeiro, Brazil

A. MorrotDepartamento de Imunologia, Instituto de Microbiologia Paulo deGóes, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil

P. A. MeloLaboratório de Farmacologia das Toxinas, Instituto de CiênciasBiomédicas, Universidade Federal do Rio de Janeiro, Rio de Janeiro,Brazil

Eur J Clin Microbiol Infect Dis (2015) 34:497–509DOI 10.1007/s10096-014-2253-1

http://www.sccmec.org/

site infections (SSI), and implantable device infections (IDI)[9, 10]. The bacterial ability to produce biofilm has beenrecognized as an important virulence factor for the pathogen-esis of CR-BSI and IDI [3, 6]. Different surface proteins havebeen implicated in biofilm formation/accumulation, includingfibronectin-binding proteins A and B (FnBPAB), S. aureussurface protein G (SasG) and staphylococcal protein A (Spa)[11–13]. In addition, many different types of proteins such asautolysins, proteases, and PBP2a have also been reported tomediate S. aureus biofilm [14, 15].

Unlike methicillin-susceptible S. aureus (MSSA), theMRSA population displays a highly clonal nature.Molecular-epidemiological studies have shown that only afew pandemicMRSA lineages are responsible for the majorityof infections worldwide [1, 16]. These pandemic staphylococ-cal strains have evolved over the years, acquiring new geneticelements via horizontal transfer and accumulating mutationsthat have altered gene expression and function [16, 17].

Until the beginning of the 1980s, MRSA infections werelimited to hospital settings, affecting mainly immunocompro-mised patients or those undergoing an invasive procedure.However, from the end of the 1980s to the beginning of the1990s, MRSA outbreaks began to emerge among healthyindividuals who lacked classical risk factors for MRSA infec-tions, and these strains were named community-acquiredMRSA (CA-MRSA) [18–20]. In contrast to typical isolatesof HA-MRSA, CA-MRSA are generally more susceptible tonon-β-lactam antibiotics, frequently harbor lukSFPVL genes(encoding for Panton-Valentine leukocidin; PVL), and oftencause skin/soft tissue infections (SSTI); however, invasiveinfections including necrotizing pneumonia, osteomielitis,and endocarditis have been reported [19, 20]. It was suggestedthat the occurrence of certain toxin repertoire and/or increasedexpression of specific toxins, in addition to the ability to evadethe host immune system, are factors that might be involved inthe dissemination and increased virulence of successful CA-MRSA strains [21]. The virulence factors frequently associat-ed with CA-MRSA pathogenesis include alpha-toxin (α-toxinor alpha-hemolysin; HLA), phenol-soluble modulins (PSM),the arginine catabolic mobile element (ACME; carried only byUSA300 isolates), and PVL [5, 7, 21–24].

A significant change in the epidemiology ofMRSAwas theentry of CA-MRSA lineages in hospital settings located indifferent regions [1, 25]. Reports on healthcare-associatedinfections involving CA-MRSA isolates related to USA300(ST8-SCCmecIV), USA400/MW2 (ST1-SCCmecIV), andUSA1100 (ST30 SCCmecIV) clones have periodically beenpublished [26–29].

Studies have reported the emergence of ST1-SCCmecIVisolates in Rio de Janeiro (related to USA400/MW2 clone)that caused healthcare-associated diseases [30, 31]. It is in-triguing that, despite the genetic relationship with the North-American MW2/USA400 clone, these USA400-related

isolates from Rio were PVL-negative, multiresistant, andmostly recovered from nosocomial bloodstream infections(BSI) [30]. In view of the unusual epidemiological circum-stances involving ST1-SCCmecIV infections in Rio deJaneiro, the aim of this study was to compare in vitro andin vivo aspects of the virulence of USA400-related isolatesfromBrazil and the United States, in order to identify potentialvirulence attributes that might account for the epidemiologicalchanges observed for the Brazilian variant.

Materials and methods

S. aureus isolates and molecular characterization

The 60 USA400/MW2-related isolates studied here were partof a convenience collection of our laboratory and were ob-tained from patients admitted to three different hospitals inRio de Janeiro as part of standard clinical care. Thirty-threeisolates were recovered from BSI (55 %) and nine fromcatheter tips (CT; 15 %). The remaining were isolated fromcolonization (C; 16.7 %), pneumonia (P; 6.7 %), skin/softtissue infections (SSTI; 3.3 %), and urinary tract infections(UTI; 3.3 %). All isolates were genotyped as CC1-ST1-SCCmecIV using pulsed-field gel electrophoresis (PFGE),SCCmec typing, restriction and modification tests (RM tests),and multilocus sequence typing (MLST) [30]. The NorthAmerican isolate USA400-0051 (ST1-SCCmecIV, PVL pos-itive), obtained from Paul Dunman (University of Nebraska,USA), and the isolate MW2 (ST1-SCCmecIV, PVL positive),obtained from Renata Rabello (Universidade FederalFluminense, Niterói/Rio de Janeiro, Brazil), were used in thisstudy as representatives of the North American MW2/USA400 clone. The S. aureus RN6390B, a gift fromRichard Novick (New York University, USA), was used as acalibrator in the gene expression experiments. The S. aureusisolate GV51 (ST239-SCCmecIII) and the S. pyogenes isolate75–194 were used as positive and negative controls, respec-tively, for testing in vitro biofilm development. This study wasapproved (#136/07) by the Human Research EthicsCommittee from the Federal University of Rio de Janeiro,RJ, Brazil.

Assessment of genes encoding staphylococcal exotoxinsby PCR

All 60 isolates of ST1-SCCmecIV were used for these assays.Additionally, we included the North-American isolatesUSA400–0051 and MW2 for comparison. DNAwas isolatedby thermal lyses, using the method described previously [32].We analyzed 13 enterotoxin genes (sea-e, seg-l, sen and seo),PVL genes (lukSPVL and lukFPVL), and the genes for LuKDEleukocidin (lukDE) and toxic shock syndrome toxin (tst). The

498 Eur J Clin Microbiol Infect Dis (2015) 34:497–509

specific primers used for amplifying segments of these genesare described in Table 1. The selection of these genes wasbased on previous reports of the virulence gene repertoire ofCA-MRSA isolates [5, 21, 38]. ThemecA gene was used as aninternal control for the PCR reactions. PCR amplificationswere performed according to standard protocols [200 μM ofeach dNTP, 0.01 μM of each primer, 0.125 U of DNA Taqpolymerase (Promega, Madison, Wisconsin, USA), 1×en-zyme buffer with MgCl2 and 100 ng of the DNA]. Reactionswere performed in a Veriti 96-well thermal cycler (AppliedBiosystems, Foster City, California, USA), using the follow-ing program: pre-denaturation at 95 °C for 4 min; 35 cycles of95 °C for 30 s, 30 s at the appropriate annealing temperature(Table 1), and 72 °C for 45 s; and a final extension at 72 °C for45 s. The PCR amplicons were visualized using a LumiBisGel Imager (DNRBio-Imaging Systems, Jerusalem, Israel)

after electrophoresis in running buffer containing 0.5 μg/mLethidium bromide.

Bacteremia mouse model

Among the 60 ST1-SCCmecIV isolates analysed, weselected two isolates (08–28 and 07–59) for experimen-tal challenges in mice based on three main characteris-tics: (1) a PFGE pattern identical to the epidemic(predominant) ST1-SCCmecIV variant from Brazil, (2)BSI association, and (3) exoprotein-gene pattern andresistance profile representatives of the group isolates.In addition, the two North-American ST1-SCCmecIVisolates (MW2 and USA400–0051), representatives ofthe MW2 and USA400 clones, were also included inthese studies for comparison. The bacteremia model

Table 1 Primers used to detect virulence genes

Gene Primer Sequence (foward / reverse) Annealingtemperature (°C)

Amplicon Reference Control strain

mecA mecA MECA P4 MECA P7 5′-TCCAGATTACAACTTCACCAGG-3′5′-CCACTTCATATCTTGTAACG-3′

55 162 [33] WB69

lukSPVL lukFPVL NewPVL-1 NewPVL-2 5′-ATCATTAGGTAAAATGTCTGGACATGATCCA-3′ 5′-GCATCAAGTGTATTGGATAGCAAAAGC-3′

55 433 [34] WB77

lukE-lukD LUKDE-1 LUKDE-2 5′-CCATCGTTTATCACTACACTAT-3′5′-CATAGTCAACAACATTTACAGC-3′

55 601 [35] WB72

sea SEA-1 SEA-2 5′-GACTTGTATGGTGCTTATTATG-3′5′-AATACTGTCCTTGAGCACCA-3′

55 361 [36] MW2

seb SEB-1 SEB-2 5′-GCTGATAAATACAAAGATAAATACG-3′ 5′-ATCCCGTTTCATAAGGCGAG-3′

55 336 [36] ATCC14458

sec mpSEC-1 mpSEC-2 5′-GTAAAGTTACAGGTGGCAAAACTTG-3′ 5′-CATATCATACCAAAAAGTATTGCCGT-3′

55 297 [34] WB72

sed SED-1 SED-2 5′-GAATTAAGTAGTACCGCGCTA-3′5′-GCTGTATTTTTCCTCCGAGAGT-3′

55 491 [36] ATCC13565

see SEE-1 SEE-2 5′-CTGAAAACAAAGAGAGTGATG-3′5′-GTGTAAATAATGCCTTGCCTGAA-3′

57 379 [36] ATCC27664

seg SEG-1 SEG-2 5′-AATTATGTGAATGCTCAACCCG-3′5′-TATGGAACAAAAGGTACTAGTTC-3′

53 634 [36] WB69

seh SEH-1 SEH-2 5′-TCACATCATATGCGAAAGCAG-3′5′-CATCTACCCAAACATTAGCAC-3′

55 371 [36] WB81

sei SEI-1 SEI-2 5′-CTCAAGGTGATATTGGTGTAG-3′5′-TATTCTTTGCCTTTACCAGTGT-3′

55 453 [36] WB69

sej mpSEJ-1 mpSEJ-2 5′-CCTTTAGTTTACAGCGATAGC-3′5′-GGTATGAAACACGATTAGTCC-3′

55 578 [36] ATCC13565

sek SEK-1 SEK-2 5′-CTTGATGTTTTTGGTATTAGTTA-3′5′-TGTGTCCTCTACTACCCAAT-3′

53 382 [36] MW2

sel SEL-1 SEL-2 5′-TAACGGCGATGTAGGTCCA-3′5′-CGCCAAAACTATTCCCCTC-3′

53 498 [36] USA100

sen mpSEN-1 mpSEN-2 5′-TAATGCTGAAGTAGACAAAAAAG-3′ 5′-AAACTCTGCTCCCACTGAAC-3′

55 615 [36] ATCC19095

seo mpSEO-1 mpSEO-2 5′-TAGAGAGTTTGTGTAAGAAGTC-3′5′-TTAAATTCAGCAGATATTCCATCT-3′

55 180 [36] WB69

tst p152 p154 5′-ATCGTAAGCCCTTTGTTG-3′5′-TGGATATAAGTTCCTTCGC -3′

55 334 [37] RN8846

Eur J Clin Microbiol Infect Dis (2015) 34:497–509 499

protocol in mice was performed as previously described[23] using Swiss female mice (six weeks old). Briefly,S. aureus isolates were grown to mid-exponential phase(OD600nm=0.7), washed once with sterile 1×phosphate-buffered saline (PBS), pH 7.4, resuspended in PBS andadjusted to contain 1x109 CFU/mL. A volume of100 μL of the adjusted bacterial suspension wasinjected into the retro-orbital vein of the anaesthetizedmouse. Control animals received the same volume ofsterile PBS. After inoculation, mouse health and diseaseadvancement were monitored every hour (up to 20 h).To accurately analyse mouse behaviour after infection,we used published guidelines to score the clinical signsof a systemic infection [39, 40]. Briefly, the signsevaluated, in an approximate order of increasing sever-ity, included ruffled hair, changes in food and waterintake, hunched back, tremor ataxia, and lethargy. Theearly recognition of the clinical signs indicating anirreversible progression to death was imperative to min-imize suffering. The time point for blood sampling waswhen the animals presented with a combination of twoor three minor signs or a single major sign, indicatingthe need for immediate euthanasia, for reducing animalsuffering. In cases of animals that have no sign ofinfection, or only one minor sign, the time point forblood sampling and euthanasia was 20 h after the in-jection. Serum γ-IFN was used with dosage using theMouse IFN-gamma DuoSet (R & D Systems;Minneapolis, MN, USA) according to the manufac-turer’s instructions. We pooled sera from six mice eachper isolate for each test. γ-IFN was chosen because itwas a good marker of infection severity in a similarbacteremia model [23]. This study was approved(# IMPG018) by the Human Resea rch E th i c sCommittee from the Federal University of Rio deJaneiro, RJ, Brazil.

In vivo competitive fitness assay

Because several host factors may interfere with bacterial fit-ness during the course of an infection, we performed anin vivo competition assay using the bacteremia model in mice.The experiment was performed as described above, exceptthat mice were inoculated with a suspension of a mixedculture (1:1) of Brazilian and North-American isolates (pair#1: MW2 and 08–28; pair #2: MW2 and 07–59; pair #3:USA400–0051 and 08–28; or pair #4: USA400 and 07–59).Brazilian and North-American isolates were separated by theantibiotic erythromycin, as the Brazilian isolates carry the ermgene [30]. Thus, blood from each animal challenged with thecompetitive pair was plated in TSAwith and without erythro-mycin (5 μg/mL) for counting CFU/mL. For this experiment

we used 12 mice for each pair of isolates, in order to achievestatistical significance.

Foreign-body mouse model

Before initiating the animal model, the ability of these isolatesto form biofilm in vitro was tested for representatives of theBrazilian (08–28 and 07–59) and North-American (MW2 andUSA400–0051) isolates. For these assays, biofilm was testedusing 96-well inert polystyrene microtiter plates (Nunclon;Nunc A/S, Roskilde, Denmark) as previously described [41].The biofilm unit (BU) was defined as indicated previously [3]and the isolates were classified as non-producers (BU ≤0.230),weak (0.230< BU ≤0.460), moderate (0.460< BU ≤0.920) orstrong producers (BU >0.920), as suggested. A total of threeindependent experiments, four replicates each, were per-formed for each isolate tested. Because biofilm is under phasevariation, the three highest BU values were used for thestatistical calculations.

The mouse subcutaneous catheter implant model was de-scribed in detail previously [41]. Briefly, two intravenouspolyurethane catheter segments (model C-UDLM-953 J;Cook Medical, Bloomington, USA) were implanted in theback of each anaesthetized young-adult BALB/c male mouse.Infection was induced 24 h after the implantation procedureby injecting a mid-exponential growth phase culture(106 CFU/10 μL) of the Brazilian (08–28 and 07–59) andNorth-American (MW2 and USA400–0051) isolates into thelumen of the implanted catheter segment. The animal waseuthanised three days post-infection, and the catheter seg-ments were surgically removed. After washing of the catheter,biofilm accumulation was determined by counting thecatheter-adherent bacteria by CFU determination. Three inde-pendent experiments were performed. This animal study wasapproved (#IMPPG013) by the Ethics Committee for AnimalCare and Use from the Federal University of Rio de Janeiro,RJ, Brazil.

Autolysis assay

Since the release of bacterial cell wall fragments can contrib-ute to the early recruitment of cells of the immune system andthereby lead to decreased bacterial virulence [42], autolysinactivity was measured in representatives of the Brazilian (08–28 and 07–59) and North-American (MW2 and USA400–0051) isolates, as previously described [43]. Briefly, a logphase culture was adjusted to OD620nm=1 in glycine-TritonX-100 buffer (50 mM glycine in 0.01 % p/v Triton X-100).One autolytic unit was defined as the activity required to causea decrease of 50 % in the initial OD620nm=1 within 30 min ofincubation at 37 °C.


Gene expression

The transcription level of relevant virulence genes(sasG, spa, fnbA, fnbB, hla and psmα3) and the regu-lator gene rnaIII (codifying for the RNAIII effectormolecule of the Agr system) were analysed for repre-sentatives of the Brazilian (08–28 and 07–59) andNorth-American (MW2 and USA400–0051) isolates.The sasG, spa, fnbA, and fnbB genes have been asso-ciated with biofilm formation/accumulation, and hla,psmα3 and spa with severity of some CA-MRSA infec-tions. For RNA preparations, bacterial cells, grown inTSB (18 h/37 °C; 250 rpm), were collected in thebeginning of the exponential phase (OD600nm=0.3) forsasG, fnbA, fnbB and spa, and in the stationary phase(OD600nm=3.0) for rnaIII, psma3 and hla gene tran-scription analyses. These assays were carried out usingtwo time points because these surface proteins are in-creased in the log phase (agr-down-regulation) and theexported proteins in the stationary phase (agr-up-regu-lation). Total RNA was prepared using the RNeasy Minikit (Qiagen; Maryland, USA) and quantified by aNanodrop Lite (Uniscience, Thermo Fisher Scientific,Wilmington, DE, USA). The RNA quality was analysedby RNA-gel electrophoresis. The reverse-transcriptionquantitative PCR (qRT-PCR) was carried out using aPower SYBR Green RNA-to-CT™ 1-Step Kit (AppliedBiosystems; Foster City, CA, USA) following manufac-turer instructions, and using the ΔΔCt comparativemethod. The primers and run conditions used for eachgene and for the endogenous control rRNA 16S werepreviously described [44]. The run was performed in theStep One™ Real Time PCR System (Appl iedBiosystems). Two biological and three technical repli-cates were used. Data were analysed using the Step OneSoftware 2.2 (Applied Biosystems). For gene expressionanalysis, a difference of one cycle, about two-foldchange, was considered relevant.

Statistical calculations

The Student’s t-test (unpaired data) was used in theexperiments of biofilm, bacterial count, interferon-gamma dosage, and gene expression to compare theNorth-American versus the Brazilian isolates, with thesignificance level of 0.1. Unless otherwise stated, datawere obtained from at least three independent experi-ments using duplicate or triplicate. For graphic repre-sentations, the data from each isolate were plotted sep-arately, and for the statistical analyses, all data from theNorth-American isolates studied were grouped (group 1)and tested against all data from the Brazilian isolates(group 2) for each experiment performed.

Results

Assessment of genes encoding exoproteins

All Brazilian isolates of ST1-SCCmecIV and the two NorthAmerican representatives of MW2/USA400 clone tested pre-sented the genes encoding enterotoxin H (SEH) and LukDEleukocidin (except the Brazilian isolate 06–02). However,different from the North American isolates, none of the 60Brazilian isolates tested amplified with primers for lukSPVL,lukFPVL, or for the genes encoding the enterotoxins A, C, or K(sea, sec or sek). The Brazilian ST1 isolates were homoge-neous concerning the virulence profiles of the genes tested(Table 2).

Comparative analysis of virulence in the bacteremia model

The post-infection behaviour of the mice was constantly mon-itored until the 20th hour after infection. Mice challenged withNorth American ST1 isolates (USA400–0051 and MW2)frequently showed signs of more severe systemic infection.However, mice inoculated with the Brazilian isolates (08–28 e07–59) showed moderate signs of infection, such as ruffledhair. The control animals (inoculated with PBS) remainedhealthy until the end of the experiment. From a group of 38–40 animals tested for each isolate, only ten did not showinfection signs. All of these ten mice were inoculated withthe Brazilian isolates (8 with 08–28 and 2 with 07–59). Fifty-five mice showed ruffled hair, hunched backs, and slightmobility impairment. Among these, 32 were injected with08–28 and 23 with 07–59 isolates. Eighty-two exhibited ex-treme mobility impairment, hunched backs, swelling of theface and nose, inability to eat or drink, and/or respiratorydistress. Among these, 13 mice were inoculated with 07–59isolate, 37 were inoculated with USA400–0051 isolate and 33with the MW2 isolate. Spontaneous death occurred for onlyten animals from the total 160 inoculated (6.3 %), three ofwhich were challenged with USA400–0051 isolate and sevenwith MW2 isolate (Fig. 1a).

The MRSA count in the blood of infected animals wasperformed in a total of 3–4 blood pools/isolate. The averageCFU in the blood was 1.9×105 and 1.8×105 CFU/mL forNorth-American isolates USA400-0051 and MW2, and 3×104 and 4×104 CFU/mL for Brazilian isolates 08–28 and 07–59, respectively. The mean differences between American andBrazilian isolates was significant (p<0.03), corresponding toa 90 % increase for the first group (Fig. 1b).

The IFN-γ concentration in the blood of infected animalswas determined in blood pools of six infected animals and atleast three independent assays were carried out for each groupof infected mice. Consistent with data from animal behaviourand blood bacteria count, the results revealed that mice infect-ed with North American isolates showed an increased IFN-γ


concentration in blood (USA400–0051=2027.5±470.0pg/mL and MW2=2515.0±878.0 pg/mL) when comparedwith those infected with Brazilian isolates (08–28=391.0±280.0 pg/mL and 07–59=272.0±153.0 pg/mL, p<0.0001;Fig. 1c).

In vivo competitive fitness

For the fitness assay, a total of 12 animals were used for eachcompetitive pair. When mice were simultaneously challengedwith a pair of Brazilian and North-American isolates, anattenuation of the signals of systemic infection was clearlynoticed (Fig. 2a), despite the fact that the number of CFUrecovered from the blood was higher (90 %; p<0.001) for

North American isolates compared with those obtained forBrazilian isolates (Fig. 2b). The levels of interferon in themixed cultures were (MW2+08–28):344±168 pg/ml;(MW2+07–59): 159±68 pg/ml; (USA400–0051+08–28):913±865 pg/ml and (USA400–0051+07–590): 712±608 pg/ml. It is noteworthy that, although still higher thanthose observed by the animals individually infected withBrazilian isolates, the levels of IFN-γ decreased significantlyin the mixed infection compared to those observed for animalsindividually inoculated with North American isolates, show-ing an 84–86 % reduction in the IFN-γ for the MW2 mixedculture and 55–65 % for the USA400–0051 (p<0.001) mixedculture (Fig. 2c). This data are in accordance with the obser-vation that the signals of systemic infection were attenuated in

Table 2 Exoprotein gene detection of North American and Brazilian MRSA isolates belonging to ST1-SCCmecIV lineage

Isolates sea seb sec sed see seg seh sei sej sek sel sen seo tst lukDE lukSFPVL

USA400-0051a +b − + − − − + − − + − − − − + +

MW2a + − + − − − + − − + − − − − + +

05–65c − − − − − − + − − − − − − − + −06–02 − − − − − − + − − − − − − − − −

a North-American isolates b + positive detection;—negative detection c Isolate 05–65 was chosen as the Brazilian representative strain, since 59 isolatesshowed identical profiles for the presence or absence of exoproteins genes tested, with 06–02 isolate as the only exception among the 60 isolates

Fig. 1 Comparative analysis of virulence in the bacteremia model. aSigns of infection presented by the inoculated mice. b Bacteria countingin mouse blood. The difference between American isolates (group 1) andBrazilian isolates (group 2) was significant (p<0.03). c IFN-gammadosage in mouse blood. MW2: Mice were inoculated with MW2 isolate,representative of the North American MW2 clone; USA400: Mice wereinoculated with the isolate USA400–0051, representative of the USA 400

clone; 08–28 and 07–59: Mice were inoculated with 08–28 or 07–59isolates, representatives of the Brazilian isolates of the ST1-SCCmecIVlineage. The difference between groups 1 and 2 was extremely significantp<0.0001. For bacterial counting and interferon dosage each bloodsample was a pool of six animals. The Student t test was used to comparethe two groups of isolates


mice inoculated with mixed cultures, in relation to the animalsinoculated with MW2 or USA400–0051 individually(Fig. 2a).

Foreign-body mouse model

Data from in vitro biofilm showed that the Brazilian (07–59and 08–28) and North American (MW2 and USA400–0051)isolates exhibited a moderate/strong biofilm phenotype,whose BU varied from 1.09±0.052 to 0.493±0.023. In theforeign body catheter model, an increased recovery of bacteriaadhered to the catheters was seen for North American isolates(USA400–0051: 1.73×108±1.55×108 CFU/mL and MW2:

1.28×108±1.03×108 CFU/mL) when compared (p=0.02) tothe Brazilian isolates (08–28: 6.98×106±6.07×106 CFU/mLand 07–59: 5.29×106±6.01×106). These data correspondedto a decay of 97 % and 96 % when comparing MW2 with 07–59 and 08–28, respectively, and a decay of 71 % and 70 %when comparing USA400–0051 with 07–59 and 08–28, re-spectively (Fig. 3).

Autolysis assay

There was no significant difference between the autolyticactivity determined for North American isolates (USA400–0051=0.75U andMW2=0.71U) and the Brazilian isolate 08–

Fig. 2 In vivo competitivefitness. a Signs of infectionpresented by the inoculatedmouse. b Bacteria counting inmouse blood. The differencebetweenAmerican isolates (group1) and Brazilian isolates (group 2)was significant (p<0.03). TheStudent t-test was used tocompare the two groups. c IFN-gamma dosage in mouse blood.08–28+USA400: Mice wereinoculated with a mixed culture of08–28 (Brazilian ST1 isolate) andUSA400–0051 (North-Americanisolate, representative of theUSA400 clone). 08–28+MW2:Mice were inoculated with amixed culture of 08–28 andMW2(North-American, representativeof the MW2 clone) isolates. 07–59+USA400: Mice wereinoculated with a mixed culture of07–59 (ST1 from Brazil) andUSA400–0051 isolates. 07–59+MW2: Mice were inoculated witha mixed culture of 07–59 andMW2 isolates. The difference inthe IFN-gamma dosage in theblood of mice inoculated withmixed culture (group 1) versussingle cultures of the Americanisolates (group 2) was verysignificant p<0.001. The Studentt-test was used to compare the twogroups


28 (0.73U). The Brazilian isolate 07–59 showed an autolyticactivity of 1.4U (Fig. 4); however, this difference was notstatistically significant.

fnbPA, sasG, spa, pmsα3, hla and rnaIII expression

Increased levels of fnbA transcripts were observed for 07–59(RQ=1.2±0.026) and MW2 (RQ=1±0.05) when comparedwith 08–28 (RQ=0.569±0.045) and USA400–0051 (RQ=0.139±0.072) isolates (Fig. 5a). However, fnbB transcriptwas only detected for the isolate MW2 (RQ=267±0.047;Fig. 5b). sasG expression of the Brazilian isolates 08–28 and07–59 (RQ=38±0.078 and RQ=40±0.032, respectively) wasdouble that of the North-American isolates USA400–0051

and MW2 (RQ=22±0.021 and 16.51±0,04 respectively;Fig. 5c). Finally, protein A gene (spa) transcription was de-creased in the 07–59 (RQ=0.353±0.026) and USA400–0051(RQ=0.333±0.038) isolates compared to the isolates MW2(RQ=0.867±0.036) and 08–28 (RQ=1.027±0.033) (Fig. 5d).It is remarkable that rnaIII and psmα3 expression was in-creased for North American isolates [USA400–0051(RQrnaIII=22±0.082 and RQpsmα3=350±0.116) and MW2(RQrnaII=36±0.094 and RQpsmα3=246±0.064)], comparedto Brazilian isolates [07–59 (RQrnaIII=16±0.18 andRQpsmα3=148±0.043) and 08–28 (RQrnaIII=13±0.187 andRQpsmα3=123±0.11)] (Figs. 5e and 5g). No significant dif-ference was observed in the hla expression for NorthAmerican isolates (USA400–0051: RQ=2.84±0.131 andMW2: RQ=4.21±0.087) and the Brazilian isolate 07–59(RQ=3.5±0.035); however, hla transcripts were undetectablefor the isolate 08–28 (Fig. 5f).

Discussion

A number of strategies were used to compare some virulencetraits of typical HA-MRSA and CA-MRSA isolates of theST1-SCCmecIV lineage that infect hospitalized and commu-nity patients in Brazil and in the United States, respectively.The results showed that both Brazilian and North-AmericanST1 isolates harbored seh and lukDE genes, except for theBrazilian isolate 06–02 that lacked lukDE. However, unlikethe North American isolates [38], none of the Brazilian iso-lates tested carried the genes encoding PVL (lukSFPVL),confirming previous data from our group [30].

Among the large number of toxin genes that can be trans-ferred among S. aureus strains, lukSFPVL is the toxin locusmost consistently present in CA-MRSA strains from differentgeographic origins. Indeed, production of PVL is more com-monly observed among CA- than HA-MRSA isolates [45].The role of PVL in CA-MRSA virulence is relatively contro-versial since lukSFPVL-negative CA-MRSA strains of variousgenetic backgrounds have been reported associated withcommunity-acquired SSTI worldwide [21]. A large-scale ep-idemiological study found no correlation between the pres-ence of lukSFPVL genes and the clinical outcome of compli-cated skin and skin structure infections, such as large abscess-es [46]. However, using rabbit models of CA-MRSA pneu-monia, a significant difference in pathogenesis between iso-genic lukSFPVL-positive and -negative strains was demon-strated [47, 48]. Indeed, it was shown that administration ofPVL was sufficient to cause pneumonia in rabbits [47]. Bothclinical and epidemiological data have provided evidence thatthe severity of some CA-MRSA infections, including necro-tizing fasciitis and pneumonia, seem to be related to PVLproduction [2, 49]. In addition, Jin et al. [50] showed that

Fig. 3 Foreign-body mouse model. Percentage of biofilm accumulatedon catheters implanted in mice and contaminated with the Brazilian ST1-SCCmecIV isolates 08–28 or 07–59 or with the North-American ST1-SCCmecIV isolates USA400–0051 (representative of the USA400 clone)or MW2 (representative of the MW2 clone). The number of viablebacteria recovered from the implanted catheter (CFU/mL) was trans-formed in percentage, considering as reference (100 %) the average valueof the isolate MW2. The difference between American isolates (group 1)and Brazilian isolates (group 2) was significant (p=0.02). The Student t-test was used to compare the two groups

Fig. 4 Autolysis assay. The autolytic activity was based on the absor-bance decay as determined by measuring the optical density (OD) at620 nm. 07–59 and 08–28: Brazilian isolate representatives of the ST1-SCCmecIV lineage. MW2 and USA400–0051: North American isolaterepresentatives for MW2 and USA400 clones, respectively. The differ-ence betweenAmerican isolates (group 1) and Brazilian isolates (group 2)was not significant. The Student t-test was used to compare the twogroups


PVL inhibited osteoblast proliferation, induced osteoblast ap-optosis, and repressed bone formation and mineralization. Amore recent study with a lukSFPVL knock-out, derived fromthe USA300 strain, was carried out using human keratinocytecell cultures and a rabbit skin infection model. The resultsindicated that PVL facilitated the bacteria to escape from

endosomes, replicate intracellularly, and induce apoptosis,thus implicating PVL as an important virulence factor in thenecrotizing skin infections by CA-MRSA [51].

Although apparently convincing, the lack of PVL amongBrazilian ST1 isolates is not sufficient to imply this loss isresponsible for the entry of these ST1 variants in hospital

Fig. 5 Gene expressiondetermined by qRT-PCR usingthe ΔΔCT comparative method.(a) fnbA (b) fnbB (c) sasG (d) spa(e) rnaIII (f) hla and (g) psmα3.07–59 and 08–28: Brazilianisolate representatives of the ST1-SCCmecIV lineage. MW2 andUSA400: North Americanisolates MW2 and USA400–0051, representatives for MW2and USA400 clones, respectively.RN6390B was used as acalibrator in the gene expressionexperiments


settings. However, it is intriguing that these lukSFPVL-nega-tive isolates have been recently causing health-care associatedinfections [30, 31] and that the only community-onset report-ed in Brazil was a case of endocarditis involving a patient thathad a chronic renal hemodialysis catheter implanted in thejugular vein [52].

It is worth mentioning that in addition to the lack of PVL,the genes encoding for the staphylococcal superantigens(SSag) SEA, SEC, and SEK were also absent in the genomeof the Brazilian isolates although all of them were present inthe North American isolates [38]. The role of SSag in foodpoisoning and toxic shock syndrome is well established.However, only more recently has the importance of SSag inother staphylococcal diseases been documented. For example,studies have shown that, when compared with the isogenic seaknock-out, infection with S. aureus Newman strain inducedSEA-dependent stimulation of IFN-γ, IL-12, and chemokineresponses, resulting in increased infiltration of neutrophils intothe liver and a consequent higher number of abscesses, whichcontained an increased amount of viable bacteria [53]. Inanother work it was suggested that the presence of sea along-side the PVL gene in the CA-MRSA ST772 may contribute tothe fitness and virulence of isolates of this lineage, since bothneutrophil proliferation and cytotoxicity were markedly re-duced in a phage-cured wild-type strain [54]. The importanceof SEC in CA-MRSA pathogenicity has also been suggestedby Salgado-Pabón et al. [55], who found that lethal sepsis,infective endocarditis, and kidney infections in rabbits causedby the MW2 strain were critically dependent on the produc-tion of a high amount of the SSAg. In contrast, the isogenicstrain lacking the sec gene, the major SSAg of MW2, showedattenuated virulence. Studies have also demonstrated thatSEK displays many of the biological activities associated withother SSAg, including superantigenicity, pyrogenicity, abilityto enhance the lethal effect of endotoxin, and lethality in arabbit model when administered subcutaneously [56].Regardless of the importance of SSAg for pathogenesis ofCA-MRSA in animal models, it is of note that isolates of avery successful HA-MRSA lineage (ST239), which is glob-ally spread and is responsible for most hospital infectionsworldwide, lack the classical SSAg [57], indicating classicalSSAg are not essential for the pathogenesis of hospital-associated infections.

It was previously suggested that the presence of a largernumber of genetic determinants conferring antimicrobial re-sistance in HA-MRSA, compared to CA-MRSA, can result ina high biological cost, reflected in a reduction of fitness forHA-MRSA [21]. Generally CA-MRSA strains carrySCCmecIV or V, which are smaller than those present intypical HA-MRSA strains, which harbor, more frequently,SCCmecI, II, and III, and contain inserted mobile geneticelements and, in some cases, other antimicrobial resistancegenes [21]. Contrasting this hypothesis, some HA-MRSA

clones, like the pediatric clone, also carry SCCmecIV [35].Even though both Brazilian and North American ST1 isolatescarry SCCmecIV, previous data showed that the majority ofthe Brazilian isolates showed multidrug resistance [30], beingresistant to ciprofloxacin, erythromycin, and clindamycin,while the North American isolates were susceptible to thesedrugs [20, 58]. Antimicrobial resistance is a common exampleof beneficial mutation under strong positive selection. Thesemutations potentially enable the bacterium to survive expo-sure to various antibiotics, but also these mutations typicallyreduce or inactivate pre-existing bacterial activities related toenzymatic, regulatory, or transport systems [59].Consequently, a mutation can be beneficial in one type ofenvironment and toxic in a different environment [60]. Basedon that, it is reasonable to speculate that the expression ofsome genes may be silenced or modified in the Brazilianisolates by the insertion of mobile genetic elements carryingresistant genes, affecting fitness and virulence of theseisolates.

Altogether, this evidence seems to support the hypothesisthat the Brazilian USA400 could be less virulent than theNorth American isolates. The high pathogenic potential ofNorth American isolates was confirmed by the recovery of alarger number of bacterial cells from the animal blood, theseverity of bacteremia signs, and by the significantly increasedproduction of IFN-γ, when compared with Brazilian isolates.In addition to an enhanced potential of virulence in animalmodels, both USA400–0051 and MW2 showed better fitnessin mice in comparison with the Brazilian isolates analysed.Other studies have also demonstrated that CA-MRSA strainswere highly pathogenic to the nematode Caenorhabditiselegans compared to HA-MRSA. However, differently fromthe work presented herein, they tested strains of very distinctgenetic backgrounds [61, 62]. Taken together, these datacorroborate previous studies, indicating that CA-MRSAstrains are more virulent or have better fitness than HA-MRSA strains [21, 61].

The data from the competitive fitness assay showed theunexpected result that the disease signs and IFN-γ productionin animals inoculated with mixed cultures were significantlyattenuated compared with the animals that had individuallybeen infected with North American ST1 isolates, even thoughwe recovered more bacteria from the blood of the infectedanimals inoculated with North American isolates. In otherwords, the pathogenic potential of MW2 and USA400–0051was clearly affected by the mixed infection with the lessvirulent Brazilian ST1 isolates.

Although not well established, the role of genetic back-ground in the pathogenic potential and spread of a specificMRSA clone has already been suggested. It was hypothesizedthat pathoadaptive mutations may improve the virulence ofpathogenic bacteria by adding, modifying or deleting genesrequired for survival in host tissues under specific conditions,


and thus interfere with the pathogenesis of the ancestral bac-teria. Such mutations could involve a gene that directly in-creases fitness or, indirectly, in a gene that would eliminate ormodify a particular characteristic, such as an antivirulencegene (AVG) [63, 64]. An AVG is a gene whose expressionin a pathogen is incompatible with the virulence of thatpathogen. Therefore, an AVG must be inactivated, deleted,or differentially regulated to prevent its expression from inter-fering with bacterial virulence [64]. A classical demonstrationof pathoadaptive mutation affecting an anti-virulence gene isthe loss of lysine decarboxylase (LDC) expression (that blocksthe action of Shigella enterotoxins) in Shigella species thathave evolved from LDC-expressing Escherichia coli [63].Actually, studies on AVG have not been extensivelyresearched in S. aureus and most research has used Gram-negative bacteria as a model [64]. In 2012, Delauné et al. [42]produced a constitutively active form of WalR, a positiveregulator of autolysins in S. aureus. Using a murine model,the WalR phosphomimetic construction showed a significantreduction in virulence compared to the wild-type strain. Theauthors suggested that the virulence attenuation had probablyoccurred due to the increase of the autolytic activity in theWalR phosphomimetic construction, since fragments of mu-rein hydrolase could trigger an early inflammatory responseand thus eliminate the bacteria [42]. Based on that, we decidedto test the autolytic activities of the ST-1 isolates. Althoughlower activity was detected for the Brazilian isolate 07–59,this difference was not statistically significant. The concept ofantagonistic pleiotropy proposes that a gene whose expressionwas advantageous in one environment may be detrimental inanother environment [65]. Thus, theoretically, it is possiblethat an undiscovered antivirulence factor might have beensecreted or shed by the Brazilian isolates, interfering withthe virulence of the North American isolates in animals inoc-ulated with mixed cultures. Additionally, these resultsreflected the importance of the use of animal models in com-parative studies of virulence and fitness.

The fact that the Brazilian ST1 isolates have frequentlybeen collected, not only from bacteremia but also from colo-nization sites [30], may reflect a high ability for colonizationin the nosocomial environment. In fact, when we analysed thetranscriptional expression of virulence factors, we observedthat sasG transcript for the Brazilian isolates was about twicethat of the North American isolates. The role of SasG inbiofilm formation/accumulation and nasal colonization hasalready been demonstrated [4, 13]. Only MW2 expressedfnbB and, corroborating these results, Shukla et al. [38] de-tected fnbB in less than 1 % in USA400 isolates. Comparedwith the other isolates tested, the MW2 strain showed in-creased expression for all genes analysed (except sasG), con-sistent with the fact that this strain is the prototype for thehighly virulent North American CA-MRSA strains related topediatric deaths in 1999 [18].

Concerning the exoproteins, the increase in the psmα3expression for North American, compared with theBrazilian, isolates was consistent with the increase in agr-RNAIII expression for the same isolates. A previous studyalso suggested that, when compared with HA-MRSA strains,the increased expression of psmα3 for CA-MRSA strains wasrelated to a parallel increase of agr-RNAIII, because the agrlocus can positively regulate psmα operon [66]. More impor-tant, studies based on animal models suggest that the psmα3operon could be associated with the severity of skin and softtissues infections and pneumonia [5, 22].

In conclusion, the mechanisms associated with the entry ofa less virulent ST1-SCCmecIV (PVL-negative USA400-relat-ed) variant in Brazilian hospitals were likely to be multifacto-rial, involving the loss of important bacterial toxins, gain ofmultiresistant traits, gene modification leading to differentialgene expression of virulence factors, and probably other im-portant mechanisms to improve nosocomial spread, coloniza-tion, and adaptation to the hospital environment, thus estab-lishing itself as an important pathogen affecting patients withhigh risk for MRSA infections. Ongoing studies involvingwhole genome sequencing of CA-MRSA and HA-MRSA ofST1-SCCmecIV will certainly contribute to the understandingof the evolutionary mechanisms and the driving forces in-volved in the divergence of variants of this MRSA lineage.

Acknowledgments This work was supported in part by ConselhoNacional de Desenvolvimento Científico e Tecnológico (CNPq),Fundação de Amparo à Pesquisa do Rio de Janeiro (FAPERJ),Coordenação de Aperfeiçoamento de Pessoal de Nível Superior(CAPES), and by European Commission's Seventh Framework Pro-gramme (FP7), through the Marie Curie International Research StaffExchange Scheme NANO_GUARD (PIRSES-GA-2010–269,138).

Conflict of interest The authors declare that they have no conflict ofinterests.

Reference

1. Figueiredo AM, Ferreira FA (2014) The multifaceted resources andmicroevolution of the successful human and animal pathogenmethicillin-resistant Staphylococcus aureus. Mem Inst OswaldoCruz 109:265–278

2. Gillet Y, Issartel B, Vanhems P et al (2002) Association betweenStaphylococcus aureus strains carrying gene for Panton-Valentineleukocidin and highly lethal necrotising pneumonia in young immu-nocompetent patients. Lancet 359:753–759

3. Amaral MM, Coelho LR, Flores RP et al (2005) The predominantvariant of the Brazilian epidemic clonal complex of methicillin-resistant Staphylococcus aureus has an enhanced ability to producebiofilm and to adhere to and invade airway epithelial cells. J InfectDis 192:801–810

4. Corrigan RM, Rigby D, Handley P, Foster TJ (2007) The role ofStaphylococcus aureus surface protein SasG in adherence and bio-film formation. Microbiology 153:2435–2446


5. Li M, Cheung GY, Hu J et al (2010) Comparative analysis ofvirulence and toxin expression of global community-associatedMRSA strains. J Infect Dis 202:1866–1876

6. Sanchez CJ Jr, Mende K, BeckiusML et al (2013) Biofilm formationby clinical isolates and the implications in chronic infections. BMCInfect Dis 13:47

7. Surewaard BG, de Haas CJ, Vervoort F et al (2013) Staphylococcalalpha-phenol soluble modulins contribute to neutrophil lysis afterphagocytosis. Cell Microbiol 15:1427–1437

8. Ryu S, Song PI, Seo CH et al (2014) Colonization andinfection of the skin by S. aureus: immune system evasionand the response to cationic antimicrobial peptides. Int J MolSci 15:8753–8772

9. Rasmussen RV, Fowler VG Jr, Skov R, Bruun NE (2011) Futurechallenges and treatment of Staphylococcus aureus bacteremia withemphasis on MRSA. Future Microbiol 6:43–56

10. Wilke M, Grube R (2013) Update on management options in thetreatment of nosocomial and ventilator assisted pneumonia: review ofactual guidelines and economic aspects of therapy. Infect Drug Resist7:1–7

11. O'Neill E, Pozzi C, Houston P et al (2008) A novel Staphylococcusaureus biofilm phenotype mediated by the fibronectin-binding pro-teins, FnBPA and FnBPB. J Bacteriol 190:3835–3850

12. Merino N, Toledo-Arana A, Vergara-Irigaray M et al (2009) ProteinA-mediated multicellular behavior in Staphylococcus aureus. JBacteriol 191:832–843

13. Geoghegan JA, Corrigan RM, Gruszka DT et al (2010) Role ofsurface protein SasG in biofilm formation in Stapylococcus aureus.J Bacteriol 192:5663–5673

14. Houston P, Rowe SE, Pozzi C et al (2011) Essential role for the majorautolysin in the fibronectin-binding protein-mediated Staphylococcusaureus biofilm phenotype. Infect Immun 79:1153–1165

15. Pozzi C, Waters EM, Rudkin JK et al (2012) Methicillin resistancealters the biofilm phenotype and attenuates virulence inStaphylococcus aureus device-associated infections. PLoS Pathog8:e1002626

16. Uhlemann AC, Otto M, Lowy FD, DeLeo FR (2014) Evolution ofcommunity- and healthcare-associated methicillin-resistantStaphylococcus aureus. Infect Genet Evol 21:563–574

17. Rasigade JP, Vandenesch F (2014) Staphylococcus aureus: a patho-gen with still unresolved issues. Infect Genet Evol 21:510–514

18. CDC (Centers for Disease Control and Prevention) (1999) Fourpediatric deaths from community-acquired methicillin-resistantStaphylococcus aureus-Minnesota and North Dakota, 1997–1999.JAMA 282:1123–1125

19. Vandenesch F, Naimi T, Enright MC et al (2003) Community-acquired methicillin-resistant Staphylococcus aureus carryingPanton-Valentine leukocidin genes, worldwide emergence. EmergInfect Dis 9:978–984

20. Como-Sabetti K, Harriman KH, Buck JM et al (2009) Community-associated methicillin-resistant Staphylococcus aureus: trends in caseand isolate characteristics from six years of prospective surveillance.Public Health Rep 124:427–435

21. Otto M (2013) Community-associated MRSA: what makes themspecial? Int J Med Microbiol 303:324–330

22. Wang R, Braughton KR, Kretschmer D et al (2007)Identification of novel cytolytic peptides as key virulencedeterminants for community-associated MRSA. Nat Med 13:1510–1514

23. Li M, Diep BA, Villaruz AE et al (2009) Evolution of virulence inepidemic community-associated methicillin-resistant Staphylococcusaureus. Proc Natl Acad Sci USA 106:5883–5888

24. Thurlow LR, Joshi GS, Clark JR et al (2013) Functional modularityof the arginine catabolic mobile element contributes to the success ofUSA300 methicillin-resistant Staphylococcus aureus. Cell HostMicrobe 13:100–107

25. Stratén AM, Frei R, Adler H et al (2009) Emergence of SCCmec typeIV as the most common type of methicillin-resistant Staphylococcusaureus in a university hospital. Infection 37:44–48

26. Scribel LV, Silva-Carvalho MC, Souza RR et al (2009)Clinical and molecular epidemiology of methicillin-resistantStaphylococcus aureus carrying SCCmecIV in a universityhospital in Porto Alegre, Brazil. Diagn Microbiol Infect Dis65:457–461

27. Machuca MA, Sosa LM, González CI (2013) Molecular typing andvirulence characteristic of methicillin-resistant Staphylococcusaureus isolates from pediatric patients in Bucaramanga, Colombia.PLoS One 8:e73434

28. Diekema DJ, Richter SS, Heilmann KP et al (2014) Continuedemergence of USA300 methicillin-resistant Staphylococcus aureusin the United States: results from a nationwide surveillance study.Infect Control Hosp Epidemiol 35:285–292

29. El-Mahdy TS, El-Ahmady M, Goering RV (2014) Molecular char-acterization of methicillin-resistant Staphylococcus aureus isolatedover a 2-year period in a Qatari hospital from multinational patients.Clin Microbiol Infect 20:169–173

30. Silva-Carvalho MC, Bonelli RR, Souza RR et al (2009) Emergenceof multiresistant variants of the community-acquired methicillin-re-sistant Staphylococcus aureus lineage ST1-SCCmecIV in two hospi-tals in Rio de Janeiro, Brazil. DiagnMicrobiol Infect Dis 65:300–305

31. Cavalcante FS, Schuenck RP, Ferreira DC et al (2014) Methicillin-resistant Staphylococcus aureus: spread of specific lineages amongpatients in different wards at a Brazilian teaching hospital. J HospInfect 86:151–154

32. Pacheco AB, Guth BE, Soares KC et al (1997) Random amplificationof polymorphic DNA reveals serotype-specific clonal clusters amongenterotoxigenic Escherichia coli strains isolated from humans. J ClinMicrobiol 35:1521–1525

33. Oliveira DC, de Lencastre H (2002)Multiplex PCR strategy for rapididentification os structural types and variants of the mec element inmethicillin-resistant Staphylococccus aureus. Antimicrob AgentsChemother 46:2155–2161

34. Jarraud S, Mougel C, Thioulouse J et al (2002) Relationships be-tween Staphylococcus aureus genetic background, virulence factors,agr groups (alleles), and human disease. Infect Immun 70:631–641

35. DeMiranda OP, CarvalhoMCS, Ribeiro A et al (2007) Emergence inBrazil of methicillin-resistant Staphylococcus aureus isolates carry-ing SCCmec IV that are related genetically to the USA800 clone. ClinMicrobiol Infect 13:1165–1172

36. Ribeiro A, Coronado AZ, Silva-Carvalho MC et al (2007) Detectionand characterization of international community-acquired infectionsby methicillin-resistant Staphylococcus aureus clones in Rio deJaneiro and Porto Alegre cities causing both community- andhospital-associated diseases. Diagn Microbiol Infect Dis 59:339–345

37. Lindsay JA, Ruzin A, Ross HF et al (1998) The gene for toxic shocktoxin is carried by a family of mobile pathogenicity islands inStaphylococcus aureus. Mol Microbiol 29:527–543

38. Shukla SK, Karow ME, Brady JM et al (2010) Virulence genes andgenotypic associations in nasal carriage, community-associatedmethicillin-susceptible and methicillin-resistant USA400Staphylococcus aureus isolates. J Clin Microbiol 48:3582–3592

39. Acred P, Hennessey TD, Macarthur-Clark JA et al (1994) Guidelinesfor the welfare of animals in rodent protection tests. A Report of theRodent Protection Test Working Party. Lab Anim 28:13–18

40. Olfert E, Bhasin J, Latt R et al (1998) The CCAC guidelines on:choosing an appropriate endpoint in experiments using animals forresearch, teaching and testing. Canadian Council on animal care.1998. Available at: http://www.ccac.ca/Documents/Standards/Guidelines/Appropriate_endpoint.pdf. Accessed 13 December 2013

41. Ferreira FA, Souza RR, Bonelli RR et al (2012) Comparison ofin vitro and in vivo systems to study ica-independent staphylococcusaureus biofilm. J Microbiol Methods 88:393–398


http://www.ccac.ca/Documents/Standards/Guidelines/Appropriate_endpoint.pdf

http://www.ccac.ca/Documents/Standards/Guidelines/Appropriate_endpoint.pdf

42. Delauné A, Dubrac S, Blanchet C et al (2012) The WalKR systemcontrols major virulence genes and is involved in triggering the hostinflammatory response. Infect Immun 80:3438–3453

43. De Jonge BL, De Lencastre H, Tomasz A (1991) Suppression ofautolysis and cell wall turnover in heterogeneous Tn551mutants of amethicillin-resistant Staphylococcus aureus strain. J Bacteriol 173:1105–1110

44. Ferreira FA, Souza RR, De Sousa MB et al (2013) Impact of agrdysfunction on virulence profiles and infections associated with anovel methicillin-resistant Staphylococcus aureus (MRSA) variant ofthe lineage ST1-SCCmecIV. BMC Microbiol 13:93

45. Watkins RR, David MZ, Salata RA (2012) Current concepts on thevirulence mechanisms of methicillin-resistant Staphylococcusaureus. J Med Microbiol 61:1179–1193

46. Bae IG, Tonthat GT, Stryjewski ME et al (2009) Presence of genesencoding the panton-valentine leukocidin exotoxin is not the primarydeterminant of outcome in patients with complicated skin and skinstructure infections due to methicillin-resistant Staphylococcus aureus:results of a multinational trial. J Clin Microbiol 47:3952–3957

47. Labandeira-Rey M, Couzon F, Boisset S et al (2007) Staphylococcusaureus Panton-Valentine leukocidin causes necrotizing pneumonia.Science 315:1130–1133

48. Diep BA, Chan L, Tattevin P et al (2010) Polymorphonuclear leuko-cytes mediate Staphylococcus aureus Panton-Valentine leukocidin-induced lung inflammation and injury. Proc Natl Acad Sci USA 107:5587–5592

49. Lina G, Piémont Y, Godail-Gamot F et al (1999) Involvement ofPanton-Valentine leukocidin-producing Staphylococcus aureus in pri-mary skin infections and pneumonia. Clin Infect Dis 29:1128–1132

50. Jin T, Zhu YL, Li J et al (2013) Staphylococcal protein A, Panton-Valentine leukocidin and coagulase aggravate the bone loss and bonedestruction in osteomyelitis. Cell Physiol Biochem 32:322–333

51. Chi CY, Lin CC, Liao IC et al (2014) Panton-Valentine leukocidinfacilitates the escape of Staphylococcus aureus from humankeratinocyte endosomes and induces apoptosis. J Infect Dis 209:224–235

52. Damasco PV, Cavalcante FS, Chamon RC et al (2013) The first casereport of non-nosocomial healthcare-associated infective endocarditisdue to methicillin-resistant Staphylococcus aureusUSA400 in Rio deJaneiro, Brazil. Infection 41:851–854

53. Xu SX, Gilmore KJ, Szabo PA et al (2014) Superantigens subvert theneutrophil response to promote abscess formation and enhanceStaphylococcus aureus survival in vivo. Infect Immun 82(9):3588–3598

54. Prabhakara S, Khedkar S, Shambat SM et al (2013) Genome se-quencing unveils a novel sea enterotoxin-carrying PVL phage inStaphylococcus aureus ST772 from India. PLoS One 8:e60013

55. Salgado-Pabón W, Breshears L, Spaulding AR et al (2013)Superantigens are critical for Staphylococcus aureus Infective endo-carditis, sepsis, and acute kidney injury. MBio 4(4):e00494–13

56. Orwin PM, Leung DY, Donahue HL et al (2001) Biochemical andbiological properties of Staphylococcal enterotoxin K. Infect Immun69:360–366

57. Costa MO, Beltrame CO, Ferreira FA et al (2013) Complete genomesequence of a variant of the methicillin-resistant Staphylococcusaureus ST239 Lineage, strain BMB9393, displaying superior abilityto accumulate ica-independent biofilm. Genome Announc 1:e00576–13

58. Popovich KJ, Weinstein RA, Hota B (2008) Are community-associated methicillin-resistant Staphylococcus aureus (MRSA)strains replacing traditional nosocomial MRSA strains? Clin InfectDis 46:787–794

59. Anderson KL (2005) Is bacterial resistance to antibiotics an appro-priate example of evolutionary change? Creation Res Soc Q 41:318–326

60. Schneider D, Lenski RE (2004) Dynamics of insertion sequenceelements during experimental evolution of bacteria. Res Microbiol155:319–327

61. Day SR, Moore CM, Kundzins JR, Sifri CD (2012) Community-associated and healthcare-associated methicillin-resistantStaphylococcus aureus virulence toward Caenorhabditis eleganscompared. Virulence 3:576–582

62. Wu K, Zhang K, McClure J et al (2013) A correlative analysis ofepidemiologic and molecular characteristics of methicillin-resistantStaphylococcus aureus clones from diverse geographic locationswith virulence measured by a Caenorhabditis elegans hostmodel.Eur J Clin Microbiol Infect Dis 32:33–42

63. Day WA Jr, Fernández RE, Maurelli AT (2001) Pathoadaptive mu-tations that enhance virulence: genetic organization of the cadAregions of Shigella spp. Infect Immun 69:7471–7480

64. Bliven KA, Maurelli AT (2012) Antivirulence genes: insights intopathogen evolution through gene loss. Infect Immun 80:4061–4070

65. Deng W, Burland V, Plunkett G 3RD et al (2002) Genome sequenceof Yersinia pestis KIM. J Bacteriol 184(16):4601–4611

66. Queck SY, Jameson-Lee M, Villaruz AE et al (2008) RNAIII-independent target gene control by the agr quorum-sensing system:insight into the evolution of virulence regulation in Staphylococcusaureus. Mol Cell 32:150–158
