Multiresistant ST59-SCCmec IV-t437 clone with strong biofilm-forming capacity was identified predominantly in MRSA isolated from Chinese children

Background This study aimed to investigate the clinical and molecular epidemiology and biofilm formation of Staphylococcus aureus (SA) isolated from pediatricians in China. Methods SA strains were isolated from Beijing Children’s hospital from February 2016 to January 2017. Isolates were typed by multilocus sequence typing (MLST), spa and SCCmec typing (for Methicillin-resistant SA [MRSA] only). Antimicrobial susceptibility testing was performed by agar dilution method except sulphamethoxazole/trimethoprim (E-test method). Biofilm formation and biofilm associated genes were detected. Results Totally 104 children (41 females and 63 males; median age, 5.2 months) were enrolled in this study, in which 60 patients suffered from MRSA infection. Among the 104 cases, 54.8% were categorized as community associated SA (CA-SA) infections. The children under 3 years were more likely to occur CA-SA infections compared with older ones (P = 0.0131). ST59-SCCmec IV-t437 (61.7%) was the most prevalent genotype of MRSA, and ST22-t309 (18.2%), ST5-t002 (9.1%), ST6-t701 (9.1%), ST188-t189 (9.1%) were the top four genotypes of methicillin-sensitive SA (MSSA). All the present isolates were susceptible to linezolid, vancomycin, trimethoprim-sulfamethoxazole, mupirocin, tigecyclin, fusidic acid. No erythromycin-susceptible isolate was determined, and only a few isolates (3.8%) were identified as susceptible to penicillin. Multi-drug resistant isolates were reponsible for 83.8% of the ST59-SCCmec IV-t437 isolates. The isolates with strong biofilm formation were found in 85% of MRSA and 53.2% of MSSA, and in 88.7% of ST59-SCCmec IV-t437 isolates. Biofilm formation ability varied not only between MRSA and MSSA (P = 0.0053), but also greatly among different genotypes (P < 0.0001). The prevalence of the biofilm associated genes among ST59-SCCmec IV-t437 clone was: icaA (100.0%), icaD (97.3%), fnbpA (100.0%), fnbpB (0), clfA (100%), clfB (100%), cna (2.7%), bbp (0), ebpS (88.5%), sdrC (78.4%), sdrD (5.4%), and sdrE (94.5%). Conclusions These results indicated strong homology of the MRSA stains isolated from Chinese children, which was caused by spread of multiresistant ST59-SCCmec IV-t437 clone with strong biofilm formation ability. The MSSA strains, in contrast, were very heterogeneity, half of which could produce biofilm strongly. Electronic supplementary material The online version of this article (10.1186/s12879-017-2833-7) contains supplementary material, which is available to authorized users.


Background
Staphylococcus aureus (SA) is an important Grampositive pathogen which can cause diseases ranging from minor to potentially life-threatening community associated and hospital-associated infections, such as skin and soft tissue infections (SSTIs), bacteremia, pneumonia, osteomyelitis and endocarditis [1]. SA also has the ability to form biofilm in biological and indwelling medical devices surfaces [2]. The successful eradication of SA infection in patients become difficult once biofilm formed, since biofilm can protect SA from the damage of antibiotics, host immune system, and so on [2]. In addition, Savage et al. found that SA biofilms could promote horizontal spread of antibiotic resistance determinants, which were mainly through increasing the frequency of plasmid transfer events by both conjugation and mobilization [3]. Thus, biofilm forming ability of SA has drawn considerable interest from researchers over the past decades.
Biofilm formation can be divided into at least three major stages: initial attachment, biofilm maturation, and dispersal [4]. Initial attachment is a crucial stage of transition from an individual planktonic cell to a biofilm. Attachment is mediated mainly through a family of surface proteins, referred to as microbial surface components recognizing adhesive matrix molecules (MSCRAMMs), such as clumping factor A (ClfA), clumping factor B (ClfB), elastin binding protein (EbpS), serine-aspartate repeat protein C (SdrC), SdrD, SdrE, bone sialoproteinbinding protein (Bbp, isoform of SdrE), fibronectinbinding proteins A (FnBPA) and B (FnBPB), collagen adhesin (Cna) [5]. During the stages of biofilm maturation, multilayered biofilm formation is related to the production of polysaccharide intercellular adhesin (PIA), which is synthesized by the enzymes encoded by the intercellular adhesion (ica) operon, mainly including icaR (intercellular adhesion regulator) and icaA, B, C, and D [6]. Among these genes, icaA and icaD are most extensively studied and play a more important role in the biofilm formation than other genes [7].
Although many studies have reported the phenotypic and genotypic basis for biofilm production in SA clinical strains isolated from different infectious diseases and different countries [8][9][10], little is known regarding the biofilm formation ability of SA clinical strains isolated from Chinese, especially children. According to our knowledge, only the prevalence of adhesion genes was ever reported among SA strains isolated from children in china, but these studies didn't assess the biofilm formation ability of bacteria [11,12].
Considering the adverse effect of biofilm formation on SA mediated infectious diseases [2,3] as well as shifts of major clones in a given region over time [13], the present study aimed to investigate the genotype characteristics, antimicrobial susceptibility, biofilm-forming ability and the prevalence of biofilm associated genes among SA clinical strains, which were collected from the biggest tertiary-care teaching hospital for children in Beijing, China.

Bacteria isolates
This study was performed in Beijing Children's Hospital in China. It was reviewed and approved by the Ethics Committee of Beijing Children's Hospital affiliated to Capital Medical University. No ethical problems existed in this study.
Once SA was detected from Bacteriology Laboratory in our hospital, the isolates were collected and stored, but bacteria isolated from throat swab, vaginal secretions, and defecate were not included. Only one strain was included from each patient. A total of 209 isolates were collected during the studied period. Of the 209 isolates, 19 were collected from outpatients (lack of epidemiological information), 86 were identified as colonizing strains, and only 104 were considered to have caused clinical infections. Thus, the 104 pathogenic bacteria were selected for further study. These strains were isolated from several clinical sources, including respiratory tract (27 form sputum, 15 from bronchial alveolar lavage fluid), skin and soft tissue (11 from pus, 8 from secretions, 13 from secretions of omphalitis, 5 from eye secretions), sterile sites (20 from blood, 2 from joint effusion and 2 from pleural effusion), and pipe end (1 isolate). SA infections were categorized as hospital associated (HA) or community associated (CA) according to the definitions established previously [14].
The identification of the SA isolates was performed by colony morphological characteristics, coagulase test, and nuc gene detection. The MRSA isolates were screened with cefoxitin discs and were confirmed by detecting the carrying situation of the mecA gene by polymerase chain reaction (PCR) [11]. All strains were stored at −80°C until use.

Extraction of genomic DNA
A typical colony was cultivated on blood agar at 37°C for 24 h. Bacteria genomic DNA was extracted using Nucleic Acids Isolation & Purification kit (Saibaisheng gene technology Ltd., China) according to the manufacturer's instructions.

Molecular genotyping analysis
MLST was performed as described by Enright et al. previously [15]. The seven housekeeping gene (arcC, aroE, glpF, gmk, pta, tpi and yqiL) sequences were compared with known alleles from the MLST database (http://saureus.mlst.net/), and the allelic profiles (allele numbers) and ST types were determined based on the database.
The the polymorphic X region of spa gene was amplified as previously described [16], and the spa type was determined by submitting the sequencing data to the SA spa type database (http://spaserver.ridom.de).

Antimicrobial susceptibility testing
The susceptibility of the isolates against penicilin G, cefuroxime, gentamicin, rifampin, ciprofloxacin, clindamycin, erythromycin, chloramphenicol, tetracycline, linezolid, vancomycin, mupirocin, tigecycline and fusidic acid were tested with the agar dilution methods. Susceptibility to sulphamethoxazole/trimethoprim was determined by E-Test method. Minimum inhibitory concentration for tigecycline and fusidic acid were interpreted using European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoints for Staphylococcus spp. [18]. The MIC of other antibiotics were interpreted using the Clinical and Laboratory Standards Institute (CLSI) breakpoints for Staphylococcus spp. [19]. ATCC29213 was used as the quality control. For MRSA, multi-drug resistance (MDR) was defined as isolates resistant to ≥ 3 classes of non-β-lactam antimicrobials [20], whereas MDR was defined as isolates resistant to ≥ 3 classes of antibiotics including β-lactam antibiotics for Methicillin-sensitive SA (MSSA).

Biofilm formation assays
Biofilm forming ability was assessed using tissue culture plate method (TCP), as described by Xu et al. [28], with slight modification. Briefly, All MRSA strains were grown in tryptic soy broth (TSB) (OXOID, USA) containing 0.25% glucose overnight at 37°C. Bacterial concentrations were adjusted to a concentration of 0.5 on the McFarland scale (~1.5 × 10 8 CFU/mL), and diluted in TSB containing 0.25% glucose to a final concentration of 10 6 CFU/ml. The biofilm assay was performed in 96-well flat-bottom plates (Corning, USA) at 37°C for 48 h. Because 48 h of growth has been optimal for SA, and biofilms are sufficiently mature at this time point [29,30]. Subsequently, wells were washed twice 0.9% sodium chloride, and fixed by methanol for 15 min. After air dried, wells were stained with 0.1% crystal violet for 5 min. The microtiter plate was then rinsed with PBS and air dried, and the stained biofilm was resuspended for quantification in 33% glacial acetic acid. The optical density (OD) of each stained well was measured at 590 nm using an CLARIOstar Microplate reader (BMG LABTECH, Germany). Each isolate was tested in three repetition. The negative control wells contained broth only.
The cut-off OD value (ODc) was defined as the arithmetic mean of the absorbance of negative controls with three times addition of standard deviation. The following classification was applied for the determination of biofilm formation: no biofilm production (OD ≤ ODc), weak biofilm production (WBF, ODc < OD ≤ 2ODc, WBF), moderate biofilm production (2ODc < OD ≤ ODc, MBF), and strong biofilm production (4ODc < OD, SBF).

Statistical analysis
SAS JMP Statistical Discovery v11.0 was used for statistical analysis. Categorical variables were analyzed using Chi-squared (χ2) test or Fisher's exact test. The OD values used to assess the biofilm formation didn't coincided with normal distribution in any cases, so Wilcoxon rank sum test was used to compare the biofilm formation ability between two groups. In addition, when compared the biofilm formation ability among three or more groups, Kruskal-Wallis test followed by Steel-Dwass test were used. P < 0.05 was considered statistically significant.

Clinical characteristics
A total of 104 children (41 females and 63 males; median age, 5.2 months) were enrolled in this study, and 60 patients suffered from MRSA infection. Their clinical characteristics were shown in Table 1. Approximately 74.0% (74/104) of the patients were less than 3 years old. By CDC criteria, 54.8% (57/104) were categorized as community associated infections, and 45.2% (47/104) were categorized as hospital associated infections. The modes of acquisition (hospital vs. community) were similar among MRSA and MSSA (Table 1) and different sites of infections (Fig. 1a). Children under 3 years were more likely to occur community associated infections compared with older children (P = 0.0131) (Fig. 1b). SSTIs (35.58%, 37/104) and pneumonia (42.3%, 44/104)
Combined analysis of MLST, spa and SCCmec types (for MRSA only) indicated that ST59-SCCmec IV-t437 (61.7%, 37/60) was the most prevalent clone among MRSA isolates. The top 4 genotypes of MSSA were

Antimicrobial resistance
Antimicrobial susceptibility test results were shown in Table 2. All isolates in this study were susceptible to trimethoprim-sulfamethoxazole, linezolid, vancomycin, mupirocin, tigecyclin, fusidic acid. Only 2 isolates were resistance to rifampin. All isolates were non-susceptible to erythromycin, and nearly all isolates (96.2%, 100/104) were non-susceptible to penicillin. The non-susceptibility rates to cefuroxime, clindamycin, and tetracycline were significantly higher in MRSA than MSSA (P < 0.05). However, the non-susceptibility rate to gentamicin was significantly lower in MRSA than MSSA (P = 0.0069). About 76.7% (46/60) of MRSA and 77.3% (34/44) of MSSA were MDR isolates.

Biofilm production
We further compared the biofilm formation ability of SA isolated from patients with different infections (Fig. 3c). Strains isolated from patients with SSTIs could product much higher biofilm than strains isolated from patients with pneumonia (P = 0.0036) and sterile site infections (P = 0.0281). Table 4 showed the prevalence of biofilm associated genes among MRSA and MSSA isolates. For MRSA, all isolates were positive for icaA, fnbpA, clfA, clfB and only one strain was icaD negative. The prevalence rates for fnbpB, cna and bbp were very low, their carrying rates were 3.3%, 10.0% and 1.7%, respectively. The prevalence For MSSA, all isolates tested were positive for icaA, icaD, clfA, clfB. Only two strains didn't harbor ebpS. The prevalence rate of fnbpA, fnbpB, cna, bbp, fib, sdrC, sdrD,sdrE ranged from 9.1% to 86.4%. All isolates of ST59-SCCmec IV-t437 MRSA clone didn't harbour fnbpB and bbp genes. Statistically significant differences of fnbA, fnbB, cna, sdrD between MRSA and MSSA were found (P < 0.05). However, only fnbpA was more likely to be presented in MRSA, other significantly different genes were more likely to be present in MSSA.

Discussion
This study provided important information on the clinical and molecular epidemiology and biofilm formation ability of SA isolated from pediatricians in China. To our knowledge, this is the first study to report the biofilm production of SA clinical strains isolated from Chinese children.
We found that SA infections were more inclined to affect infants. Children under 3 years of age accounted for 74.0% of the total cases with SA infections in the present study. This result was similar to the study reported by Iwamoto et al., which showed that 39.0% of the total 876 pediatric cases were among infants [31]. Furthermore, Suryadevara et al. estimated populationbased incidence of invasive SA infection in children <19 years of age (1996 to 2006), and found that the incidence of MSSA and MRSA infections was highest in children 0 to 4 years of age [32]. The reason why infants are more likely to be infected may be due to that infants are frequently colonized by SA, and the carriage of SA was highest in the first three months of life (25.4%) [33], whereas nasal carriage of SA is an important risk factor for SA infection [34]. In addition, infants were more likely to occur community associated infections compared with older children in our study.
Our results revealed that ST59-SCCmec IV-t437 was the most prevalent clone both in CA-and HA-MRSA isolates. In this study, the prevalence rate of MRSA ST59 clone (76.7%) was much higher than previously reported (35.8%, MRSA strains were isolated from Chinese children from 2004 to 2012) [11]. What's more, we need to note that although ST59 was the predominant clone in the MRSA isolates, ST239 clone also accounted for 22.0% in the previous research [11]. However, ST239 was disappeared in our study. ST59 and ST239 were usually community associated and hospital associated clones in China, respectively [35,36]. The increasing prevalence rate of ST59 and the disappear of ST239 suggested the significant penetration of CA-MRSA clone into hospitals, and even replaced HA-MRSA clone. Indeed, many studies have indicated that CA-MRSA clones are beginning to replace HA-MRSA clones as the predominant cause of hospital infections around the world, such as USA, Greece, Denmark, Uruguay, Korea, Tunisia, and Algeria [37].This maybe due to that CA-MRSA clone carries a shorter SCCmec (usually type IV and V) than HA-MRSA clone (usually type I, II and III), which believed to minimized the fitness cost [38]. In addition, pvl may be involved because CA-MRSA clones were more likely to carry pvl, but pvl negative CA-MRSA strains can also cause outbreaks in healthcare settings [39]. Further studies are still needed on this issue.
For MSSA clinical strains, there were diverse genotypes and no dominant clone was identified. The top three MLST types were ST22 (20.5%), ST5 (11.4%) and ST398 (11.4%), which differed from those detected in other regions, such as Europe and Australia [40,41]. In addition, the most frequent  MLST types of MSSA clinical isolates in this study were also different from previous research which showed that ST88, ST25, ST7, ST2155, and ST188 were the top five MLST types for MSSA strains isolated from Chinese children [42]. These results indicate that the molecular characteristics of MSSA may also have regional characteristics, and the common genotypes are also changing with time. Therefore, molecular epidemiological investigations of MSSA strains are also very important, and have great significance to control MSSA clinical infection in a given region.
CA-MRSA clones are usually considered susceptible to most antibiotics other than methicillin and beta-lactams [43]. But in our study, ST59-SCCmec IV-t437 clone, the most prevalent clone both in CA-MRSA and HA-MRSA isolates, showed relative high resistant rates to erythromycin, clindamycin, tetracycline, chloramphenicol, and even ciprofloxacin. What's more, the MDR rate of this clone had reached 83.8%. These results were consistent with a previous research which demonstrated that resistance to non-β-lactams, especially to clindamycin, was high in CA-MRSA isolates from Chinese children, and  the MDR rate for ST59 clone was 67.9% [44]. Multiresistant CA-MRSA clone has also been reported in other countries. For example, CA-MRSA USA300 isolates are becoming more resistant to a variety of antibiotics, including erythromycin, levofloxacin, mupirocin and tetracycline, and have spread to Europe, South America and Australia [45]. This phenomenon should arouse the attention of clinicians when making treatment protocols for patients potentially infected with these bacteria. In addition, MSSA isolates were more susceptible to cefuroxime, clindamycin, and tetracycline than MRSA isolates. But the resistance rate of MSSA to penicillin and erythromycin reached also nearly 100%, which indicated that penicillin and erythromycin may not be suitable for Chinese children with SA infection. Furthermore, our data demonstrated that the biofilm formation abilities of SA strains are generally high: 83.3% of MRSA and 54.5% of MSSA showed SBF. The generally high biofilm production of SA strains obtained from Chinese children call for greater attention in the treatment of SA infectious diseases, especially indwelling medical device infection. We also found that MRSA strains could produce significantly higher biofilm than MSSA strains. This result was consistent with Kwon et al. describing that the rate of biofilm positivity in MRSA strains was significantly higher than in MSSA strains (37.9% vs. 14.3%, P < 0.05) [46]. The morphological studies of Jones et al. also indicated that the MRSA biofilm was thicker than the MSSA biofilm [47]. However, many other studies failed to establish a link between oxacillin resistance and biofilm formation ability [48][49][50]. Different results of these studies may be due to the following reasons. Firstly, the predominant clone of MRSA has regional characteristics, and MRSA strains can express either low level heterogeneous resistance or high-level, homogeneous resistance to   methicillin [51]. These phenomena make the relationship between methicillin resistance and biofilm formation become more complicated. Secondly, the mechanisms of biofilm formation of MRSA and MSSA are different, biofilm formation ability of MRSA and MSSA maybe influenced by the expression level of their respective regulatory mechanism. Researches have shown that MSSA strains form PIA-mediated biofilms whereas MRSA strains form biofilms independent of PIA, but requiring surface proteins and firmly regulated by accessory gene regulator (agr) system [51]. Further studies are still needed to explore the relationship between methicillin resistance and biofilm formation ability. In addition, our results showed that a correlation between the clonal lineage and biofilm formation might be existed. What need to be stressed was that 83.8% of the ST59-SCCmec IV-t437 clone, the most prevalent clone of MRSA, showed SBF. The ability of ST59-SCCmec IV-t437 clone to form strong biofilm may contribute to its dominance and multi-drug resistance in China. What's more, all MSSA strains belonging to ST188-t189 showed especially strong biofim formation ability. Although we found that MRSA could produce significantly higher biofilm than MSSA, the extremely high biofilm formation ability of MSSA ST188-t189 isolates indicated that biofilm formation might be more closely related with clonal lineage. The relationship between clonal lineage and biofilm formation has been supported by several other studies. Naicker et al. [50] found that MLST CC5 might be associated with high biofilm formation. Croes et al. [52] also demonstrated that strains associated with MLST CC8 were markedly more often classified as strong biofilm former. Furthermore, Atshan et al. [53] found that isolates belonging to similar spa, SCCmec, and MLST types had similar abilities to produce biofilms, and isolates of different spa types showed high variation in their ability to produce biofilms. These researches, including ours, suggest that clonal lineage might be good predictors of biofilm production.
To understand the molecular mechanism of SA biofilm formation, we detected the frequency of 12 selected genes in biofilm formation. In the present study, all isolates harbored icaA, clfA and clfB, and only one strain didn't harbor icaD. Similar to our study, several other studies also reported a high prevalence rate of these genes [54,55]. A comparative analysis between MRSA and MSSA isolates regarding the presence of all tested genes showed that fnbpA were more inclined to be present in MRSA, whereas fnbpB, cna, sdrD were more likely to be present in MSSA. However, a previous study didn't find any correction between methicillin resistance and the prevalence of biofilm associated genes [51]. This discrepancy may be due to that specific clonal complexes of SA may contain a unique combination of surface-associated and regulatory genes [56], and the distribution of clonal lineage have regional characteristics. Further researches are still needed to evaluate the expression of these genes in SA.

Conclusions
In summary, our results indicated strong homology of the MRSA stains isolated from Chinese children, in which multiresistant ST59-SCCmec IV-t437 clone with strong biofilm formation ability was determined predominantly. The MSSA strains, in contrast, were very heterogeneity. The generally high MDR rate and biofilm production of SA in this study should arouse the attention of pediatrician in China. In addition, significant differences were found between MSSA and MRSA regarding biofilm formation and several biofilm associated genes (fnbA, fnbB, cna, sdrD), and an correlation between clonal lineage and biofilm formation might also be existed. Investigation of biofilm production and related molecular mechanisms of SA will ultimately promote the treatment of biofilm mediated infections. analysed the data and proofed the article. JZ, WZ, HX, HZ and WL collected and identified S. aureus clinical strains, and analysed the data. All authors read and approved the final manuscript.

Ethics approval and consent to participate
The study mainly used the bacterial isolates from the biological specimens obtained during patients' clinical diagnosis and management, and had no any threat to the subjects' rights and health. The applications for exemption of written informed content and ethical review had been approved by the Ethics Committee of Beijing Children's Hospital Affiliated to Capital Medical University according to national regulations. Thus, only verbal consent was obtained from the patient's legal guardian.

Consent for publication
Not applicable.