A total of 371 Staphylococcus spp. isolated from positive blood cultures between May 2011 and May 2012 were provided by the Laboratory of Microbiology at the Teaching Hospital of Botucatu Medical School, UNESP - Univ Estadual Paulista (City of Botucatu, São Paulo State, Brazil).

The isolates included in the study were obtained on the day following growth detection in a Bactec 9120 system and after confirmation of containing probable Staphylococcus spp. by Gram staining (Gram-positive cocci) and positive catalase tests.

The blood culture bottles were disinfected with 70% alcohol and 3mL of blood was collected with a 5mL syringe and 70×25-mm needle into two microtubes (1.5mL per microtube). One microtube was used for seeding onto blood agar for phenotypic identification and the other microtube was used for extraction of bacterial DNA for multiplex PCR.

Phenotypic identification was performed at the Laboratory of Bacteriology, Department of Microbiology and Immunology, Botucatu Biosciences Institute, UNESP. The samples were seeded into blood agar and incubated for 24h at 37°C. After isolation from blood agar, the bacteria were submitted to Gram staining for confirmation of morphological staining characteristics, catalase test for characterization of the genus Staphylococcus, and coagulase test for differentiation between S. aureus and CoNS.

Before DNA extraction, the samples were treated to remove interfering elements in the blood culture. The sample was centrifuged at 850×g for 2min.9 The supernatant was transferred to another microtube and again centrifuged at 11,000×g for 1min. The supernatant was carefully discarded, leaving only the pellet. Five hundred microliters of lysozyme buffer (1M EDTA, 0.1M NaCl, 10mM Tris, pH 8.0, and 5% Triton X-100) was then added to the pellet and the mixture homogenized in a vortex. Next, 800μL benzyl alcohol (Sigma) was added, the mixture homogenized in a vortex, and the microtube centrifuged at 7000×g for 5min.10 Two phases formed in the microtube, the lower phase contained the alcohol with interfering substances (inhibitors) and the upper phase the bacteria.

For DNA extraction, 320μL of the liquid was removed from the upper phase and transferred to another microtube and 10μL lysozyme was added (10mg lysozyme in 1mL 10mM Tris–HCl). The microtube was homogenized in a vortex and incubated at room temperature for 15min, homogenizing the microtube at five-minute intervals. After this period, 10μL proteinase K was added and the mixture was incubated for 15min at 56°C, homogenizing the microtube at five-minute intervals in a vortex. The liquid (330μL) was transferred to a silica column provided by the Illustra extraction kit (GE Healthcare). This column was placed inside a tube and centrifuged at 11,000×g for 1min. Lysis solution (500μL) was added and the column centrifuged at 11,000×g for 1min. Next, 500μL washing solution was added and the column was again centrifuged at 11,000×g for 3min. The tube was discarded and the column was transferred to a microtube. Elution solution (200μL) pre-warmed to 70°C in water bath was then added to the column. After incubation at room temperature for 1min, the microtube was centrifuged at 11,000×g for 1min and the DNA obtained was stored in a freezer at −20°C.

The isolates were submitted to in vitro susceptibility testing by the disk diffusion method on Mueller-Hinton agar plates at the Laboratory of Microbiology, Department of Internal Medicine, University Hospital, Botucatu Medical School, UNESP, according to the Clinical and Laboratory Standards Institute criteria.11

First, PCR was standardized using a range of annealing temperatures to establish the adequate annealing temperature for all primers (Table 1). Amplification was performed in an Eppendorf thermocycler. The amplification conditions were: 92°C for 3min, followed by 30 cycles of DNA denaturation at 92°C for 1min, annealing at 56°C for 1min, and extension at 72°C for 3min. The final reaction volume was 35μL containing 10.2μL autoclaved Milli-Q water, 3.5μL deoxyribonucleotide triphosphates (dNTP) (Life Technologies), 0.8μL Taq DNA polymerase (Biotools), 2.5μL MgCl2-free buffer (Biotools), 2.5μL MgCl2 (Biotools), 0.14μM of each 16S primer, 0.86μM of each MRS primer, 0.35μM of each SAU primer, 0.57μM of each COA primer, and 3.5μL of bacterial DNA. The amplification products were analyzed by electrophoresis on 3% agarose gel at 70V. The following controls were included in all amplification reactions: ATCC 33591 (mecA-positive S. aureus), ATCC 25923 (mecA-negative S. aureus), ATCC 12228 (mecA-negative S. epidermidis), and negative control (water).

The sensitivity of multiplex PCR was evaluated by amplifying serial dilutions of S. aureus and CoNS control strains, starting with a dilution of 10−1 (corresponding to 107 CFU/mL) until 10−7 (10CFU/mL) based on a 0.5 McFarland standard (1.5×108CFU/mL).

The specificity of multiplex PCR was evaluated by PCR using strains of other microbial species isolated directly from a culture and from positive blood culture bottles containing other bacterial species. The following species were analyzed: Enterococcus faecalis, Enterococcus faecium, Streptococcus viridans, Streptococcus pneumoniae, Candida spp., Gram-positive bacilli, Pseudomonas aeruginosa, Escherichia coli, Klebsiella pneumoniae, Acinetobacter baumannii, Enterobacter cloacae, Enterobacter aerogenes, Salmonella spp., Morganella morganii, and Proteus mirabilis.

The isolates not carrying mecA gene although resistant to oxacillin by the disk diffusion test were submitted to the fowling tests to confirm if they were β-lactamase hyperproducers, and to confirm the resistance profile: nitrocefin disk test for β-lactamase hyperproduction; simplex PCR for mecA gene detection, cefoxitin by disk diffusion, and amoxicillin/clavulanic acid disk test for resistance confirmation.

The kappa statistic was calculated to assess agreement between the methods used for the identification of Staphylococcus sp. A level of significance of 5% was adopted for all tests. Statistical analyses were performed using SSPS software version 19.0 (SSPS Inc., Chicago, IL).

Identification by the phenotypic method detected 85 (23.0%) S. aureus strains and 286 (77.0%) CoNS strains in the 371 blood cultures analyzed. The disk diffusion test revealed oxacillin resistance in 43 (50.6%) S. aureus isolates and in 236 (82.5%) CoNS isolates.

Multiplex PCR was standardized for genotypic identification of the genus Staphylococcus (16S rRNA gene), to differentiate coagulase-positive staphylococci and CoNS (coa gene), to identify S. aureus (S. aureus-specific gene), and to detect oxacillin resistance (mecA gene) directly from blood culture bottles (Fig. 1).

Fig. 1.

Electrophoresis gel (3% agarose) of the products of multiplex amplification in blood cultures for the detection of Staphylococcus spp. and the mecA gene. Lane 1: 100–2000bp Ladder; lane 2: ATCC 33591 (mecA-positive S. aureus); lane 3: ATCC 25923 (mecA-negative S. aureus); lane 4: ATCC 12228 (mecA-negative S. epidermidis); lane 5: negative control; lanes 6 and 7: mecA-positive S. aureus isolate; lanes 8 and 9: mecA-negative CoNS isolate; lane 10: mecA-negative S. aureus isolate.

(0.07MB).

Amplification was performed on all 371 isolates included in the study. In the case of the 85 S. aureus isolates identified phenotypically, there was 100% agreement with the multiplex PCR results, with amplification of the 16S rRNA, S. aureus and coa genes. Among the 286 CoNS isolates included, 100% agreement between the two methods was only observed in terms of amplification of the 16S rRNA gene.

Amplification of the mecA gene was observed in 43 (50.6%) of the 85 S. aureus isolates; of these, 42 were resistant to oxacillin by the disk diffusion test and one isolate was susceptible. Among the 42 (49.4%) mecA gene-negative isolates, 41 were susceptible by the disk diffusion method and one isolate was resistant. The mecA gene was amplified in 225 (78.7%) of the 286 CoNS isolates; of these, 218 were resistant to oxacillin by the disk diffusion test and seven were susceptible. Among the 61 (21.3%) mecA gene-negative CoNS isolates, 43 were classified as susceptible by the disk diffusion test and 18 were resistant.

Among the 19 isolates not carrying mecA gene although resistant to oxacillin by disk diffusion test, all were positive by nitrocefin disk test and cefoxitin disk diffusion test. On the other hand, six isolates were resistant to amoxicillin/clavulanic acid by disk diffusion test.

In the sensitivity test, amplification of fragments was observed up to a dilution of 10−5 (103CFU/mL) for CoNS and of 10−2 (106CFU/mL) for S. aureus, as shown in Fig. 2. Analysis of specificity using isolates of other species revealed no amplification of any fragments of the genes investigated by multiplex PCR (Fig. 3).

Fig. 2.

Sensitivity of multiplex PCR. Electrophoresis gel showing the amplification of different dilutions of CoNS and S. aureus. Lane 1: 100–2000bp molecular weight marker; 2: mecA-positive S. aureus ATCC; 3: S. epidermidis ATCC; 4: CoNS 10−7 dilution (10CFU/mL); 5: 10−6 (102CFU/mL); 6: 10−5 (103CFU/mL); 7: 10−4 (104CFU/mL); 8: 10−3 (105CFU/mL); 9: 10−2 (106CFU/mL); 10: 10−1 (107CFU/mL), 11: S. aureus 10−7 dilution (10CFU/mL); 12: 10−6 (102CFU/mL); 13: 10−5 (103CFU/mL); 14: 10−4 (104CFU/mL); 15: 10−3 (105CFU/mL); 16: 10−2 (106CFU/mL); 17: 10−1 (107CFU/mL); 18: 100–2000bp molecular weight marker.

(0.06MB).

Fig. 3.

Electrophoresis gel illustrating the specificity of multiplex PCR. Lane 1: 100–2000bp Ladder; 2: mecA-positive S. aureus ATCC; 3: mecA-negative S. aureus ATCC; 4: CoNS ATCC; 5: Salmonella spp.; 6: Proteus mirabilis; 7: Morganella morganii; 8: Klebsiella pneumoniae; 9: Pseudomonas aeruginosa; 10: Acinetobacter baumannii; 11: Enterobacter cloacae; 12: Enterobacter aerogenes; 13: E. coli; 14: E. faecalis; 15: E. faecium; 16: Streptococcus viridans; 17: Streptococcus pneumoniae; 18: Candida spp.; 19: Gram-positive bacilli.

(0.05MB).

The cost of multiplex PCR and the automated Vitek 2 system, the system used for direct bacterial identification in blood cultures at the Laboratory of Microbiology, Department of Internal Medicine, University Hospital, Botucatu Medical School, UNESP, was lower than the phenotypic method (Table 2).