Rapid Identification of Staphylococcus aureus and the mecA Gene from BacT/ALERT Blood Culture Bottles by Using the LightCycler System

Abstract

One hundred BacT/ALERT blood culture bottles growing gram-positive cocci in clusters were cultured and studied by LightCycler PCR for the sa442 and mecA genes. PCR was 100% sensitive and specific for detecting Staphylococcus aureus and methicillin resistance in S. aureus but was less accurate for methicillin resistance in coagulase-negative staphylococci.

Bacteremia with Staphylococcus aureus often indicates a serious medical condition. Coagulase-negative staphylococci (CoNS), however, may be contaminants. Consequently, when gram-positive cocci in clusters are seen in Gram stains of signal-positive blood culture bottles, two important questions arise. The first is whether the organism is S. aureus or a CoNS. The second question is whether the organism is susceptible to methicillin or not. The first distinction requires 18 to 24 h by conventional techniques. Determination of susceptibility requires an additional 6 to 10 h, if the organism is in pure culture. Earlier clarification of these questions would likely have significant clinical impact. Martineau et al. have identified a DNA segment, the sa442 gene, unique to S. aureus (3). The major determinant of methicillin resistance in staphylococci, the mecA gene, occurs in both S. aureus and CoNS (6, 8). These genes have been used for rapid identification of methicillin-resistant S. aureus (MRSA) in bacterial subcultures on solid media (5) and directly in BACTEC blood culture bottles using the LightCycler system (7).

The BacT/ALERT FA and BacT/ALERT FN bottles used with the BacT/ALERT blood culture system contain carbon particles to adsorb inhibitors of bacterial growth; although useful in this regard, the particles interfere with PCR and fluorescence detection in the LightCycler system. Any attempt to remove the particles may also diminish the amount of target bacteria present and thereby possibly decrease the sensitivity of subsequent PCR. This study was conducted to examine the feasibility of detection of S. aureus and identification of methicillin resistance directly in positive BacT/ALERT (bioMerieux, Inc., Durham, N.C.) bottles using the LightCycler system.

Routine blood culture bottles were incubated in the BacT/ALERT automated continuous monitoring system. Bottle contents that signaled positive were identified, Gram stained, and inoculated onto 5% sheep blood agar; if gram-positive cocci in clusters were seen, a 0.5-ml aliquot for PCR was also immediately removed. The charcoal and the bacteria, both suspended in the BacT/ALERT blood culture medium, were separated by differential centrifugation. The aliquot was centrifuged at 850 × g for 2 min to remove the charcoal. The supernatant was centrifuged at 11,500 × g for 5 min. The resulting pellet was resuspended in 200 μl of a lysis buffer (5) containing 1% Triton X-100, 0.5% Tween 20, 10 mM Tris-HCl (pH 8.0), and 1 mM EDTA and incubated in a screw-cap reaction tube at 100°C for 10 min. This was again centrifuged at 850 × g for 2 min, and the supernatant was removed and stored at −20°C for future testing.

The sa442 gene and the mecA gene were amplified on a LightCycler, using previously described primers (3, 5): Sa-442F (5′-GTCGGTACACGATATTCTTCACG-3′), Sa442-R (5′-CTCTCGTATGACCAGCTTCGGTAC-3′), mecA-F (5′-CAAGATATGAAGTGGTAAATGGT-3′), and mecA-R (5′-TTTACGACTTGTTGCATACCATC-3′). We, however, performed the sa442 gene and the mecA gene assays separately, rather than the duplex reaction as previously described (5); this was done because we noted a slightly diminished detection of the mecA gene under duplex reaction conditions (unpublished data). Product detection was accomplished using previously described hybridization probes labeled with fluorescein (fam) and fluorophore molecules (Red 640 and Red 705) (3, 5): Sa442-HP-1 (5′-TACTGAAATCTCATTACGTTGCATCGGAA-[fam]3′), Sa442-HP-2 (5′[Red 705]-ATTGTGTTCTGTATGTAAAAGCCGTCTTG-p3′), mecA-HP-1 (5′-CAGGTTACGGACAAGGTGAAATACTGATT-[fam]3′), and mecA-HP-2 (5′[Red 640]-ACCCAGTACAGATCCTTTCAATCTATAGCG-p3′). All oligonucleotide primers and probes were obtained from BioChem (Salt Lake City, Utah). The sa442 amplification mixture consisted of 3.0 mM MgCl₂, a 0.25 μM concentration of each sa442 primer, a 0.2 μM concentration of each sa442 probe, and 2 μl of 10× LightCycler FastStart DNA Master Hybridization Probes mixture (Roche) in a volume of 15 μl. The mecA amplification mix was similar, except that the mecA primers and probes were present at concentrations of 0.5 and 0.2 μM, respectively. Five microliters of template DNA extract was added to obtain a reaction volume of 20 μl for each capillary tube. A suspension of Tris-EDTA and an extract of an MRSA, ATCC no. 33592 (American Type Culture Collection, Manassas, Va.), were used as the negative and positive controls, respectively.

The reaction protocol was as follows: an initial FastStart DNA Taq polymerase activation phase at 95°C for 10 min; a 45-cycle amplification phase consisting of a 95°C segment for 10 s, a 50°C segment for 10 s, and a 72°C segment for 20 s; a melt phase from 45 to 75°C with a temperature transition rate of 0.1°C/s; and a rapid cooling phase. The presence of amplified DNA was measured by detection of energy emitted at 640 nm (for the presence of the mecA gene) and at 705 nm (for the presence of the sa442 gene). The temperature at which the hybridization probes dissociated from the target sites was determined by melting curve analysis, as provided for by the LightCycler software. This served as an independent indicator of the specificity of hybridization.

The cultures were processed per routine, using colonial morphology and slide coagulase tests. Latex agglutination (Staphaurex; Murex BioTech Ltd., Dartford, England) and tube coagulase tests were performed in ambiguous cases. Susceptibility testing was performed using the Vitek (bioMerieux, Inc., Hazelwood, Mo.) system. For CoNS, only isolates considered clinically significant were tested for susceptibility (by the Vitek method), as per routine laboratory policy. For the purposes of this study, the remaining CoNS isolates were tested by the Kirby-Bauer oxacillin disk method, according to NCCLS guidelines (4); the Vitek system was used for ambiguous cases. Detection of S. aureus and the mecA gene by real-time PCR was compared with conventional identification and susceptibility testing. The sensitivity and specificity of the real-time PCR assays were determined using the conventional methods as the reference standard.

One hundred consecutive positive bottles analyzed in this manner contained 34 S. aureus and 66 CoNS isolates. Both positive and negative controls reacted appropriately. The real-time sa442 PCR assay was 100% sensitive and 100% specific for detecting S. aureus (Table 1). A single isolate, initially identified as a CoNS by conventional testing, was positive for the sa442 gene and was subsequently confirmed to be S. aureus by further biochemical testing. The PCR assay for the mecA gene was 100% sensitive and 100% specific for detecting methicillin resistance in S. aureus and was 77.4% sensitive and 92.3% specific for detecting methicillin resistance in CoNS (Table 2). The melting temperatures of the sa442 and mecA gene probes were 65.12 ± 1.36°C (mean ± 2 standard deviations; n = 34), and 66.03 ± 0.63°C (mean ± 2 standard deviations, n = 66), respectively.

TABLE 1.

sa442 PCR testing of staphylococci from blood culture bottles^a

Organism	No. of samples detected		Total no. of samples
	PCR+	PCR−
S. aureus	34	0	34
CoNS	0	66	66

^aSensitivity = 100%; specificity = 100%. PCR⁺, number of isolates that were positive by the sa442 PCR assay; PCR⁻, number of isolates that were negative by the sa442 PCR assay.

TABLE 2.

Results of mecA PCR testing on methicillin-resistant and methicillin-susceptible staphylococci^a

Organism	No. of samples detected		Total no. of samples
	PCR+	PCR−
All staphylococci (n = 100)b
Methicillin resistant	65	12	77
Methicillin susceptible	1	22	23
S. aureus (n = 34)c
Methicillin resistant	24	0	24
Methicillin susceptible	0	10	10
CoNS (n = 66)d
Methicillin resistant	42	11	53
Methicillin susceptible	1	12	13

^aPCR⁺, number of isolates that were positive by the mecA PCR assay; PCR⁻, number of isolates that were negative by the mecA PCR assay.

^bSensitivity = 84.42%; specificity = 95.65%.

^cSensitivity = 100%; specificity = 100%.

^dSensitivity = 77.36%; specificity = 92.31%.

The timely and accurate information provided by these tests would help clinicians identify S. aureus bacteremia, determine methicillin resistance, and thereby modify antimicrobial therapy 24 to 36 h earlier than would be possible otherwise. The mecA assay performed less optimally with the CoNS, and this discrepancy has been noted before (1). This may be because of resistance to methicillin by mechanisms other than the mecA gene and indicates that mecA gene detection may not be accurate enough to guide therapy for CoNS infections.

The sensitivity of PCR could be compromised by the presence of PCR inhibitors or other physical factors. We did not observe PCR inhibition in our small study sample. Blood culture systems differ in their medium compositions and results with other systems may vary. PCR performed directly on aliquots from BACTEC bottles was found to have a sensitivity of 96% for the detection of S. aureus (7). Further testing is necessary to fully understand the significance of PCR inhibition in blood culture systems. In BacT/ALERT bottles, the presence of carbon particles is a significant obstacle to PCR but was overcome in our study by a simple differential centrifugation process. Although other systems are not hampered by the presence of carbon particles, there may be other impediments to PCR that are unique to each culture system that need to be addressed. In our study, PCR also identified an S. aureus isolate that had been misidentified as a CoNS by conventional methods. Altered colonial morphology, possibly due to antimicrobial therapy, caused the misidentification; medical record review revealed that the patient was on therapy for native aortic valve endocarditis known to be caused by S. aureus from previous cultures. This observation suggests that PCR testing may be more sensitive than traditional methods for identifying S. aureus bacteremia in patients on antimicrobial therapy.

Conversely, it is conceivable that misleading results may occur, particularly if a mixture of methicillin-susceptible S. aureus and mecA-positive CoNS coincidentally is present within the same blood culture bottle (2). The assay described here would correctly characterize the bottle as containing S. aureus but would inaccurately assign it the status of methicillin resistant. Therefore, we suggest that routine culture with standard agar plates continue to be used to identify cultures that contain mixtures of bacteria.

We conclude that the LightCycler sa442 gene assay accurately detects the presence of S. aureus in BacT/ALERT blood culture bottles and that the mecA gene assay reliably determines methicillin resistance in S. aureus but not in CoNS, which may be resistant by additional mechanisms. The significance of this study is that it demonstrates the feasibility of direct detection of S. aureus and identification of MRSA in another major blood culture system, thereby decreasing the time for obtaining definite identification and key susceptibility results by approximately 36 h.