Direct mecA Detection from Blood Culture Bottles by Branched-DNA Signal Amplification

Abstract

A branched-DNA (bDNA) signal amplification method was used to detect the mecA gene directly from blood culture broth growing staphylococci. BACTEC blood culture bottles with positive growth indices and containing staphylococcus-like organisms as shown by Gram stain were tested for the presence of the mecA gene. Comparison of test results was done among 225 patients (one blood culture from each patient). Compared with PCR, the sensitivity and specificity of the bDNA method are 100 and 99%, respectively. The bDNA test is carried out in a 96-well format and requires approximately 6 h to perform. Our preliminary results suggest that direct detection of the mecA gene by bDNA signal amplification is (i) sensitive enough to detect mecA directly from blood culture bottles without the requirement for subculture and (ii) as sensitive and specific as the PCR-based method.

Staphylococci are the most common etiologic agents of nosocomial bloodstream infections (10). Recently, methicillin-resistant Staphylococcus aureus has been recognized as an important pathogen for both hospitalized patients and possibly for community-acquired infections (2, 7, 12). Rapid identification of methicillin-resistant S. aureus and methicillin-resistant coagulase-negative staphylococci from patients may also be important for the implementation of appropriate antibiotic therapy, since less expensive β-lactam antibiotics can potentially be used in strains lacking the mec gene. Currently, the most widely used methods for determining methicillin resistance include agar or broth dilution and disk diffusion. All of these tests require the isolation and subculture of the organisms.

In most methicillin-resistant staphylococci, the mecA gene is localized on the bacterial chromosome. The mecA gene encodes a penicillin binding protein (PBP2a) which has a low affinity for β-lactams (4, 16). Because the expression of the mecA gene is highly variable and dependent on different factors (5, 11, 15, 16, 18), phenotypic methods such as agar or broth dilution and disk diffusion have been shown to be less accurate than genotypic detection methods such as PCR (1, 6, 7, 9, 10, 13, 14). This is especially true for coagulase-negative staphylococci (8). The conserved nature of the mecA gene in all staphylococcal species has enabled the development of molecular diagnostic techniques that specifically target this gene. These assays are not dependent on phenotypic expression.

Despite their promise, however, most genotypic detection tests are time-consuming and results are not usually obtained more rapidly than with conventional methods. We previously reported the successful utilization of a branched-DNA (bDNA) signal amplification assay to detect the mecA gene in 416 clinical staphylococcal isolates (8). In the present study, we used this assay to detect the mecA gene directly in blood culture broth containing clinical isolates of staphylococci without the need for subculture.

(This work was presented in part of the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 28 September to 1 October, 1997.)

To evaluate mecA detection by the bDNA method, an aliquot was obtained from each of consecutive positive BACTEC blood culture bottles (Becton Dickinson, Sparks, Md.), representing samples from 225 patients. These blood culture bottles all showed positive growth indices by the BACTEC 9240 system and contained clusters of gram-positive cocci by Gram staining. The comparison of the bDNA and PCR methods was done with one blood culture from each patient. All isolates grown in the BACTEC blood culture bottles were subcultured onto 5% sheep blood agar plates for identification purposes and for PCR confirmation of the bDNA results. S. aureus isolates were identified by a positive coagulase test. Coagulase-negative staphylococci were not speciated.

Blood cultures that were positive for nonstaphylococcal agents were assayed for mecA in order to assess the specificity of the assay. Aliquots from 40 blood culture bottles that were positive for at least one of the following gram-positive organisms or yeasts were obtained (some bottles had more than one organism present): Streptococcus pneumoniae (five isolates), viridans group Streptococcus (nine isolates), Abiotrophia spp. (nutritionally variant Streptococcus; one isolate), Enterococcus faecalis (five isolates), Enterococcus faecium (six isolates), Corynebacterium species (one isolate), Actinomyces odontolyticus (one isolate), Candida albicans (one isolate), Candida parapsilosis (four isolates), and Bacillus spp. (two isolates).

The selection of PCR primers and amplification and detection methods were performed as described previously (6).

The mecA bDNA test was performed with bDNA reagents provided by Bayer Diagnostics (Emeryville, Calif.). The procedures were performed according to the manufacturer's instructions and were similar to those described previously (8) with the exceptions that 50 μl of sample was added to 150 μl of lysis solution and heated for 10 min at 120°C, and 25 μl of a dilution was approximately equal to a 1.0 McFarland standard. ATCC 33591 and ATCC 12600 were included as positive and negative controls, respectively. As previously described, signal-to-noise ratios of 3.0 or greater were considered positive for the presence of the mecA gene (8).

All 40 specimens from blood bottles that were positive for gram-positive organisms other than staphylococci and yeasts were negative by the bDNA assay (data not shown).

Compared with PCR (Table 1), the overall sensitivity, specificity, and positive and negative predictive values for the bDNA method are 100, 99, 99, and 100%, respectively. One specimen (which grew S. aureus) that was positive by bDNA (signal-to-noise ratio = 10.1) but negative by PCR was detected in one of the two blood culture bottles obtained from the same patient on the same day; however, turbid broth from both bottles was tested and only one bottle gave discrepant results (bDNA positive but PCR negative). The result remained the same after repeating both bDNA and PCR assays. Oxacillin MIC determination was done for the isolate by agar dilution, with plates incubated for both 24 h at 35°C and 48 h at 30°C in NaCl-containing medium. The MIC was 1 μg/ml (susceptible) under both conditions. The reason for the inconsistency between bDNA and the other tests is not clear at this time, but it could be due to some unknown factors within the bottle or to an analytical error.

TABLE 1.

Comparison of bDNA test results and PCR^a

Result of bDNA	No. of isolates by PCR
	Positive	Negative
Positive	122 (16)b	1c (1)b
Negative	0	102 (33)b

^aA total of 225 staphylococcus-positive blood culture bottles were used, representing 1 culture from each of 225 patients. The 225 isolates included 50 S. aureus isolates and 175 coagulase-negative staphylococci. 

^bValues in parentheses indicate S. aureus isolates. 

^cTwo blood culture bottles were obtained from the same patient on the same day. Both bottles were tested and only one bottle gave a discrepant result. 

Several reports have shown that genetic detection of mecA is more accurate than phenotypic methods, especially for coagulase-negative staphylococci (9, 10, 13, 17). In a previous study (8), by using high concentrations of salt or gradually increasing oxacillin concentration in the medium, we demonstrated that oxacillin resistance could be induced in 6 of 10 mecA-positive isolates that had been classified as oxacillin sensitive by the disk diffusion method.

One report that describes the utilization of molecular diagnostic methods to detect mecA directly from blood culture bottles has been published (3). The advantages of the bDNA test are that it does not require a complicated sample preparation procedure and is not prone to inhibition by sodium polyanetholesulfonate and other substances known to be inhibitory to PCR. The bDNA assay results in the amplification of the chemiluminescent signal, rather than amplification of the DNA target; therefore, contamination by the amplification product is not an issue. The test is carried out in a 96-well format and is easy to perform. It requires approximately 6 h (after a positive growth index has been detected) to perform, making it possible to obtain results the same day that a positive blood culture is identified. In our laboratory, this would result in a 2- to 3-day savings in time to first result. Such methods should reduce antibiotic costs by permitting the use of the less expensive antibiotics (narrow-spectrum β-lactams versus vancomycin) for treatment of the susceptible strains and decreasing the opportunity for other organisms to acquire vancomycin resistance. In summary, our preliminary results suggest that direct detection of the mecA gene by bDNA signal amplification is (i) sensitive enough to detect mecA directly from blood culture bottles without the requirement for subculture and (ii) as sensitive and specific as PCR-based methods.