Multiplex Real-Time PCR Assay for Detection of Methicillin-Resistant Staphylococcus aureus and Associated Toxin Genes

Abstract

We describe a real-time PCR assay for the detection of methicillin-resistant Staphylococcus aureus and genes encoding toxic shock syndrome toxin 1 and Panton-Valentine leukocidin. Rapid screening and detection of toxins is a useful tool for surveillance studies and outbreak investigations involving large numbers of isolates.

TEXT

In the course of outbreak investigations and surveillance studies, laboratories often screen large numbers of Staphylococcus aureus isolates for common strain characteristics, including mecA-mediated methicillin resistance and the presence of clinically relevant, strain-predictive toxin genes. Conventional methods for the identification of methicillin-resistant S. aureus (MRSA) and detection of toxin genes often require several different assay types and can be time- and labor-intensive. To address these issues, we designed a multiplex real-time PCR (MRStox) that simultaneously detects nuc, an S. aureus-specific marker that encodes staphylococcal thermonuclease (6); mecA, which confers methicillin resistance (9); and two toxin genes, tst (toxic shock syndrome toxin 1 [TSST-1]) and lukS-PV (Panton-Valentine leukocidin [PVL]) (5). Previously characterized MRSA isolates from our repository, including those collected through the Active Bacterial Core Surveillance (ABCS) project (n = 147), and prospectively collected presumptive MRSA isolates from our ABCS collection (n = 1,637) were evaluated. Also included were the S. aureus type strain (ATCC 12600), six coagulase-negative Staphylococcus (CNS) type strains (S. saprophyticus ATCC 15305, S. epidermidis ATCC 14990, S. hominis ATCC 27844, S. kloosii ATCC 43959, S. lugdunensis ATCC 43809, and S. schleiferi subsp. schleiferi ATCC 43809), and one Enterococcus type strain (E. faecalis ATCC 29212). DNA was obtained from 18- to 24-h cultures as previously described (5). Real-time PCR primers and probes were used at the concentrations indicated (Table 1) in 25-μl reaction mixtures containing 5 μl DNA and 20 μl LightCycler 480 (LC480) Probe Master mix (Roche Diagnostics, Indianapolis, IN). The assay utilized an LC480 real-time PCR instrument according to the following cycling parameters: 1 cycle of 95°C for 10 min, followed by 40 cycles consisting of 95°C for 15 s, 60°C for 40 s, and 72°C for 20 s. Ramp rates were 4.4°C/s, 2.2°C/s, and 4.4°C/s, respectively. Positive, negative, and no-template controls were included for each run (BAA-1752 [lukS-PV positive, tst negative, mecA positive, nuc positive], ATCC 51650 [lukS-PV negative, tst positive, mecA negative, nuc positive], ATCC 14990 [S. epidermidis, lukS-PV negative, tst negative, mecA negative, nuc negative], and deionized water). A sigmoid-shaped amplification plot with a crossing threshold (C_T) between cycles 9 and 38 was interpreted as a positive result. lukS-PV and tst results were compared to results previously obtained with our standard toxin duplex real-time PCR assay (5). The presence of mecA was compared to results obtained via oxacillin and cefoxitin disk diffusion testing (3). The presence of nuc, confirming S. aureus identification, was compared to results obtained with the Staphaurex (Remel, Lenexa, KS) latex agglutination test. Discrepant results were repeated by both methods and compared to expected results for additional tests, including a positive tube coagulase test, fermentation of mannitol and trehalose, a positive Voges-Proskauer test, and detection of l-pyrrolidonyl arylamidase activity (2). For this evaluation, the result obtained via the tube coagulase test was considered the final arbiter of S. aureus versus non-S. aureus species identity. Discrepant results for the presence of tst and lukS-PV were resolved by repeating the MRStox assay and the duplex assay described by Limbago et al. (5). Discrepant results for the presence of mecA were resolved by comparing the results obtained by repeating this multiplex method to those obtained by a simplex mecA PCR (4). Repeated identical results for the presence of tst, lukS-PV, and mecA were considered correct results.

Table 1.

Primers and probes used in this study

Gene assayed	Primer or probe name	Sequence (5′-3′)	Final concn (μM)	Fluorophorea	Mean CT* (SD)	Reference
lukS-PV	PVLTAQF	AAT GAA ATG TTT TTA GGC TCA AGA CA	1			This study
lukS-PV	PVLTAQR	TGG ATA ACA CTG GCA TTT TGT GA	1			This study
lukS-PV	PVLTAQT	AG CAA CTT AAA TGCb TGG ACA AAA CTT CTT GGA A	1	HEX	21.60 (1.60)	This study
tst	STSTQF	CCC TTT GTT GCT TGC GAC A	1			This study
tst	STSTQR	GCT TTT GCA GTT TTG ATT ATT TGA TT	1			This study
tst	STSTQT	TC GCT ACA GAT TTTc ACC CCT GTT CCC TTA TCA T	1	CalRd610	25.28 (1.89)	This study
mecA	LTmecAF	AAA GAA CCT CTG CTC AAC AAG T	1			8
mecA	LTmecAR	TGT TAT TTA ACC CAA TCA TTG CTG TT	1			8
mecA	LTmecAHP2	CC AGA TTA CAA CTT CAC CAG GTT CAA CTb	1	Quas670	21.26 (1.93)	8
nuc	LTnucF	AAA TTA CAT AAA GAA CCT GCG ACA	3			8
nuc	LTnucR	GAA TGT CAT TGG TTG ACC TTT GTA	3			8
nuc	LTnucHP1	AA TTT AAC CGT ATC ACC ATC AAT CGC TTTb	3	FAM	21.78 (1.97)	8

^aHEX, 2,5,2′,4′, 5′,7′-hexachloro-6-carboxyfluorescein; FAM, 6-carboxyfluorescein; CalRd610, CalFluor Red 610; Quas670, Quasar 670.

^bThe underlined T is Black Hole Quencher 1.

^cThe underlined T is Black Hole Quencher 2.

All 148 retrospective S. aureus isolates were positive for the nuc target, while all 7 CNS and Enterococcus isolates were negative. Of 1,637 prospectively collected S. aureus isolates, 1,633 (99.8%) were positive for the nuc target; the four nuc negative isolates remained PCR negative upon repeat testing. Two of the nuc PCR-negative isolates were identified as atypical S. aureus strains via the biochemical tests described above (2) and thermonuclease activity (thermonuclease agar; Remel, Lenexa, KS), similar to the phenomenon described by van Leeuwen et al. (10). Further biochemical testing (ID32 STAPH; bioMérieux SA, Marcy l'Etoile, France) identified the other two as a Staphaurex-positive S. saprophyticus and a Rothia sp. exhibiting autoagglutination (Table 2; sensitivity, 0.999; specificity, 1.00).

Table 2.

Results of MRStox gene detection

MRStox result	No. of isolates with indicated results
	lukS-PV		tst		nuca		mecA
	Positive	Negative	Positive	Negative	Positive	Negative	Positive	Negative
Positive	27	1b	8	1b	1,781	0	141	0
Negative	2c	118	0	139	2	9	1	13d

^aBased on species identification (S. aureus, positive; non-S. aureus, negative).

^bPositive by both methods upon repeat testing.

^cNegative by both methods upon repeat testing.

^dOne isolate was methicillin resistant by oxacillin and cefoxitin broth microdilution but not by mecA PCR.

The presence of mecA was detected in 141/142 MRSA isolates; mecA was not detected in 6/6 methicillin-susceptible S. aureus isolates or in any of the methicillin-susceptible CNS or Enterococcus isolates (Table 2; sensitivity, 0.993; specificity, 1.00). The lone mecA-discordant S. aureus isolate was also found to be negative for mecA by a conventional simplex PCR assay (4), though it was resistant to oxacillin and cefoxitin by broth microdilution (3). The phenomenon of oxacillin-resistant strains lacking mecA is uncommon but has been described by others (1, 7); these should be reported as MRSA (3).

The presence or absence of PVL was concordant for 145/148 (98.0%) S. aureus isolates by both methods. Upon repeat testing by both methods, the MRStox results were unchanged; the one false-negative and two false-positive results obtained with the conventional duplex PCR were resolved to match those obtained via the MRStox assay.

The presence or absence of TSST-1 was concordant for 147/148 (99.3%) S. aureus isolates. Upon repeat testing of the discordant isolate, the MRStox result remained unchanged while the false-positive result obtained with our conventional toxin assay was resolved (Table 2). The mean C_T for all four targets ranged from 21.26 for mecA to 25.28 for tst (Table 1). As expected, the lukS-PV and tst genes were not detected in any CNS or Enterococcus isolates.

The reproducibility of the MRStox assay between operators was evaluated with a subset of 82 S. aureus isolates. Reproducibility between operators resulted in 82/82 congruous results (data not shown). Positive, negative, and no-template controls for all tests were as expected.

In conclusion, the MRStox assay is reliable, reproducible, and rapid and may be more sensitive and specific than our previous assay. The data can be very useful in assessing strains of S. aureus for investigation of outbreaks and for surveillance studies.