Correlation between Genotype and Phenotypic Categorization of Staphylococci Based on Methicillin Susceptibility and Resistance

Abstract

Positive correlation between methicillin and oxacillin susceptibility test results and the detection of the mecA gene was observed for Staphylococcus aureus, S. epidermidis, and S. haemolyticus as well as among mecA⁺ strains of other species of coagulase-negative staphylococci (CNS). However, at least 50% of the mecA-negative strains of these other species of CNS were falsely classified as methicillin and oxacillin resistant.

Methicillin and oxacillin resistance (MR) in staphylococci is due to the acquisition of the mecA gene, which encodes the low-affinity penicillin-binding protein PBP2a (3, 4). Presence of the mecA gene defines the staphylococcus as MR, while absence of the gene from a staphylococcal strain indicates methicillin susceptibility (MS).

mecA⁺ strains can differ in their level of expression of MR (4). Strains expressing low-level MR (i.e., heterogenously MR) can be difficult to identify by MIC testing; it is often difficult to identify mecA⁺ strains of coagulase-negative staphylococci (CNS) (11). Thus, correctly categorizing staphylococci as MR or MS based on MIC test results has been a challenge. Over the years, the National Committee for Clinical Laboratory Standards (NCCLS) has modified the MIC interpretative criteria for MR and MS so that the MS or MR phenotype correlates better with the mecA genotype (Table 1). While Staphylococcus aureus and CNS once shared the same MIC interpretative criteria for methicillin and oxacillin (8), the two groups now have different MIC interpretive criteria for determining MR and MS (9, 10). More recently, the oxacillin MIC breakpoints for CNS were lowered to allow for increased detection of mecA⁺ S. epidermidis strains, and oxacillin testing is no longer recommended for S. saprophyticus, since mecA-negative strains of this species often phenotypically appear to be resistant (10).

TABLE 1.

NCCLS definitions (interpretive criteria) for MS and MR^a

Species	Compound	MIC (μg/ml)
		Susceptible	Resistant
S. aureus	Methicillin	<16	≥16
	Oxacillin	<4	≥4
CNS	Methicillinb
	Oxacillin (current)	<0.5	≥0.5
	Oxacillin (previous)	<4	≥4
S. saprophyticusc

^aSee references 9 and 10. 

^bNo recommendation. 

^cNo methicillin or oxacillin testing recommended. 

In this study, we assessed the correlation between genotype and phenotypic (MS or MR) categorization of staphylococci. A total of 442 clinical isolates, including 155 S. aureus strains (120 mecA⁺ strains and 35 mecA-negative strains) and 287 CNS strains (104 mecA⁺ strains and 183 mecA-negative strains), were evaluated. S. aureus speciation was based on a positive coagulase test result, and CNS speciation was done using the API-Staph system (bioMérieux, Hazelwood, Mo.).

The detection of the mecA gene was done according to the PCR assay described by Bignardi et al. (1). Precautions were taken to prevent the samples from being contaminated by each other or by the skin of laboratory personnel. These precautions included the use of prealiquoted reagents, gloves, disposable pipettes, and disposable tips with aerosol resistant filters; in addition, the preparation of the amplification reaction mixtures and the analysis of the amplified product were performed in separate areas. Included in every set of PCRs were positive (MR S. aureus strain A27283) and negative (S. aureus strain ATCC 29213) target DNA controls. The methicillin and oxacillin MICs were determined by the NCCLS-recommended agar dilution method (7), using Mueller-Hinton agar supplemented with 2% NaCl and a bacterium inoculum of 5 × 10⁴ CFU/spot. The plates were incubated at 35°C for 24 h. The lowest drug concentration that prevented visible growth was considered the MIC endpoint. Methicillin was obtained from Bristol-Myers Squibb Co. (Syracuse, N.Y.), and oxacillin was obtained from Sigma Chemical Co. (St. Louis, Mo.).

The use of the recently determined oxacillin MIC interpretative breakpoints resulted in mecA-negative and mecA⁺ S. aureus strains being correctly classified as MS and MR, respectively (Table 2). Though mecA-negative and mecA⁺ S. aureus strains were delineated by their methicillin MICs, overlap in the methicillin MIC ranges of the two populations resulted in the misclassification of one mecA-negative S. aureus strain as MR.

TABLE 2.

Correlation between oxacillin MIC and presence of the mecA gene^a

Staphylococcus species	mecAb	No. of strains tested	No. of strains for which oxacillin MIC (μg/ml) was:
			0.06	0.13	0.25	0.5	1	2	4	8	16	32	≥64
S. aureus	−	35		2	8	12	8	5
	+	110									6	9	96
S. epidermidis	−	12		11	1
	+	79					1	9	16	11	9	7	26
S. haemolyticus	−	10		1	9
	+	18							2			1	15
S. hominis	−	25	1	8	7	5	3		1
	+	13					1	4				1	7
S. simulans	−	18			6	8	3			1
	+	2											2
S. wameri	−	19			10	6	3
	+	1											1
S. capitis	−	8		2	2				1	2			1
	+	1									1
S. cohnii	−	10		1			6			3
S. xylosus	−	6			1		5
S. saprophyticusc	−	66				14	50	2

^aPresently, the NCCLS definition for oxacillin susceptibility is, for S. aureus, a MIC of ≤4 μg/ml and, for all other species listed (except S. saprophyticus), a MIC of ≤0.5 μg/ml. Previously, oxacillin susceptibility for all species listed here was defined as a MIC of ≤4 μg/ml. 

^b−, absence of gene; +, presence of gene. 

^cPresently, oxacillin testing is not recommended for S. saprophyticus. 

Likewise, more overlap was observed in the methicillin MIC distribution than in the oxacillin MIC distribution for mecA-negative and mecA⁺ coagulase-negative strains (Tables 2 and 3), corroborating the NCCLS recommendation that oxacillin be used for phenotypic categorization of MS and MR. When the current oxacillin MIC breakpoint of <0.5 μg/ml for MS (10) was used, the mecA-negative S. epidermidis and S. haemolyticus strains were correctly segregated from those that were mecA⁺, whereas when the previous oxacillin MIC breakpoint of <4 μg/ml for MS (9) was used, 13% of mecA⁺ S. epidermidis strains were miscategorized as MS (Table 4).

TABLE 3.

Correlation between methicillin MIC and presence of the mecA gene^a

Staphylococcus species	mecAb	No. of strains	No. of strains for which methicillin MIC (μg/ml) was:
			0.5	1	2	4	8	16	32	≥64
S. aureus	−	35		1	9	22	2	1
	+	110						2	6	102
S. epidermidis	−	12		1	10	1
	+	79		1		5	14	20	6	33
S. haemolyticus	−	10	1	1	7	1
	+	18			1			1		15
S. hominis	−	25	7	5	2	8	3
	+	13		2	1	2				8
S. simulans	−	18	1	6	8	2	1
	+	2								2
S. wameri	−	19		1	6	9	3
	+	1								1
S. capitis	−	8			2	5	1
	+	1								1
S. cohnii	−	10	1	1	3	2	1	2
S. xylosus	−	6			2	4
S. saprophyticus	−	66			1	48	17

^aPresently, the NCCLS definition of methicillin susceptibility for S. aureus is a MIC of ≤8 μg/ml. There are no recommended methicillin breakpoints for CNS. 

^b−, absence of gene; +, presence of gene. 

TABLE 4.

Ability of the oxacillin resistance breakpoint in predicting methicillin susceptibility among mecA-positive, coagulase-negative staphylococcal strains

Staphylococcus species	No. of strains	No. of strains (%) correctly identified as methicillin resistant by oxacillin resistance breakpoint of:
		≥4 μg/mla	≥0.5 μg/mlb
S. epidermidis	79	87	100
S. haemolyticus	18	100	100
S. hominis	13	61	100
S. simulans	2	100	100
S. wameri	1	100	100
S. capitis	1	100	100

^aPrevious MIC breakpoint. 

^bCurrent MIC breakpoint. 

According to the current oxacillin breakpoint, mecA⁺ CNS strains were correctly grouped as MR. No mecA⁺ strains of S. saprophyticus, S. cohnii, and S. xylosus were encountered, and mecA⁺ strains of S. capitis, S. warneri, and S. simulans were uncommon. Conversely, 50% or more of the mecA-negative strains of S. hominis, S. warneri, S. capitis, S. cohnii, and S. xylosus were falsely categorized as MR according to the current oxacillin breakpoint for MS (<0.5 μg/ml) (10) (Table 5).

TABLE 5.

Ability of the oxacillin susceptibility breakpoint in predicting methicillin susceptibility among mecA-negative, coagulase-negative staphylococcal strains

Staphylococcus species	No. of strains	No. of strains (%) correctly identified as methicillin susceptible by oxacillin susceptibility breakpoint of:
		<4 μg/mla	<0.5 μg/mlb
S. epidermidis	12	100	100
S. haemolyticus	10	100	100
S. hominis	25	96	64
S. simulans	17	94	35
S. wameri	19	100	53
S. capitis	8	50	50
S. cohnii	10	70	10
S. xylosus	6	100	16
S. saprophyticus	66	100	—c

^aPrevious MIC breakpoint. 

^bCurrent MIC breakpoint. 

^c—, no oxacillin testing recommended. 

Our findings were similar to those of Hussain et al. (5; L. Stoakes, D. Diagre, V. Fitzgerald, R. Lannigan, and Z. Hussain, Abstr. 100th Gen. Meet. Am. Soc. Microbiol., abstr. C-327, 2000). They observed that use of the new NCCLS oxacillin breakpoint resulted in the correct classification of all CNS strains with mecA as MR and of mecA-negative strains of S. epidermidis and S. haemolyticus as MS. They also noted that the oxacillin test often led to the incorrect determination of mecA-negative strains of S. simulans, S. warnerii, S. cohnii, and S. saprophyticus as MR. While Hussain et al. reported that use of the current oxacillin breakpoint resulted in the correct categorization of mecA-negative strains of S. hominis and S. capitis, our results indicated that 50 to 64% of these strains were grouped inappropriately as MR (Table 5). These discrepancies might be due to differences between the strains analyzed in the two studies or to a difference in CNS speciation. The strains used in this study were clinical trial isolates collected from a wide geographical area. The API-Staph system used in this study was reported to correctly identify 87% for clinical CNS isolates, a frequency the same as that obtained with the conventional method based on selected reactions from the Kloos and Schleifer scheme described by Gahrn-Hansen et al. (2). The species most commonly encountered in the clinical setting are S. epidermidis, S. hominis, S. haemolyticus, and S. warneri (6, 12).

In summary, the current NCCLS-recommended oxacillin breakpoints led to the accurate categorization of S. aureus, S. epidermidis, S. haemolyticus, and mecA⁺ strains of other species of CNS. With the use of the methicillin MIC interpretative criteria, most mecA-negative strains were correctly differentiated from mecA⁺ S. aureus. However, use of the most recently published oxacillin breakpoint resulted in the misclassification of mecA-negative isolates of other species of CNS as MR. Laboratory workers should be aware that while the new oxacillin breakpoints can be used to accurately categorize the most commonly encountered staphylococcal species of clinical significance (i.e., S. aureus, S. epidermidis, and S. haemolyticus) as MS or MR, errors in the phenotypic classification of mecA-negative strains belonging to other CNS species are frequent.