Genotypic Diversity of Methicillin-Resistant Staphylococcus aureus Isolates in Korean Hospitals

Abstract

Ninety-six methicillin-resistant Staphylococcus aureus (MRSA) isolates from eight Korean hospitals were analyzed by multilocus sequence typing, SCCmec typing, and spa typing. The predominant genotype was ST5-MRSA-II of clonal complex 5, which was found in 36 isolates from six hospitals, but ST239-MRSA-III was also common. Overall, results showed a notable genotypic diversity of MRSA strains circulating in Korean hospitals.

Widespread emergence of methicillin-resistant Staphylococcus aureus (MRSA) has been a serious clinical problem worldwide since the 1980s. One of the main factors for the global emergence of MRSA is the dissemination of epidemic MRSA clones, which was documented by molecular typing methods such as pulsed-field gel electrophoresis, multilocus sequence typing (MLST), and SCCmec typing (3, 6, 7, 13). In a previous study, we identified two major MRSA clones in Asian countries (8). Based on MLST and SCCmec typing, clonal complex 5 (CC5)-MRSA-II was found to be prevalent in Korea and Japan, while CC239-MRSA-III (or -IIIA) was found to be prevalent in other Asian countries. However, since MRSA isolates analyzed in the previous study had been obtained from a single hospital, data could not represent the nationwide distribution of MRSA genotypes in Korea.

In the present study, we performed MLST, SCCmec typing, and spa typing with MRSA isolated in 2003 and 2004 from eight tertiary care hospitals in various regions of Korea.

MRSA isolates.

A total of 96 nonduplicate MRSA isolates from eight tertiary care hospitals in various parts of Korea were analyzed in this study (Table 1). All MRSA isolates were collected during the period between August and October, 2003, except those from the Chungnam National University hospital, which were collected in May 2004. All isolates were obtained from blood or urine specimens from patients who had clinical manifestations of staphylococcal infection. In vitro susceptibility to oxacillin was determined by broth microdilution testing according to CLSI guidelines (1).

TABLE 1.

Distribution of sequence types of methicillin-resistant Staphylococcus aureus isolates from Korean hospitals

Localitya	Clonal complex (CC) and no. of sequence types (ST)
	CC5			CC239		CC89		CC1		Singletons
	ST5	ST518	ST328	ST239	ST519	ST89	ST509	ST1	ST514	ST254	ST72	ST30	ST121	ST513	ST515	ST516	ST517
Seoul-A (14)	7			2						1	2	1		1
Seoul-B (6)	3			2		1
Seoul-C (13)	7			4							2
Daegu (14)	5	1		7									1
Busan (14)	7			2	2	1					1	1
Gwangju (11)	5		1	1		1	1			1	1		1
Daejeon (15)	6			4						1	1				1	1	1
Anyang (9)	1			1				2	1	3		1
Total (96)	41	1	1	23	2	3	1	2	1	6	7	3	1	1	1	1	1

^aThe number of isolates from each hospital is shown in parentheses. Seoul-A, Samsung Medical Center; Seoul-B, Kangbuk Samsung Hospital; Seoul-C, Kyunghee University Medical Center; Daegu, Kyungbook National University Hospital; Busan, Dong-A University Hospital; Gwangju, Chonnam National University Hospital; Daejeon, Chungnam National University Hospital; Anyang, Hallym Medical Center.

Molecular characterization.

MLST was performed and analyzed as described previously (4, 8). The SCCmec types were determined by a multiplex PCR as described previously, which identified types I to IV and their variants, IA, IIIA, and IVA (12). Nontypeable types were defined as those showing unexpected fragments or lacking some fragments in multiplex PCR and so not corresponding to previously defined SCCmec types (Table 2). The MRSA genotypes were defined based on sequence type (ST) and SCCmec type as proposed by Robinson and Enright (14). spa typing was performed with each isolate as previously described (7, 9, 15). The Panton-Valentine leukocidin (PVL) gene was screened by the PCR method of Lina et al. (10).

TABLE 2.

Definition of nontypeable SCCmec types (NT-A to NT-E) compared with well-characterized SCCmec types

SCCmec type (no. of isolates)	Locusa								ST (no. of isolates)
	A	B	C	D	E	F	G	H
I (IA) (0)	+			+			(+)
II (43)		+	+	+			+		ST5 (34), ST72 (1), ST89 (3), ST239 (2), ST328 (1), ST509 (1), ST518 (1)
III (19)			+		+	+		+	ST5 (2), ST30 (1), ST72 (2), ST239 (14)
IIIA (8)			+		+	+			ST5 (1), ST30 (1), ST239 (4), ST519 (2)
IV (2)				+					ST5 (1), ST254 (1)
IVA (12)				+			+		ST1 (2), ST5 (2), ST72 (3), ST239 (3), ST254 (2)
NT-A (3)				+		+	+		ST254 (3)
NT-B (5)					+		+		ST30 (1), ST121 (1), ST513 (1), ST514 (1), ST516 (1)
NT-C (1)				+	+			+	ST72 (1)
NT-D (2)		+	+	+	+	+		+	ST515 (1), ST517 (1)
NT-E (1)			+		+			+	ST5 (1)

^aLocus described by Oliveira et al. (12). A, downstream of pls gene; B, kdp operon; C, mecI gene; D, dcs gene; E, region between pI258 and Tn554; F, region between Tn554 and orfX; G, left junction between IS431 and pUB110; H, left junction between IS431 and pT181. +, present; (+), present in type IA but absent in type I.

In MLST analysis, 96 MRSA isolates from eight tertiary care hospitals in Korea showed 17 STs (Table 1). The most prevalent STs were ST5 and ST239, which were found in 41 (42.7%) and 23 (23.6%) isolates, respectively. Both STs were isolated from all eight hospitals. Six STs (ST72, ST254, ST30, ST89, ST1, and ST519) were found in two to seven isolates, and the other nine STs were found only in one isolate. Of the 17 STs, 9 (ST328, ST509, ST513, ST514, ST515, ST516, ST517, ST518, and ST519) were newly identified in this study, which represented 10 MRSA isolates (10.4%). Clustering analysis by eBURST program (5) showed two major clonal complexes (CC5 and CC239), two minor clonal complexes (CC89 and CC1), and eight singletons (Table 1). CC5 and CC239 were most prevalent in most hospitals except the Hallym Medical Center in Anyang.

The SCCmec types of Korean MRSA isolates included types II (43 isolates), III (19 isolates), IIIA (8 isolates), IV (2 isolates), and IVA (12 isolates). However, twelve MRSA isolates (12.5%) were nontypeable (NT) by the multiplex SCCmec typing method (Table 2). These nontypeable SCCmec types could be classified into five types by comparing amplified band patterns (Table 2). These isolates with NT SCCmec types showed 10 different STs. ST5, the most predominant ST found in 41 isolates, showed six different SCCmec types: II, III, IIIA, IV, IVA, and NT-E (Table 3). Of these genotypes, the most representative one was ST5-MRSA-II, which was found in 34 isolates. ST239, another prevalent ST in 23 Korean isolates, showed four SCCmec types, II, III, IIIA, and IVA (Table 3).

TABLE 3.

Comparison of genotypic diversity between MRSA isolates of CC5 and CC239

CC	ST	SCCmec	spa	No. of isolates
CC5	ST5	II	TJMBMDMGMK	21
CC5	ST5	II	TJMBBMDMGMK	7
CC5	ST5	II	TJMBBMDMK	1
CC5	ST5	II	TJMBMAMGMK	1
CC5	ST5	II	TJMB # DMGMKa	1
CC5	ST5	II	TJMGMKK	1
CC5	ST5	II	TMBBMDMGMK	1
CC5	ST5	II	TMBMDMGMK	1
CC5	ST5	III	TJMBMDMGMK	2
CC5	ST5	IIIA	TJMBMDMGMK	1
CC5	ST5	IV	TJMBBMDMGMK	1
CC5	ST5	IVA	TJMBMDMGMK	2
CC5	ST5	NT-E	TJMBMDMGMMK	1
CC5	ST518	II	TJMB # DMGMKa	1
CC5	ST318	II	TJMBBMDMGMK	1
CC239	ST239	II	WGKAOMQ	2
CC239	ST239	III	WGKAOMQ	14
CC239	ST239	IIIA	WGKAOMQ	4
CC239	ST239	IVA	WGKAOMQ	3
CC239	ST519	IIIA	WGKAOMQ	2

^a# denotes new repeat of spa, AAAGAAGACGGCAACAAGTCTGGT.

In this study, 25 distinct spa types were identified in 96 MRSA isolates from Korea. spa types showed correlation with STs by MLST (Table 3), as documented in other studies (2, 11). MRSA isolates which belonged to CC5 (ST5, ST518, and ST328) showed nine different spa types with similar motifs (DMGMK) (Table 3). However, MRSA isolates belonging to CC239 showed only one spa type, WGKAOMQ (Table 3). Two spa types (WGKAKAOMQ and WFKAOMQ) containing the same motif (KAOMQ) with WGKAOMQ were shown in three isolates with ST30.

Panton-Valentine leukocidin (PVL) gene, which is assumed to be associated with community-associated MRSA (10), was not detected in any of the MRSA isolates.

As documented in the previous study (8), present results confirmed that ST5-MRSA-II is the most prevalent genotype in Korean MRSA isolates. However, in this study we recognized that MRSA isolates with the ST239-MRSA-III (or -IIIA) genotype, which were not detected in the previous study (8), are also present in Korea. Although in the previous study only 13 isolates from a single hospital were investigated (8), the presence of ST239-MRSA-III (or -IIIA) in MRSA isolates obtained in 2003 and 2004 might suggest that the clones with this genotype have recently been introduced in Korea. The diversity of spa types could support this speculation. While isolates with ST5-MRSA-II showed eight different spa types, isolates with ST239-MRSA-III (or -IIIA) showed a single spa type, which might suggest the recent introduction of these strains.

Based on MLST and SCCmec typing, MRSA isolates from the Hallym Medical Center in Anyang showed a genotype distribution different from that seen with the other hospitals. Many MRSA isolates from this hospital showed ST1 or ST254 with SCCmec type IVA, which is one of the characteristics of community-associated MRSA. Further epidemiologic study to evaluate this unique distribution of genotypes in this hospital is under way.

In summary, we investigated the diversity of MRSA isolates from eight Korean tertiary care hospitals. The most prevalent clone was ST5-MRSA-II, followed by ST239-MRSA-III (or -IIIA). Strains with ST239-MRSA-III (or -IIIA) might have been recently introduced into Korean hospitals. Although these data may not completely reflect the overall genotype distribution of MRSA strains from Korean hospitals due to the relatively small number of isolates investigated, the results showed a notable genotypic diversity of MRSA strains circulating in Korean hospitals.