![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article 
Antimicrobial Agents and Chemotherapy, January 2007, p. 264-274, Vol. 51, No. 1
0066-4804/07/$08.00+0 doi:10.1128/AAC.00165-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
Juntendo University, Graduate School of Medicine, Department of Infection Control Science, Tokyo, Japan,1 Juntendo University, Department of Bacteriology, Tokyo, Japan,2 Public Health Research Institute, Newark, New Jersey,3 Faculté de Médecine Laennec, Centre National de Référence des Staphylocoques, IFR62, INSERM E02030, 7 rue guillaume Paradin, 69008 Lyon, France4
Received 8 February 2006/ Returned for modification 7 April 2006/ Accepted 9 October 2006
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Tn554
integrated into the J2 regions of type II and III SCCmec
elements; and M-PCR 6 identified plasmids pT181 and pUB110 integrated
into J3 regions. The system was validated with 99 MRSA strains carrying
SCCmec elements of different types. The SCCmec types
of 93 out of the 99 MRSA strains could be assigned. The SCCmec
type assignments were identical to those made with a PCR system that
uses numerous primer pairs to identify genes or gene alleles. Our
system of six M-PCRs is thus a convenient and reliable method for
typing SCCmec
elements. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
SCCmec elements typically share four characteristics: first, they carry the mec gene complex (mec) consisting of the methicillin resistance determinant mecA and its regulatory genes and insertion sequences; second, they carry the ccr gene complex (ccr) consisting of ccr genes that are responsible for the mobility of the element and its surrounding sequences; third, they have characteristic directly repeated nucleotide sequences and inverted complementary sequences at both ends; and last, they integrate into the 3' end of an open reading frame (ORF), orfX.
Despite these similarities,
the structures of SCCmec elements are rather divergent.
Allotypic differences that are used for SCCmec type
definitions have been identified in both ccr and mec.
Five types of ccr and four classes of mec have been
reported. ccr types 1 to 4 carry the ccrA and
ccrB genes, which share approximately 80% identity with each
other, and the type 5 ccr carries the ccrC gene
(10,
11,
17,
19). Four classes of the
mec gene complexes have been identified among
methicillin-resistant staphylococcal strains of various species: class
A mec, consisting of
IS431mec-mecA-mecR1-mecI; class B
mec, consisting of
IS431mec-mecA-
mecR1-IS1272;
class C mec, consisting of
IS431mec-mecA-mecR1-IS431;
and class D mec, consisting of
IS431mec-mecA-mecR1 with
no insertion sequences downstream of mecR1 identified
by PCR as of yet (13). In
S. aureus strains, mec classes A, B, and C have been
identified. Insertion sequences have sometimes been found to be
integrated in or around the class A mec. A class A
mec carrying IS431 downstream of mecI was
found in Staphylococcus haemolyticus
(13). Recently, Shore et
al. identified MRSA strains carrying class A mec with an
insertion of IS1182 in and around the mecI gene and
designated them classes A3 and A4
(23).
The SCCmec element type has been defined by the combination of ccr type and mec class. In MRSA strains, six types of SCCmec elements, that is, six combinations of ccr and mec, have been reported (Table 1). These six SCCmec elements have been further classified by differences in regions other than ccr and mec, which are designated junkyard (J) regions. The J regions comprise three parts: J1 (the region between ccr and the right-flanking chromosomal region), J2 (the region between mec and ccr), and J3 (the region between orfX and mec). The J regions are not always specific to each SCCmec type, but certain J regions are commonly shared among certain types of SCCmec elements. Of the three regions, we regard J1 as being the most fundamental, because we presume that it reflects the original form of SCC into which a mec gene complex integrated. Moreover, several different J1 regions have been identified in type II and type IV SCCmec elements (7, 15, 17, 20, 23). The presence or absence of integrated plasmids encoding drug resistance genes in the J3 regions of SCCmec elements can also be used as markers to classify SCCmec elements further in epidemiological studies (1, 19).
|
Tn554) and plasmids (pUB110 or
pT181). |
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Nomenclature of SCCmec elements. In this paper, we use our proposed nomenclature for SCCmec elements (2). As shown in Table 1, the SCCmec element type (defined by the combination of ccr and mec allotypes) is indicated with roman numerals, while the SCCmec element subtype (defined by differences in the junkyard regions) is indicated with Arabic numbers separated by a period, where each number indicates the structure of the J regions (J1, J2, and J3, respectively) according to the chronological priority of the description. It should be noted that the description of the J regions has been modified from our original proposal according to the suggestion of Kunyan Zhang (Calgary University, Canada).
M-PCRs. Chromosomal DNA was extracted from MRSA strains by using the small-scale phenol extraction method and was used as a template (8).
The primer pairs used for PCR experiments are listed in Table 2. M-PCR 1 for ccr type assignment contained two primers to identify mecA and eight primers used for the identification of five ccr genes: four primers including a common forward primer (common to ccrB1-3) and three reverse primers specific for ccrA1, ccrA2, and ccrA3 for identifying ccr1-3 based on the differences in ccrA genes; two primers for identifying ccr4; and two primers for identifying ccr5. M-PCR 2 for mec class assignment contained four primers to identify the gene lineages of mecA-mecI (class A mec), mecA-IS1272 (class B mec), and mecA-IS431(class C mec). M-PCR 3 contained five primer pairs: one pair for identifying specific ORF in the J1 region of type I SCCmec elements and four pairs for identifying specific ORFs in the J1 regions of four subtypes of type IV SCCmec elements. M-PCR 4 contained six primer pairs: four pairs for identifying specific ORFs in J1 regions of four subtypes of type II SCCmec elements, one pair for identifying specific ORFs in the J1 region of type III SCCmec elements, and one pair for identifying specific ORFs in the J1 region of type V SCCmec elements. M-PCR 5 contained three primers: one primer specific to the J2 regions of type II and type III SCCmec elements and two primers specific to ermA and cadB, respectively, to identify the J2 regions of type II or type III SCCmec elements. M-PCR 6 contained three primers: one primer specific for mecA and two primers specific for ant(4') in plasmid pUB110 and tetK in plasmid pT181, respectively, to identify integrated plasmids in the J3 regions.
|
PCRs for the identification of SCCmercury and customary (nonmultiplexed) PCRs to identify pls were carried out with the primer pairs listed in Table 2 according to a previously described procedure (10). PCR products were visualized by agarose gel electrophoresis.
PCR-based identification and determination of part of the nucleotide sequence of the type II.4 SCCmec element. DNA fragments encompassing the entire SCCmec element of strain RN7170 were amplified by long-range PCR using the Expand High Fidelity PCR system under the same conditions as those used for M-PCR 6. The primer sets used for amplifying the DNA fragments are given in Table 2. The amplicon sizes estimated by agarose gel electrophoresis are as follows: the region from orfX to mecA, amplified with primers cR1 and mA3, was 11 kb; the region from mecA to ermA in Tn554, amplified with primers mA2 and ermA1, was 11 kb; the region from ermA in Tn554 to ccr, amplified with primers ermA3 and 2AJ1, was 12 kb; and the region from ccr to the right-flanking chromosomal region, amplified with primers cß and cL4, was 15 kb. The locations of the primers are indicated in Fig. 1. PCR products were purified with the QIAquick PCR purification kit (QIAGEN, Hilden, Germany), and nucleotide sequences of the DNA fragments from ccr to the right-flanking chromosomal region were determined by primer walking.
|
|
RESULTS AND DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
Development of two M-PCRs for J1 regions of SCCmec elements. M-PCRs 3 and 4 were developed to identify specific ORFs in the J1 region of each SCCmec element: M-PCR 3 for type I and type IV SCCmec elements carrying class B mec and M-PCR 4 for type II and type III SCCmec elements carrying class A mec and type V SCCmec elements carrying class C mec.
With M-PCR 3, the J1 regions of all five SCCmec type I or IV elements (type I.1, type IV.1, type IV.2, type IV.3, and type IV.4) (Table 1) were identified with primer pairs specific for each subtype. The sizes of the amplified DNA fragments matched those predicted for the J1 regions of the five SCCmec elements (type I.1, 154 bp; type IV.1, 458 bp; type IV.2, 726 bp; type IV.3, 259 bp; type IV.4, 1,242 bp) (Fig. 2C and Table 2).
In this study, we identified a new subtype of type II SCCmec elements carried by RN7170. By amplifying entire SCCmec region and determining the nucleotide sequence of the J1 region of the element, we designated it type II.4 SCCmec. Interestingly, the J1 region carried by type IIB, IIC, IID, and IIE SCCmec elements was the same as that of the type IV.2 SCCmec element (23). Therefore, we considered these SCCmec elements to be type II.3, type II elements carrying the third identified J1 region (although it was identified previously in the type IV.2 SCCmec). With M-PCR 4, the J1 regions of six SCCmec types, type II, III, or V (type II.1, type II.2, type II.3, type II.4, type III.1, and type V.1) (Table 1), were identified with primer pairs specific for each type. The sizes of the amplified DNA fragments matched those predicted for the J1 regions of all six SCCmec types (type II.1, 287 bp; type II.2, 1,518 bp; type II.3, 726 bp; type II.4, 2,003 bp; type III.1, 503 bp; type V.1, 1,159 bp) (Fig. 2D and Table 2).
Development of two M-PCRs to identify resistance plasmids/transposons integrated into SCCmec elements (J2 and J3 regions).
M-PCR 5 was developed to identify
transposon Tn554 or Tn554 integrated into
type II and type III SCCmec elements by targeting ORFs located
in the J2 region flanking these elements and resistance
determinants carried by the transposons (ermA by
Tn554 and cadB by Tn554). As
expected, M-PCR 5 amplified DNA fragments of 2,756 bp (Fig.
2E, lanes 1 and 2),
corresponding to CN030-ermA, and 1,540 bp (lane 3),
corresponding to CZ021-cadB.
M-PCR 6 was developed to identify differences in the J3 region based on the presence or absence of resistance plasmids pUB110 and pT181 by amplifying the J3 regions mecA-ant(4')-1 for pUB110 and mecA-tetK for pT181. As expected, the DNA fragments amplified by M-PCR 6 were 4,952 bp (Fig. 2E, lane 4) for mecA-ant(4')-1 and 7,406 bp (Fig. 2E, lane 5) for mecA-tetK. It should be noted that with M-PCR 6, plasmids pUB110 and pT181 could be detected if they were located downstream of mecA but could not be detected if they were located distant from mecA. In contrast, in the M-PCR developed previously by Oliveira and Lencastre (20), these plasmids were identified with primer sets specific for each plasmid, regardless of their relative position to mecA, leaving the possibility that these plasmids are located outside the SCCmec element.
Validation of six multiplex PCRs. We evaluated our system of six M-PCRs by examining a total of 99 MRSA strains. The results obtained with the M-PCRs were identical to those obtained by traditional methods, which mostly used a set of two primers to identify each gene or different allotypes (7, 10, 18).
M-PCR 1 amplified a single DNA fragment belonging to one of five ccr types from the chromosomal DNAs of 83 of the 99 MRSA strains. In 10 strains, two DNA fragments were amplified, signifying that these strains carry two ccrs, while in 6 strains, no DNA fragments were amplified (Table 3). Among the 10 strains that carried two ccrs, 8 carried a type 3 ccr and ccrC and 2 carried a type 1 ccr and ccrC. Since ccrC is present in SCC elements carrying the mercury resistance operon (SCCmercury) or the capsule gene cluster (SCCcap1) (2, 16), we conducted an M-PCR experiment to identify the mercury resistance operon and J region in SCCmercury by using four primers (merA2, merG, mN21, and mN22) listed in Table 2. DNA fragments indicating the carriage of the mercury resistance operon were amplified from the chromosomal DNAs of all 10 strains, and DNA fragments indicating the carriage of the J region of SCCmercury were amplified from the chromosomal DNAs of 9 of them, signifying that these 9 strains carried SCCmercury. We tentatively regarded the remaining strain as a type 1 ccr strain, presuming that ccrC might be carried by other unknown mobile genetic elements, since no SCCmec element carrying the combination of ccrC (type 5 ccr) and class A mec or class B mec has been identified yet.
|
As such, SCCmec elements carried by 93 of the 99 tested strains were classified into one of the six known types of SCCmec elements (Table 3).
With M-PCRs 3 and 4, these SCCmec elements could be further classified based on differences in the J1 region. Overall, we were able to classify the J1 regions of SCCmec elements carried by 92 of 93 strains that had an identified SCCmec type (Table 3). Interestingly, no DNA fragment from the chromosomal DNAs of the six untypeable strains was amplified by M-PCR 3 and M-PCR 4, suggesting that these strains might carry new SCCmec elements.
We used M-PCR 5 to establish
the presence or absence of Tn554 and Tn554
in the J2 regions of type II and type III SCCmec elements.
Tn554 was identified in 36 of 37 type II SCCmec
elements and 1 of 12 type III SCCmec elements.
Tn554 was identified in 11 of 12 type III
SCCmec elements (Table
3). One type III
SCCmec strain carried Tn554 instead of
Tn554 at the J2 region, as previously reported
(9).
We used M-PCR 6 to determine the presence or absence of plasmids pUB110 and pT181. Plasmid pUB110 was carried by 1 of 17 type I SCCmec elements, 35 of 37 type II SCCmec elements, 10 of 27 type IV SCCmec elements, and 3 of the 6 untypeable strains. Remarkably, all 10 tested gentamicin-susceptible MRSA strains (isolated in France in 1996) carried a type IV.3 SCCmec with an integrated plasmid pUB110. No type III SCCmec harbored pUB110 downstream of mecA. In contrast, 8 of 12 type III SCCmec elements harbored plasmid pT181, which was not identified in type I, type II, and type IV SCCmec elements.
Atypical and unclassifiable SCCmec elements. We were not able to identify the ccr gene of six mecA-positive strains with M-PCR 1. Two of them carried class A mec, the J2 region of the type II SCCmec element, and the integrated plasmid pUB110, indicating that the gene lineage pUB110-IS431-mecA-mecR1-mecI-Tn554, which is usually located in the type II SCCmec element, was carried by these two strains. Further studies to determine the nucleotide sequences of the region between Tn554 and the right-flanking chromosomal region will clarify whether these strains carry a deleted type II SCCmec or a novel region with new ccr genes. Three of them carried the mecR1 gene, whereas the mecI gene was not detected by PCR testing for the respective genes (18). One strain carried neither mecR1 nor mecI. The structures of those elements will be the subject of further investigation.
Comparison to previously reported M-PCRs. Our M-PCRs do not conflict with two previously reported M-PCRs based primarily on the identification of junkyard regions (20, 26). The M-PCR described previously by Oliveira and Lencastre (20) has the advantage that it identifies multiple loci simultaneously (e.g., the mecA gene, the mecI gene, the J1 region of type I and type II SCCmec elements, ccrC, the dcs [downstream constant sequence] region, pT181, and pUB110). The M-PCR described previously by Zhang et al. (26) has the advantage that it identifies the J1 region of eight SCCmec elements simultaneously. M-PCRs 3 to 6, which were developed to identify specific ORFs in J regions, are based on the same concept as that of the two previously reported M-PCRs.
We first used the pls (plasmin-sensitive protein) gene to identify the J1 region of the type I.1 SCCmec, but that approach was changed since only 12 of 17 strains were positive for this gene (6). In addition, the pls gene was identified with the chromosomal DNA from a type III SCCmec strain, suggesting that it could not be a specific marker for the J1 region of type I SCCmec elements. Oliveira and Lencastre and Zhang et al. also designed primers specific to the J1 region of the type I SCCmec in regions other than the pls gene. The J region of 16 of 17 type I SCCmec strains was classified as subtype 1 with both the primers described previously by Oliveira et al. and our primers.
The identification of the J1 region of the type III SCCmec element was a bit confusing because we first reported the nucleotide sequence of the type III SCCmec element carried by strain 85/2082 as being the longest one, but this turned out to be a composite of SCCmercury and a type III SCCmec element (as indicated in Fig. 1, the length of the type III SCCmec element is different from that originally reported). Both Oliveira and Lencastre and Zhang et al. happened to design primers on the nucleotide sequence of SCCmercury for the identification of type III SCCmec elements. Therefore, the identification of locus E and locus F in the M-PCR described previously by Oliveira and Lencastre and the identification of the locus used by Zhang et al. indicate the likely presence of SCCmercury and are not specific for type III SCCmec elements.
The identification of the J1 regions of type IV SCCmec elements is similar with our and primers and those described previously by Zhang et al., except for a primer pair used for the identification of a type IV.3 SCCmec element designed by Zhang et al. on a locus outside the type IV.3 SCCmec, designated IE25923.
Prospects for assignment of SCCmec elements. The typing system designed here is not final and should be developed further, since it could not identify every known difference. For example, some differences in the J3 regions of the type IV SCCmec element, such as the carriage of Tn4001 in the type IV SCCmec of strain 81/108, the different J3 region structures in type IVE and type IVF SCCmec elements reported previously by Shore et al. (23), and the presence of dcs in the type III SCCmec element reported previously by Chongtrakool et al. (2), were not identified with our system. Although the structure of the J1 region is rather specific to each type and correlates well with the SCCmec type, we want to emphasize that the identification of the J region did not correlate exactly with the type of SCCmec element. The type II.3 MRSA strain isolated in Ireland is a good example: if only the multiplex PCR identifying the J1 region had been conducted, it would have been classified as type IV.2.
It is not easy to conduct six M-PCRs for every case. We suggest that M-PCRs 1 and 2, for identifying ccr and mec, should be conducted first to assign types of SCCmec elements, and they might be enough in most of the cases for epidemiological purposes. In cases where further typing is required, we suggest to proceed with identification of the J1 region structure with M-PCR 3 or 4 or with M-PCRs developed by Zhang et al., since the structure of J1 region might reflect the structure of an SCC in which the mec gene complex was integrated. The remaining two M-PCRs, M-PCR 5 and M-PCR 6, should be conducted, if necessary, for additional typing.
Although it might be difficult to determine all SCCmec types carried by staphylococci, the determination of as many unknown SCCmec types as possible would be of help for epidemiological studies as well as for inferring the origins of MRSA strains. Further discussion is needed in order to form a consensus among staphylococcal researchers regarding how to define SCCmec element types and how to assign SCCmec elements to cope with the ever-increasing diversity of SCCmec elements.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Published ahead of print on 16 October 2006.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Clin. Vaccine Immunol. | Clin. Microbiol. Rev. |
|---|---|
| J. Clin. Microbiol. | ALL ASM JOURNALS |
