Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, May 2007, p. 1671-1677, Vol. 51, No. 5
0066-4804/07/$08.00+0 doi:10.1128/AAC.00978-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Department for Microbiology and Virology, Institute of Medical Biology, University of Tromsø, Tromsø, Norway
Received 7 August 2006/ Returned for modification 30 October 2006/ Accepted 7 February 2007
|
|
|---|
|
|
|---|
SCCmec integrates into the chromosome in a site-specific manner and unique orientation and is also precisely excised from the chromosome (10, 16). Integration of SCCmec is sequence specific at a unique site (attBSCC) located near the S. aureus origin of replication (12). attBSCC is found downstream of an open reading frame of unknown function, designated orfX, that is well conserved among clinical strains of S. aureus. SCCmec contains conserved terminal inverted and direct repeats (DRs) at the integration junctions (11, 13). Excision and integration of the SCCmec element have been observed in vivo and were found to be dependent on the products of the ccrAB genes located within the element (10). The ccrAB genes encode recombinases of the invertase/resolvase family (15).
Different SCCmec elements in staphylococci have been classified and characterized according to the combination of two parts: the ccr complex (ccrAB1, ccrAB2, ccrAB3, ccrAB4, and ccrC) and the mec complex (classes A to E) (10, 13, 16, 26). Six different major types of SCCmec (types I to VI) have been reported in staphylococci (10, 13, 16, 27). Also, several SCCmec subtypes, subtypes IIA to E (31) and subtypes IVa to IVg (14, 21, 23, 31), and SCCmec type VT (3) have been reported. In addition to S. aureus, SCCmec has been described in S. epidermidis (8, 24, 34), S. haemolyticus (32), S. hominis (8, 17), and S. warneri (8). Four SCC non-mec types have also been reported (11, 14, 17, 18, 22, 24).
Our aim was to investigate the nature of the SCCmec elements and the cognate insertion sites in a collection of staphylococcal clinical isolates of various species from Norway, with a special emphasis on S. haemolyticus.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Characteristics of S. aureus (n = 4) and CoNS (n = 38) isolates used in this study
|
|
View this table: [in a new window] |
TABLE 2. Oligonucleotide primers and PCR amplicons used in this study
|
![]() View larger version (5K): [in a new window] |
FIG. 1. Gene map of oligonucleotide primer positions for obtaining orfX and att in this study: cL1/mL1 (371 bp), cR6/cR2 (467 bp), orfX lower/orfX upper (385 bp), mR8/cR2 (637 bp), orfXA1/orfXB1 (S. hominis), and orfXA2/orfXB2 (S. aureus and S. warneri). and , primer position and direction of amplification, respectively.
|
2 (ccrAB1; 650 bp), ccrA1F/ccrA1R (ccrA1), ccrB1F/ccrB1R (ccrB1), β2/
3 (ccrAB2; 1,000 bp), ccrA2F/ccrA2R (ccrA2), ccrB2F/ccrB2R (ccrB2), β2/
4 (ccrAB3; 2,000 bp), ccrA3F/ccrA3R (ccrA3), and ccrB3F/ccrB3R (ccrB3) (8) and
F/
R (ccrC) (13). The open reading frames of ccrC6 and ccrC7 were determined by PCR with primer pairs CForw1/
R, CForw2/CRev2, CForw3/CRev3, and CForw4/CRev4 (Table 2). Those strains that were mecA negative were SCC typed based solely on the ccr gene complex sequence. For detection of the presumed functional CcrB1, the PCR primers EndB1 Forw and EndB1 Rev were used (Table 2). Inverse PCR. Inverse PCR was used as a tool for amplification of unknown regions downstream of orfX into attR and of the left junction (attL) of SCC. Only seven strains (representing various species and SCCmec types) were selected for testing by inverse PCR (Table 1; see also Fig. 3). The oligonucleotide primers were designed on the basis of the nucleotide sequences of 10 different reference strains (Table 2). Genomic DNA was isolated with ADVAMAX beads (Edge Bio Systems, Gaithersburg, MD), according to the manufacturer's instructions. Three micrograms of total DNA was digested with restriction enzyme DraI (10 U/µl; Promega), according to the manufacturer's instructions. The digestion was tested by separating the DNA by agarose gel electrophoresis before phenol extraction. Ligation was performed with T4 DNA ligase (50 U/µl; Promega). PCRs were run with the divergent primers attL-IR/attL-DR, orfXA1/orfXB1, and orfxA2/orfXB2 (Fig. 1). The PCR products contained the region of unknown sequence, which could be sequenced by automatic DNA sequencing, as described by Hanssen et al. (8).
![]() View larger version (9K): [in a new window] |
FIG. 3. Alignment of DNA sequences of DRSCC and IRSCC junctions from SCC and SCCmec elements in various CoNS strains. Shown are S. epidermidis 8-23 new SCCmec type, S. hominis 13-27 new SCCmec type, S. epidermidis 6-28 new SCCmec type, S. epidermidis 13-48 new SCCmec type, S. haemolyticus 13-14 new SCCmec type, S. haemolyticus 6-32 SCCmec nontypeable, and S. warneri 8-80 SCCmec type IVE. Capital letters, part of the SCC cassette; lowercase letters, part of the chromosome; black arrows, direction of IRs and DRs; lighter letters, chromosome; bold letters, DR; underlining, IR; DR-L, DR left junction; IR-L, IR left junction; DR-R, DR right junction; IR-R, IR right junction. There is no corresponding attR for strain 6-32. There is no corresponding attL for strain 8-80. N, one of the bases A, G, C, or T. Pairs of slashes indicate intervening omitted sequences.
|
Southern blot hybridization and dot blotting. After PFGE, Southern blot transfer of SmaI-digested genomic DNA to a positively charged nitrocellulose membrane (Roche, Mannheim Germany) by vacuum blotting (Vacugene XL system; Pharmacia Biotech) and Southern blot hybridization were carried out as described previously (8). Total DNA from the following bacteria was used as the template for probe synthesis: S. aureus NCTC 10442 (GenBank accession no. AB033763) ccrAB1, ccrA1, and ccrB1 probes; S. aureus strain N315 (GenBank accession no. D86934) ccrAB2, ccrA2, ccrB2, mecA, mecI, mecR1, attL, and attR probes; S. aureus strain 85/2082 (GenBank accession no. AB037671) ccrAB3, ccrA3, and ccrB3 probes; and S. aureus strain WIS (GenBank accession no. AB121219) ccrC probe. PCR amplicons from PCR amplification of the ccrAB, ccrC, attR, and attL genes were used as probes and were labeled as described by Hanssen et al. (8). Dot blotting was performed by standard techniques (2).
DNA sequencing. The PCR products were purified with the EXO/SAP (Shrimp Alkaline Phosphatase and Exonuclease I) PCR product presequencing kit (USB Corporation). Heterogeneity in the ccr, attR, and attL genes was identified by bidirectional DNA sequencing, as described previously (8).
Computer analyses. The nucleotide sequences and deduced amino acid sequences were edited by using Chromas software (version 2.21) and were aligned by using the BioEdit sequence alignment editor (version 5.0.9) (7). The nucleotide sequences were compared to the sequences in the GenBank database; and the protein sequences were compared to nonredundant GenBank coding sequence translations by using the BLASTN, BLASTP, and BLASTX local alignment search tools (1).
Phylogenetic trees. A phylogenetic tree (rooted) of the ccrC6 and ccrC7 genes (1,677 bp) was generated from the alignment of ccrC with the ccrC genes of other reference strains by using MEGA software (version 2.1) (neighbor joining). The topology of the phylogenetic tree was evaluated by bootstrap analyses with 1,000 replicates to give the degree of confidence intervals for each node on the phylogenetic tree.
Nucleotide sequence accession numbers. The nucleotide sequences were deposited in GenBank under accession numbers EF190467 (ccrC6; 1,677 bp in strain 25-60), EF190468 (ccrC7; 1,677 bp in strain 13-48), EF190469 (partial ccrA2; 927 bp in strain 13-48), and EF190470 (partial ccrB2; 464 bp in strain 13-48).
|
|
|---|
We report 52% (22/42) PCR ccr-untypeable CoNS strains and 1 S. aureus strain. Southern blot hybridization and dot blot hybridization performed with all five known ccr types confirmed the negative PCR results. Qi et al. (29) also reported a large number of untypeable SCCmec variants among MR-CoNS strains. This was quite different from the findings of Oliveira et al. (28), who reported 7 to 8% ccr-untypeable MRSA strains. The ccr-untypeable strains may carry a new deleted SCCmec or novel regions with ccr.
Interestingly, none of the S. haemolyticus strains contained IS1272 in SCCmec within 2 kb from mecA. S. haemolyticus is suggested to be the definite host for IS1272, while it was only secondarily acquired by S. epidermidis and S. aureus (19). Our S. haemolyticus strains that were ccrAB2 positive contained the class C1 mec gene complex. Nineteen of 23 S. haemolyticus strains contained mec complex C1. Among our S. haemolyticus strains, we found seven previously unreported SCCmec types, and most of them contained multiple ccr genes. The new SCCmec type containing ccrAB2 genes (which were 100% identical to ccrAB2 in S. aureus SCCmec types IVc and IVE), and mec complex C1, in combination with ccrC3, ccrC6, or ccrC7, might be evidence for multiple SCCs in S. haemolyticus. Six different SCCs have been reported in one S. haemolyticus strain (32), and this underlines the complexity of the SCC pattern. Proof of the extensive diversity of SCCmec in CoNS became clear when S. haemolyticus strain 8-09 was studied. It did not fit into the description of any of the reported SCCmec types. This strain contained ccrC6 and a mec complex A variant, i.e., a new SCCmec type. In addition, it gave weak ccrAB2 PCR amplification products, but we were unable to sequence the DNA of the ccrAB2 genes. However, this strain gave positive dot blot hybridization signals when probes specific for ccrAB1, ccrAB2, and ccrAB3 were used, indicating that this strain is a candidate for containing novel ccr variants.
Multiple copies of SCC and ccr. We detected multiple copies of ccr in 11 of our strains, i.e., ccrAB1/ccrC7 in S. epidermidis (n = 1), ccrAB2/ccrC6 in S. haemolyticus (n = 2), ccrAB2/ccrC3 in S. haemolyticus (n = 1), ccrAB2/ccrC7 in S. epidermidis (n = 1) and S. haemolyticus (n = 1), ccrAB3/ccrC3 in S. aureus (n = 1), ccrAB3/ccrC7 in S. epidermidis (n = 2) and S. haemolyticus (n = 1), and ccrAB2/ccrAB3/ccrC7 in S. epidermidis (n = 1) (Table 1). To our knowledge, except for the ccrAB3/ccrC3 combination (4), they are all new combinations not reported earlier. Multiple copies of the ccr genes have earlier been reported in S. aureus (3, 20, 29) and S. epidermidis (24). To our knowledge, ccrC and ccrAB have not previously been reported to be present in a single S. haemolyticus strain, while ccrC or ccrAB alone has been observed in S. haemolyticus (13). One limitation of our work is that we do not know which ccr gene is linked to SCCmec, i.e., whether it is the ccrAB or the ccrC gene. Our data indicate that it is the ccrAB gene that is linked to SCCmec, since those isolates containing ccrAB2 also have the molecular feature of SCCmec type IV or II at the attachment sites. The question remains whether ccrC is situated within the SCCmec cassette or outside the cassette in the chromosome. Long-range PCRs between ccrAB-mecA and ccrC-orfX have been performed, but the tests were unsuccessful. Full sequencing of the cassette or cohybridization will probably give us an answer to this question.
Allelic variants of ccrAB and ccrC. We found two new allelic variants of ccrC in our study, referred to as ccrC6 and ccrC7 (Table 1; Fig. 2). We followed the naming system used by Boyle-Vavra et al. (3) and continued to use that system to name the ccrC variants, and we also named ccrC in S. haemolyticus strain JCSC1435 (GenBank accession no. AP006716) ccrC5. The nucleotide sequence identities among the ccrC allelic variants (ccrC1-ccrC7) varied from 87 to 97%. The new variant, ccrC7, showed the closest nucleotide sequence identity (97%) to ccrC3 in SCCmercury, but ccrC3 was 122 bp shorter than ccrC7. ccrC7 also showed 93% nucleotide sequence identity to ccrC2 in SCCmec type VT (3) and ccrC5 in S. haemolyticus (32). ccrC6 showed the closest identity (97%) to ccrC2 in SCCmec type VT. The two new allelic variants formed distinct branches in the phylogenetic tree (Fig. 2). CcrC7 showed 99% and 95% amino acid sequence identities to CcrC3 in SCCmercury (4) and CcrC5 in S. haemolyticus (32), respectively. CcrC6 showed 97% amino acid sequence identity to CcrC2 in SCCmec type VT and CcrC3 in SCCmercury (3, 4). Worth noticing is the fact that both CcrC6 and CcrC7 are one amino acid shorter than CcrC2 in SCCmec type VT. It is not known whether the different CcrC recombinase variants have different activities. New variant ccrC7 was observed in S. epidermidis and S. haemolyticus, while ccrC6 was observed only in S. haemolyticus. Allelic variation within ccrAB in epidemic MRSA clones has been shown earlier (8, 28). It is noteworthy that none of the strains carried ccrC alone, but ccrC was always accompanied by ccrAB. This suggests that these strains have an SCCmec element closely related to the composite type III SCCmec of S. aureus. We found several combinations involving ccrC of SCCmec type V, as reported earlier by others (25). Most likely, the small size and enhanced mobility of SCCmec type V may cause recombination events with ccrC genes from various sources and may thus contribute to the diversity and allelic variation in the ccrC genes.
![]() View larger version (8K): [in a new window] |
FIG. 2. Phylogenetic tree (neighbor joining) showing the allelic variants of ccrC6 and ccrC7 (1,677 bp) compared with previously reported ccrC sequences: ccrC1 in S. aureus SCCmec type V (GenBank accession no. AB121219), ccrC3 from S. aureus SCCmec type III (GenBank accession no. AB037671), truncated ccrC from S. aureus SCCmec type III (GenBank accession no. AB047089), ccrC4 from S. aureus SCCcap1 (GenBank accession no. U10927), ccrC2 from S. aureus SCCmec type VT (GenBank accession no. AY894416), ccrC (ccrC5) from S. haemolyticus SCCmec type V (GenBank accession no. AP006716), ccrC6 of 1,677 bp in strain 25-60 (GenBank accession no. EF190467), and ccrC7 of 1,677 bp in strain 13-48 (GenBank accession no. EF190468). The outgroup was ccrB3 from S. aureus SCCmec type III (GenBank accession no. AB037671). The scale bar represents nucleotide differences.
|
S. hominis strains 13-27 and 13-33 harbored ccrAB1, which supposedly encodes a functional version of ccrB1, and a class A mec gene complex. They were therefore designated new SCCmec types. Also, mecA-negative S. hominis strain 8-39 and S. epidermidis strain 6-42 contained the presumed functional version of ccrB1 (Table 1). This is the first report of mecA-negative S. epidermidis and mecA-negative and mecA-positive S. hominis strains harboring the original version of ccrB1. The presence of the same allelic variant of ccr and the same SCCmec type in different species suggests horizontal gene transfer across the species barrier.
Attachment sites. The strains were first tested for att-flanking regions of SCCmec with primers cR2/cR6, orfX upper/orfX lower, cR2/mR8, cL1/mL1, and attL forw/attL rev; but so few strains gave positive PCR amplification products that we chose seven strains representing various SCCmec types and species and examined them by inverse PCR and sequencing (Fig. 3). The left and right boundaries were determined by comparing their nucleotide sequences with those of previously reported SCC elements.
We found DRs of 15 bp each and incomplete inverted repeats (IRs) IR-L and IR-R, composed of 10 to 28 bp at both extremities. This finding is in accordance with the observations made by Ito et al. (12). Only minor differences in DRs were observed, while there was more variation in the IRs. Interestingly, the new SCCmec type of S. epidermidis strain 8-23 (ccrAB1, ccrC7, mec complex B) contained an IR of only 10 bp, which is different from the IRs in strains harboring SCCmec type I, which had IRs of 26 to 27 bp (Fig. 3); but an IR of 7 bp in SCCcap1 has been reported (22). However, the ccrAB1 genes were identical to the deleted version of ccrAB1 in NCTC 10442 (SCCmec type I). This may indicate that strain 8-23 contains several types of unrecognized ccrAB genes or that the CcrAB1 enzymes are able to recognize IRs typical for the integration of both SCCmec type I and SCCmec type V elements. IRs are believed to act as recognition sites for transposases (5), Cre (6), and the Sin recombinase in S. aureus (30). Another possibility is that strain 8-23 once contained two separate SCC elements, as reported for S. aureus SCCmec type III-SCCmercury (4, 10, 11). This could actually be the case, since the strain contains both ccrAB1 and ccrC7.
In S. warneri strain 8-80 (SCCmec type IVc) only attR was amplified by PCR. DR-R (15 bp) was 1 bp different from SCCmec type IVc, but the DR was incomplete since we did not have the left side for comparison. The IR was suggested to consist of 28 bp (which was 100% identical to SCCmec types II and IVc) (Fig. 3). Our data indicated that attR is more conserved between the different species than attL. It was more difficult to obtain attL, maybe due to major rearrangements in this region, as reported by Luong et al. (22). Perhaps strain 8-80 is devoid of the attL IR element, as reported for S. aureus SCCcap1 (13).
S. epidermidis strain 13-48 contained a new SCCmec type carrying three different ccr genes, and thus, the attachment sites were difficult to define. However, we found attachment sites that were most similar to that found in SCCmec type III. The 28-bp IR-R of S. hominis strains 13-33 and 13-27 (new SCCmec type, ccrAB1, mec complex A) (8) was identical to the IR-R in S. hominis ATCC 27844, while IR-L was 2 bp different from that in strain ATCC 27844. It does not seem as if an original or deleted version of ccrB1 affects the specificity for the attachment site.
S. haemolyticus (new SCCmec, ccrAB2, mec complex C1) had attachment sites with DRs of 15 bp identical to the DRs in S. aureus SCCmec types II and IV and IRs of 26 bp with only minor differences compared to the IRs in SCCmec types II and IV. These observations confirm the hypothesis that the attachment site and ccr type are closely linked.
Concluding remarks. The species-independent conservation of SCCmec in this study suggests horizontal gene transfer between staphylococci. CoNS are more likely to contain several representatives of different ccr complexes and thus seem to harbor multiple or composite new types of SCCmec. The enormous diversity in SCC elements that we observed implies that the SCCmec reservoir is large in CoNS. It may perhaps be worthwhile to focus on CoNS, since they may be the breeding ground for new SCCmec elements.
We thank K. Hiramatsu of Juntendo University (Tokyo, Japan), P. Gaustad of Rikshospitalet, University Hospital (Oslo, Norway), and the Department of Microbiology, University Hospital of North Norway, for kindly providing strains. We thank T. Tessem and L.-H. Henriksen for excellent technical assistance.
Published ahead of print on 16 February 2007. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»