Clin Microbiol Infect 2012; 18: 395–400
The recent finding of a new mecA homologue, mecALGA251, with only 70% nucleotide homology to the conventional mecA gene has brought the routine testing for mecA as a confirmatory test for methicillin-resistant Staphylococcus aureus (MRSA) into question. A multiplex PCR was designed to differentiate mecALGA251 from the known mecA together with detection of lukF-PV and the spa gene fragments, enabling direct spa typing by sequencing of the PCR amplicons. The PCR analysis and subsequent spa typing were validated on a large collection (n = 185) of contemporary MRSA and methicillin-sensitive S. aureus isolates, including 127 isolates carrying mecALGA251. The mecALGA251 gene was situated in staphylococcal cassette chromosome mec type XI elements, and sequence variation within a 631-bp fragment of mecALGA251 in 79 isolates indicated a very conserved gene sequence. Following a successful validation, the multiplex PCR strategy was implemented in the routine testing of MRSA for national surveillance. Over a 2-month period, among 203 samples tested, 12 new MRSA cases caused by isolates carrying mecALGA251 were identified, emphasizing the clinical importance of testing for these new MRSA isolates.
Following its emergence in the 1960s, methicillin-resistant Staphylococcus aureus (MRSA) has been a major threat, first to hospitalized patients, but also, over the last decade, increasingly to patients outside healthcare facilities, owing to community-associated MRSA. Since its discovery, detection of the mecA gene encoding penicillin-binding protein 2a, which confers resistance to all β-lactams, has been the reference standard for detection of MRSA. A novel mecA homologue (mecALGA251) was recently reported. It has been found in S. aureus lineages typically associated with cattle, i.e. clonal complex (CC)130, CC1943, and sequence type 425 . The mecALGA251 gene shares only 70% nucleotide homology with mecA, and is found in a novel staphylococcal cassette chromosome (SCC) element designated SCCmec type XI. Thus far, isolates containing mecALGA251 have proved to be resistant phenotypically to β-lactams, but have failed to be recognized as ‘classical’ MRSA with conventional PCRs for mecA, owing to the different nucleotide composition [1,2]. The initial detection method for mecALGA251 was a non-discriminatory PCR co-detecting both mecA and mecALGA251 with the same amplicon length, requiring an additional conventional mecA PCR to distinguish the mecALGA251 genotype, making the method less useful for reference laboratory use . Specific primers for mecALGA251 were recently tested on 11 CC130 isolates from Germany, indicating the need for a specific method for detection of mecALGA251 in a single PCR reaction . The National Staphylococcus Laboratory at Statens Serum Institut (SSI) holds two major S. aureus strain collections: (i) all Danish MRSA or suspected MRSA isolates found since 1988; and (ii) all Danish S. aureus bacteraemia isolates found since 1958. All isolates have been typed using different methods (phage typing, pulsed-field gel electrophoresis, and/or spa typing), and all MRSA isolates have been confirmed by mecA detection, providing a valuable archive from which putative mecALGA251-positive isolates can be retrieved. All isolates submitted after November 2006 have been subjected to a triplex PCR detecting mecA, spa, and lukF-PV, with subsequent sequencing of the spa amplicons . To facilitate rapid detection and characterization of isolates harbouring mecALGA251, the triplex PCR has been expanded to include mecALGA251-specific primers. Herein, we report the design, validation and utility of the revised assay with a large (n = 185) and diverse set of both MRSA and methicillin-sensitive S. aureus (MSSA) isolates, including 127 mecALGA251-positive isolates. The mecALGA251-positive isolates were investigated in detail, including spa typing, partial analysis of SCCmec, and nucleotide variation in the mecALGA251 sequence.
Materials and Methods
Danish S. aureus isolates from 1975 onwards were selected for this study on the basis of: (i) possession of spa types identical or similar to those previously reported in mecALGA251-positive isolates (t742, t843, t978, t1535, t1736, t6220, t6292, t6293, t6300, t6386, t7485, t7734, t7945, t7946, or t7947); or (ii) possession of an MRSA phenotype but negative results for mecA with conventional PCR . These isolates, together with 148 mecA-negative isolates with other spa types having cefoxitin zone diameters just above the set breakpoint for MRSA detection with disk diffusion antimicrobial susceptibility testing (MSSA), were subjected to mecA/mecALGA251 PCR as described by Garcia-Alvarez et al. .
All mecA-negative and mecALGA251-positive isolates identified were spa-typed, and subjected to multilocus sequence typing (MLST) if new spa types, not previously associated with CC130 or CC1943 isolates, were found .
mecALGA251 detection and characterization
The primer sequences used in this study are listed in Table 1. All amplifications were performed on supernatants from crude DNA extracts; 1 μL of bacteria boiled in 300 μL of dH2O for 5 min and subsequently centrifuged for 5 min at 13 000 g. Initially, degenerate primers for mecA/mecALGA251 were used to verify the presence of mecALGA251 in all 127 mecA-negative isolates, as previously described . Novel primers (mecALGA251MultiFP and mecALGA251MultiRP) discriminating mecALGA251 (138 bp) from mecA (162 bp) were designed on the basis of the sequence of LGA251 (accession no. FR821779), and were implemented in a triplex PCR detecting spa, mecA, and lukF-PV . Briefly, 0.4 μM of each primer was added to a Qiagen Multiplex PCR Master kit (Qiagen, Valencia, CA, USA), and amplification was performed with the following program: 15 min at 94°C, followed by 30 cycles of 30 s at 94°C, 1 min at 59°C, and 1 min at 72°C, with a final 10 min elongation step at 72°C. PCR products were visualized on 2% E-Gels (Invitrogen, Grand Island, CA, USA).
|Primer names||Gene||Sequence (5′–3′)||Amplicon size (bp)||References|
|mecALGA251MultiFP||mecALGA251||GAAAAAAAGGCTTAGAACGCCTC||718||This study, |
Validation of the PCR was performed on 127 mecA-negative, mecALGA251-positive isolates, originating from Denmark (n = 84), England (n = 28), Scotland (n = 12), and France (n = 3); of these, 28 English isolates (13 human and 15 bovine), 12 Scottish and 24 Danish isolates (all human, spa type t843) have been described previously . Also included were 20 MSSA isolates (t002, t012, t015, t021, t065, t084, t091, t127, t216, and t230) and 30 MRSA isolates (t002, t008, t019, t024, t032, t034, t041, t044, t127, and t437) representing the predominant spa types found in Denmark in 2009 and 2010, respectively, as well as isolates representing SCCmec types I–VI, IX and X [7,8].
Routine testing with the new multiplex PCR
From 1 August 2011, the new multiplex PCR was implemented as a routine test and applied to a total of 203 suspected MRSA isolates submitted from all 15 Danish Clinical Microbiological Departments.
mecALGA251 sequence variation and SCCmec characterization
To search for evidence of sequence variation, amplification of a 720-bp fragment of mecALGA251 was performed with 79 isolates (England (n = 9), France (n = 3), and Denmark (n = 67)), representing CC130, CC425, and CC1943. PCR was performed with 0.4 μM of each primer (mecALGA251MultiFP and mecALGA251RP), 2 mM MgCl2 and the Qiagen Multiplex PCR Master kit, with the following program: 15 min at 94°C, followed by 35 cycles of 30 s at 94°C, 1 min at 50°C, and 1 min at 72°C, with a final 10 min elongation step at 72°C. DNA sequencing was performed by GATC-Biotech (Konstanz, Germany). Sequences were analysed with the CLC Genomics workbench 4.7.2 software (CLC bio, Aarhus, Denmark).
The SCC element was characterized with three PCRs to detect mecI/mecR1, ccrA/B and blaZ, with the SCCmec type XI element defined as described by Garcia-Alvarez et al. . All mecALGA251-positive isolates that were negative for ccrA and/or ccrB, as described above, were subjected to multiplex PCR for detection of other known ccr alleles .
Investigation of the collections at SSI for isolates with spa types identical or similar to the previously reported types associated with mecALGA251, or for isolates that were MRSA phenotypically but mecA-negative, revealed 84 isolates from individual patients. Of these, 40 were cefoxitin-resistant but mecA-negative; the remaining 44 (also cefoxitin-resistant) were identified primarily on the basis of their spa types, which were related to those associated with carriage of mecALGA251 previously.
Except for two bacteraemia isolates from 1975 (t843) and 1992 (t9397), all Danish isolates were recovered in the period between January 2004 and 1 October 2011.
Typing of isolates harbouring mecALGA251
The Danish isolates harbouring mecALGA251 could be grouped into two MLST CCs, CC130 (n = 71) and CC1943 (n = 13). Among the CC130 isolates, 12 different spa types were identified, dominated by t843 (n = 48). The remainder comprised related spa types as follows: t528 (n = 7), t1535 (n = 3), t1773 (n = 3), t1048 (n = 2), t3256 (n = 2), t1532, t1736, t3218, t3570, t5970 and t9397. Among the CC1943 isolates, t978 (n = 6) and the related t2345 (n = 4) and t3391 (n = 2), and a new spa type, t8835, were detected. Thirty-three UK isolates and three French isolates belonged to CC130 (t843, n = 24, t7485, n = 3; t1535 and t6293, n = 2 each; t6220, t6292, t7734, t7946 and t7947). Two UK isolates belonged to CC1943 (t7945), and six English isolates were related to a genetic lineage not found in Danish isolates (CC425), with spa types t742 (n = 2), t6292 (n = 2), t6300, and t6386. A subset of eight Danish strains were typed by MLST from the two spa clusters, and confirmed the CC130 and CC1943 clustering.
The multiplex protocol was validated on 185 isolates: (i) 127 phenotypic MRSA isolates (mecA-negative and mecA/mecALGA251-positive); (ii) 38 conventional MRSA isolates (mecA-positive and mecALGA251-negative); and (iii) 20 MSSA isolates (mecA/mecALGA251-negative). The sensitivity and specificity were 100% for identification of the three groups. Furthermore, spa amplicons from isolates containing only one spa repeat (t528) were distinguishable from both the mecA and mecALGA251 PCR fragments when analysed by gel electrophoresis (Fig. 1).
In addition, direct sequencing of the spa PCR products was not affected by the presence of the extra mecALGA251 amplicon in the sequencing reaction, and reliable spa sequences were obtained.
Routine testing with the new multiplex PCR
Twelve isolates harbouring mecALGA251 were identified among 203 isolates submitted for genetic confirmation of MRSA phenotype in a 2-month period. All were successfully spa-typed with the new protocol, and this revealed no other spa types than those included in the initial validation set; t843 was predominant.
Sequence analysis of mecALGA251 and SCCmec characterization
Sequence data from mecALGA251 (631 bp: positions 36 827–36 197 of LGA251) from 79 isolates were analysed to investigate sequence variation of the gene, and aligned with ten sequences of mecA representing all published SCCmec types (data not shown). All mecALGA251 fragments clustered with the LGA251 sequence. Only six single-nucleotide polymorphisms (SNPs) were observed among the mecALGA251 fragments sequenced, including two silent third-base mutations (pos. 36821 T>C; pos. 36235 C>T) and four conservative missense mutations (Ala346Ser, Tyr363Phe, Arg466His, and Asp482Asn). These SNPs did not correlate with spa types or differences in ccr alleles.
Additional PCRs detecting mecI/mecR, ccrA/B and blaZ indicated that mecALGA251 was contained within SCCmec type XI elements. ccrA and ccrB amplifications were, however, not obtained for eight CC130 isolates, and ccrB was not detected in a further 18 isolates (16 CC130 and two CC1943) with the primers of Garcia-Alvarez et al. The ccrA/B genes from these 26 isolates were also not detected with primers designed to detect other ccrAB allele variants , and no correlation with specific spa types was observed (Table 2).
|CC||No. of isolates (n = 185)||mecA||mecALGA251||SCCmec typea||ccrA/B (n = 79)||spa types||MLST (ST) (n = 52)||Years||Country|
|CC130||107||0||107||XI||A (n = 16), AB (n = 40), negative (n = 8)||t528, t843, t1048, t1532, t1535, t1736, t1773, t3218, t3256, t3570, t5970, t6220, t6293, t7485, t7734, t7946, t7947, t9397||(n = 42) 130, 1245, 1526, 1945||1975, 1992, 2002, 2004–2011||DK, FR, UK|
|CC425||6||0||6||XI||AB (n = 2)||t742, t6292, t6300, t6386||(n = 6) 425||2006–7, 2009–10||UK|
|CC1943||14||0||14||XI||A (n = 2), AB (n = 11)||t978, t2345, t3391, t7945, t8835||(n = 4) 1943, 1946, 2173, 2174||2004–2008, 2010–2011||DK, UK|
|CC1, CC5, CC8, CC15, CC30, CC45, CC59||20||0||0||–||ND||t002, t012, t015, t021, t065, t084, t091, t127, t216, t230||ND||2011||DK|
|CC1, CC5, CC8, CC22, CC59, CC80, CC398||30||30||0||ND||ND||t002, t008, t019, t024, t032, t034, t041, t044, t217, t437||ND||2011||DK|
|Other||8||8||0||SCCmec types I–VI, IX, X||ND||–||–||–||–|
Genotypic detection of mecA is widely used as a reference standard for identification of MRSA, whether used as a primary test or for confirmation . Phenotypically resistant (cefoxitin/oxacillin) isolates that lack mecA when tested by PCR are often identified as borderline oxacillin-resistant S. aureus (BORSA) or, more rarely, moderately resistant S. aureus (MODSA) strains. Following the discovery of mecALGA251, the main genetic basis for the MODSA MRSA phenotype may have been resolved. Using the recently published PCR detection assay  to confirm the presence of mecALGA251, we analysed BORSA and MODSA isolates at the national S. aureus reference laboratory in Denmark, as well as isolates with spa types identical or similar to those of isolates previously reported as being associated with the mecA homologue. This revealed a total of 84 mecALGA251-positive isolates, including the 24 reported recently .
On the basis of the complete genome sequence of S. aureus LGA251 (GenBank accession number FR821779), Garcia-Alvarez et al. designed a PCR to detect both mecA and mecALGA251, resulting in a single undifferentiated amplicon. In this study, a method was derived based on the LGA251 genome sequence to facilitate the detection and characterization of isolates harbouring mecALGA251 by expanding a previously published multiplex PCR (targeting mecA, spa, and lukF-PV) to include mecALGA251-specific primers. This method showed 100% specificity and sensitivity for both mecA targets when tested on 185 isolates. The PCR amplicons were readily visualized by the use of agarose electrophoresis to identify the various combinations of detected genes (Fig. 1). Sequencing of the spa PCR products was not hindered by the extra mecALGA251 amplicon in the sequencing reaction.
We discovered a number of different spa types associated with the two clonal lineages CC130 and CC1943, together with two novel MLST types, both related to CC1943. The characterization of the SCC elements in 79 isolates indicated that mecALGA251 in all isolates was contained within an SCCmec type XI element. However, some variation was observed in the PCR results with regard to detection of the ccrA/B alleles. On the basis of the other results indicating the consistent presence of SCCmec type XI, we speculate that the ccrA/B PCR-negative results are most likely to result from minor sequence differences in the SCCmec type XI-specific ccr alleles.
To investigate the sequence variation of mecALGA251, sequence analysis was performed for a 631-bp fragment (positions 36 827–36 197 of LGA251) from a total of 79 isolates, including Danish, French and UK isolates to increase the possible diversity. The analysis provides evidence indicating that mecALGA251 is highly conserved, thus supporting the existence of a clearly distinct novel mecA homologue.
We identified six SNPs with no apparent impact on the observed phenotypic resistance to cefoxitin, and that did not correlate with spa types, or with differences in ccrA/B detection.
On the basis of 60 days of screening, the finding of 12 mecALGA251-positive isolates among 203 routine isolates provides evidence to suggest that mecALGA25 is potentially significant for the occurrence of MRSA isolates in Denmark.
In conclusion, we present a technique that facilitates rapid co-detection and differentiation of both mecA variants and their Panton–Valentine leukocidin (PVL) locus together with the ability to perform typing downstream by spa sequencing to allow rapid and inexpensive typing and genetic clustering of isolates. We also present the first evidence that mecALGA251 is highly conserved, and that the corresponding SCCmec elements show some variation in the ccrA/B regions without any relation to specific spa types.
This work was carried out with technical assistance from L. R. Hansen, S. F-Madsen, E. Chapka, and J. H. Nielsen. The Danish Clinical Microbiological Laboratories are thanked for submitting S. aureus isolates to SSI.
The authors declare that they have no conflicting interests in relation to this work.