ABSTRACT
Similar to mecA, mecC confers resistance against beta-lactams, leading to the phenotype of methicillin-resistant Staphylococcus aureus (MRSA). However, mecC-harboring MRSA strains pose special difficulties in their detection. The aim of this study was to assess and compare different phenotypic systems for screening, identification, and susceptibility testing of mecC-positive MRSA isolates. A well-characterized collection of mecC-positive S. aureus isolates (n = 111) was used for evaluation. Routinely used approaches were studied to determine their suitability to correctly identify mecC-harboring MRSA, including three (semi)automated antimicrobial susceptibility testing (AST) systems and five selective chromogenic agar plates. Additionally, a cefoxitin disk diffusion test and an oxacillin broth microdilution assay were examined. All mecC-harboring MRSA isolates were able to grow on all chromogenic MRSA screening plates tested. Detection of these isolates in AST systems based on cefoxitin and/or oxacillin testing yielded overall positive agreements with the mecC genotype of 97.3% (MicroScan WalkAway; Siemens), 91.9% (Vitek 2; bioMérieux), and 64.9% (Phoenix, BD). The phenotypic resistance pattern most frequently observed by AST devices was “cefoxitin resistance/oxacillin susceptibility,” ranging from 54.1% (Phoenix) and 83.8% (Vitek 2) to 92.8% (WalkAway). The cefoxitin disk diffusion and oxacillin broth microdilution assays categorized 100% and 61.3% of isolates to be MRSA, respectively. The chromogenic media tested confirmed their suitability to reliably screen for mecC-harboring MRSA. The AST systems showed false-negative results with varying numbers, misidentifying mecC-harboring MRSA as methicillin-susceptible S. aureus. This study underlines cefoxitin's status as the superior surrogate mecC-positive MRSA marker.
INTRODUCTION
The still worrying occurrence of methicillin-resistant Staphylococcus aureus (MRSA) in many parts of the world poses a major challenge to health care systems by increasing the burden of disease. Rapid and effective MRSA identification and susceptibility testing are paramount to prevent further dissemination and to adapt antimicrobial treatment. In 2011, a novel PBP 2a-encoding mecA homologue designated mecC (originally mecALGA251) was reported with homologies on the nucleotide and protein levels of only 70% and 63%, respectively (1, 2). Later mecC was confirmed as the genetic determinant that confers methicillin resistance in S. aureus for those isolates (3). Farm and wildlife animals have been revealed as reservoirs for mecC-harboring MRSA (4, 5), and the zoonotic potential of these livestock-associated MRSA has been shown (6–8).
The limited homology of mecC to mecA and their respective proteins led to major diagnostic challenges in identification and susceptibility testing of mecC-harboring MRSA (9). In addition to obvious but easily resolved difficulties in targeting the divergent mecC nucleotide sequence by DNA-based diagnostic tests (10, 11), phenotypic approaches exhibited considerable difficulties due to comparatively low oxacillin MICs (1, 7, 8), which may be caused by differences in the mecA and mecC promoters (3). Moreover, low homology between the encoded PBP 2a proteins is the reason for the failure of existing PBP 2a agglutination tests to detect mecC-positive isolates (5, 7, 8).
In this study, we compared several routinely applied diagnostic approaches in their capability to identify mecC-harboring MRSA strains from a comprehensive, heterogeneous, and representative collection. In detail, we compared (i) three (semi)automated susceptibility testing (AST) systems, (ii) five selective chromogenic agar plates (MRSA screening plates), (iii) a cefoxitin disk diffusion test, and (iv) oxacillin broth microdilution.
MATERIALS AND METHODS
A large set of mecC-harboring MRSA isolates (n = 111) from human and animal specimens isolated in Germany, the United Kingdom, and Belgium were included in the study. All isolates were confirmed as mecC positive by PCR (12) and characterized by spa typing (t843, n = 51; t6292, n = 13; t1736, n = 6; t1535, n = 4; t3391, n = 3; t978, t9165, t742, t6902, t6521, t6220, t5930, t1773, and t11706, n = 2 each; t9910, t9738, t9280, t9123, t8842, t7914, t7603, t7189, t6300, t524, t13233, t1207, t11702, t11290, t11120, and not typeable, n = 1 each). Isolates were of human (n = 80), unknown (n = 24), bovine/bulk milk (n = 4), sheep (n = 2), and environmental (n = 1) origins. No copy isolates were included.
The following selective chromogenic agar plates were inoculated with a single colony from overnight blood agar plate cultures: (i) Oxoid Brilliance MRSA 2, (ii) bioMérieux chromID MRSA, (iii) BD BBL CHROMagar MRSA II, (iv) Bio-Rad MRSA Select, and (v) MAST Diagnostica CHROMagar MRSA. To simulate potentially low inocula of clinical specimens, nine isolates with different spa types (t843, t978, t1207, t1535, t1736, t391, t5930, t6292, and t6902) were each adjusted to 0.5 McFarland standard turbidity, and serial dilutions with the final dilution factor of 105 were prepared. Subsequently, 100 μl of the final dilutions was used to inoculate all chromogenic media (except MRSA Select from Bio-Rad due to supply constraints) and blood agar plates for growth control in triplicate. S. aureus strains USA300 and ATCC 29213 were used as positive and negative controls, respectively. Growth was evaluated after 24 h and 48 h. Automated systems were inoculated from the same plates as chromogenic media. Automated systems for susceptibility testing were used according to the manufacturers' recommendations: i.e., BD Phoenix (Becton Dickinson, Heidelberg, Germany) was executed with test panel PMIC-72, Vitek 2 (bioMérieux, Marcy l'Etoile, France) with test panel AST P580, and MicroScan WalkAway 96 Plus (Siemens Healthcare Diagnostics, Eschborn, Germany) with test panel Pos MIC 28.
Cefoxitin disk diffusion assays (cefoxitin discs, 30 μg; bestbion dx, Cologne, Germany) were performed according to EUCAST and using S. aureus ATCC 29213 as control. The EUCAST guidelines (version 7.0, valid from 1 January 2017 [inhibition zone of <22 mm, resistant]) and CLSI criteria (M100-S27, 27th ed., January 2017 [inhibition zone of ≤21 mm, resistant]) were followed in the interpretation of the results.
Oxacillin (Sigma-Aldrich, Taufkirchen, Germany) susceptibility was determined by broth microdilution, using a final inoculum of approximately 5 × 105 CFU/ml and S. aureus ATCC 29213 as quality control. MICs were interpreted according to EUCAST guidelines (version 7.0, valid from 1 January 2017 [MIC of >2 μg/ml]) and CLSI criteria (M100-S27, 27th ed., January 2017 [MIC of ≥4 μg/ml]).
RESULTS
Applicability of AST systems to detect mecC-positive isolates.Analyzing resistance toward cefoxitin and oxacillin by AST systems, different susceptibility patterns were observed. For all systems, the most frequently detected pattern was the combination of the categorization “cefoxitin-resistant, but oxacillin-susceptible,” ranging from 54.1% (Phoenix) and 83.8% (Vitek 2) to 92.8% (WalkAway) of all tested isolates (Table 1). In the WalkAway system, three isolates (2.7%) were categorized as cefoxitin and oxacillin susceptible, whereas in the Vitek 2 and the Phoenix system, 9 isolates (8.1%) and 39 isolates (35.1%), respectively, were categorized as susceptible to both. One isolate was categorized as cefoxitin susceptible and oxacillin resistant by the Phoenix system.
Susceptibility pattern testing of cefoxitin and oxacillin for mecC-positive S. aureus isolates
The MIC90 values for oxacillin were >2 μg/ml (Phoenix), 2 μg/ml (MicroScan), and 2 μg/ml (Vitek 2). The MIC90 values for cefoxitin were >8 μg/ml (Phoenix) and >4 μg/ml (WalkAway); Vitek 2 detected 91.9% of isolates as resistant to cefoxitin without reporting a MIC value. Less than 10% of isolates were tested resistant to both cefoxitin and oxacillin (Phoenix, 9.9%; MicroScan, 4.5%; Vitek 2, 8.1%).
Applicability of chromogenic MRSA screening plates for detection of mecC-positive isolates.The vast majority of isolates showed typical growth on all tested cefoxitin-containing chromogenic MRSA screening plates. Reduced growth (i.e., smaller colonies, but with characteristic MRSA-indicating color) was observed for a small fraction of isolates (Table 2). Oxoid Brilliance MRSA 2 plates showed a mixed phenotypic appearance with blue (presumptive for MRSA) and white colonies for all isolates.
Growth on selective chromogenic agar media
Additionally, a subset of nine isolates and positive control strain S. aureus USA300, tested in triplicate, showed growth on screening plates from four manufacturers using an inoculum of 100 μl from a 10−5 dilution of a 0.5 McFarland standard suspension (approximately 100 CFU/plate). MRSA Select agar plates (Bio-Rad) were not tested in this additional experiment due to supply unavailability. Negative control S. aureus ATCC 29213 exhibited no growth on chromogenic agar plates.
Applicability of cefoxitin disk diffusion and oxacillin broth microdilution test for detection of mecC-positive isolates.The cefoxitin disk diffusion test detected mecC-encoded methicillin resistance in 111/111 isolates (i.e., 100%). The oxacillin broth microdilution resulted in a categorization of 43 susceptible (38.7%) and 68 resistant (61.3%) isolates.
DISCUSSION
The occurrence of mecC-harboring MRSA has been described in humans, companion animals, and livestock in several European countries (13). While the overall prevalence of these isolates seems to be low, it has been suspected that mecC prevalence might be underestimated because of its misidentification as methicillin-susceptible S. aureus (MSSA) due to its borderline resistant phenotype. Additionally, negative results in MRSA PCR and agglutination assays if only the mecA gene (i.e., the gene encoding PBP 2a) is targeted, hamper mecC-harboring MRSA detection efforts. Furthermore, it has been shown that the prevalence of mecC-positive S. aureus isolates increased at least in Denmark and that mecC-positive MRSA isolates are also capable of causing infections in humans (4). A reliable detection of these isolates is important to ensure both an adequate treatment of mecC-harboring MRSA infections and the use of the same prevention measures as already established for mecA-harboring MRSA. This study revealed that all chromogenic media and the cefoxitin disk diffusion test were able to categorize all mecC-positive MRSA strains properly. Additionally, we were able to show for a subset of strains that inocula as low as approximately 100 CFU per plate result in growth on chromogenic media, indicating that a recovery from clinical swab samples with low MRSA loads can likely be achieved. However, these findings are limited because they could mimic the usual clinical specimen as encountered in the laboratory only partially. To various degrees, all three AST systems displayed limitations in the ability to detect mecC MRSA. While the detection rate of WalkAway (97.3%) was also high, Vitek 2 (91.9%) and particularly the Phoenix system (64.9%) showed considerably lower rates. A study by Cartwright et al. showed a detection rate of 88.7% (n = 62 mecC-positive MRSA isolates) for the cefoxitin-resistant/oxacillin-susceptible pattern using the Vitek 2 (14); similarly, this AST device detected this pattern in 83.8% of the tested isolates in our study. The oxacillin broth microdilution performed poorly, showing a detection rate of only 61.3%. This is in accordance with previous studies (15).
In conclusion, automated systems may fail to detect mecC-encoded methicillin resistance, while all chromogenic screening media displayed colonies presumptive for MRSA growth. In comparison to oxacillin, cefoxitin was confirmed as superior surrogate marker to detect mecC-harboring MRSA isolates. Discrepancies between positive screening results based on the use of chromogenic media and categorization as methicillin susceptible by AST systems should be verified by molecular assays or disk diffusion.
ACKNOWLEDGMENTS
We are grateful to B. Grünastel and F. Erdmann for expert technical assistance.
This work was supported in part by grants BMBF 03ZZ0802H and 03ZZ0805B from the German Ministry for Education and Research to K.B. and was performed under the auspices of the Paul Ehrlich Gesellschaft für Chemotherapie, Section “Basics” and its working groups “Susceptibility testing and resistance” (G.W.) and “Staphylococcal infections” (K.B.).
The Phoenix (BD) and MicroScan WalkAway (Siemens) instruments were provided free of charge during the study. The authors declare no conflicts of interest.
FOOTNOTES
- Received 23 May 2017.
- Returned for modification 7 July 2017.
- Accepted 18 September 2017.
- Accepted manuscript posted online 4 October 2017.
For a commentary on this article, see https://doi.org/10.1128/JCM.01549-17.
- Copyright © 2017 American Society for Microbiology.
