Cloning and Occurrence of czrC, a Gene Conferring Cadmium and Zinc Resistance in Methicillin-Resistant Staphylococcus aureus CC398 Isolates

Abstract

We recently reported a phenotypic association between reduced susceptibility to zinc and methicillin resistance in Staphylococcus aureus CC398 isolates from Danish swine (F. M. Aarestrup, L. M. Cavaco, and H. Hasman, Vet. Microbiol. 142:455-457, 2009). The aim of this study was to identify the genetic determinant causing zinc resistance in CC398 and examine its prevalence in isolates of animal and human origin. Based on the sequence of the staphylococcal cassette chromosome mec (SCCmec) element from methicillin-resistant S. aureus (MRSA) CC398 strain SO385, a putative metal resistance gene was identified in strain 171 and cloned in S. aureus RN4220. Furthermore, 81 MRSA and 48 methicillin-susceptible S. aureus (MSSA) strains, isolated from pigs (31 and 28) and from humans (50 and 20) in Denmark, were tested for susceptibility to zinc chloride and for the presence of a putative resistance determinant, czrC, by PCR. The cloning of czrC confirmed that the zinc chloride and cadmium acetate MICs for isogenic constructs carrying this gene were increased compared to those for S. aureus RN4220. No difference in susceptibility to sodium arsenate, copper sulfate, or silver nitrate was observed. Seventy-four percent (n = 23) of the animal isolates and 48% (n = 24) of the human MRSA isolates of CC398 were resistant to zinc chloride and positive for czrC. All 48 MSSA strains from both human and pig origins were found to be susceptible to zinc chloride and negative for czrC. Our findings showed that czrC is encoding zinc and cadmium resistance in CC398 MRSA isolates, and that it is widespread both in humans and animals. Thus, resistance to heavy metals such as zinc and cadmium may play a role in the coselection of methicillin resistance in S. aureus.

Methicillin-resistant Staphylococcus aureus (MRSA) isolates mainly belonging to CC (clonal complex) 398 have emerged recently in livestock production around the world (7, 10, 20, 22). This clonal lineage is able to persist in the farm environment and therefore constitutes a large reservoir for transmission to humans working in such environments, which raises occupational health concerns (14, 22, 27). Resistance to antimicrobials, as well as other factors, may have contributed to the success of MRSA CC398 and its emergence in the animal reservoirs (1).

Metal-containing compounds are widely used as a feed supplement or for the prevention of gastrointestinal diseases in food animals. Resistance to metals is widely disseminated among bacterial species, and we have shown previously that the use of copper in feed can coselect for resistance to macrolides and glycopeptides in Enterococcus faecium (5). Heavy-metal resistance, i.e., resistance to arsenic, mercury, and cadmium, has been described among S. aureus isolates, including MRSA (2, 17-19). We recently described a phenotypic association between reduced susceptibility to zinc and MRSA CC398 from pigs in Denmark (1). It was suspected that the genetic background for the metal resistance was genetically linked to the methicillin resistance, and it was hypothesized to be present in the staphylococcal cassette chromosome mec (SCCmec) element carried by these MRSA isolates, as metal resistance genes had been described previously in the SCCmec element type III of MRSA isolate ST239 (9).

The purpose of this study was to identify the genetic determinant(s) involved in the observed zinc resistance phenotype among the CC398 MRSA isolates from animals in Denmark and to investigate its effects on susceptibility to several metallic compounds. Furthermore, the prevalence of resistance to zinc and the putative resistance gene were determined among a larger collection of swine and human S. aureus CC398 isolates from Denmark.

MATERIALS AND METHODS

Strains.

A total of 31 MRSA and 28 methicillin-susceptible S. aureus (MSSA) CC398 strains isolated from 2000 to 2008 from pigs in Denmark (1), as well as 50 MRSA and 20 MSSA strains of CC398 isolated from humans in Denmark, were included in the screening for the putative zinc resistance gene.

The 20 human MSSA isolates were isolated from bacteremia cases. The 50 human MRSA isolates were randomly chosen among 115 CC398 MRSA isolates referred to the Statens Serum Institut. The reporting of MRSA outbreaks has been required in Denmark since November 2006.

The strain collection from animals that had been tested previously for zinc susceptibility contained 74% of the zinc chloride-resistant isolates among the MRSA and none among the MSSA strains, whereas the human isolate collection had not been tested before.

All strains were genotyped previously using spa typing, and the SCCmec elements of the MRSA isolates were typed using a multiplex PCR technique (11, 21).

Cloning of the czrC gene.

S. aureus 171 (2007-70-91-4), a zinc-resistant isolate from a pig farm in Denmark (2), was used for the amplification of the czrC gene fragment. Escherichia coli TG1 (Stratagene, Cambridge, United Kingdom) was used for the transformation with the cloned pAT18 (24), and czrC-negative strain S. aureus RN4220 (12), a 8325 restriction-deficient strain able to receive plasmid DNA from E. coli (kind gift from Hanne Ingmer, University of Copenhagen), was used as a final recipient of the DNA insert-containing vectors. Erythromycin was used in the transformation experiments to select for the vector pAT18.

In two sequenced CC398 isolates, similar SCCmec type V (5C2&5) cassettes were identified (GenBank accession no. AM990992 and S. Li, R. L. Skov, A. R. Lasen, M. Sørum, M. Wulf, A. Voss, X. Han, K. Hiramatsu, and T. Ito, unpublished data). Both contained a putative metal transporter gene empirically annotated as cadA or copA based on homology to copper homeostasis genes present in the chromosome of S. aureus. The putative metal transporter was located between nucleotides 64589 and 66514 within the SCCmec element of SO385 and downstream of the second ccrC gene. Furthermore, this gene also was observed in another recently sequenced SCCmec type VIII cassette of MRSA isolated from a patient in Canada (15, 28), as well as in the composite SCC cassette in S. epidermidis ATCC 12228 (15, 28).

Based on the results of this study, we propose to change the designation of this gene to czrC (for cadmium and zinc resistance gene C) to correlate the name with its function and the C variant annotation, as the denominations czrAB already exist to refer to the chromosome-determined zinc transporter operon found in S. aureus; however, these are not associated with methicillin resistance (13). Therefore, in the remainder of this work we will refer to this gene as czrC.

For the cloning experiments, DNA of the czrC-positive MRSA strain 171 (2007-70-91-4) was used. The inserted fragment was amplified using the primers czrC start SmaI (5′-CTCGGGCCCGATCATCCATACTG-3′) and czrC end SmaI (5′-GTCCCCGGGTCAATCGATTCGTTCATTATTTAG-3′) (restriction sites are underlined). The amplified 2.8-kb fragment was digested with EcoRI (using a restriction site in the upstream region of the czrC gene) and SmaI (New England BioLabs, Wilbury Way Hitchin, United Kingdom) and treated with shrimp alkaline phosphatase (SAP) (USB Corporation, Cleveland, OH) before ligation into pAT18. The ligation product was electroporated into TG1-competent E. coli cells (Stratagene, Cambridge, United Kingdom). Transformants were selected on brain heart infusion (BHI) agar plates containing 150 μg/ml erythromycin. The presence of the cloned fragment containing the czrC gene was confirmed by PCR amplification using primers czrC forward (5′-TAGCCACGATCATAGTCATG-3′) and reverse (5′-ATCCTTGTTTTCCTTAGTGACTT-3′). Plasmid DNA was extracted from the E. coli transformants and transferred into competent S. aureus strain RN4220 by electroporation. The transformants were selected on selective plates containing 8 μg/ml erythromycin, and the presence of the czrC gene was confirmed by PCR and the sequencing of czrC. The isogenic strain containing pAT18 without any cloned fragment was obtained by the transformation of S. aureus strain RN4220 with the empty pAT18 vector DNA.

For the confirmation of the presence of the cloned vectors, plasmid extraction was performed for both E. coli and S. aureus transformants, and the identity of the vectors was confirmed by the comparison of the electrophoresis patterns, both unrestricted and restricted with EcoRI and SmaI (data not shown).

Susceptibility testing.

MICs were determined in an agar dilution assay for zinc chloride (0.25 to 16 mM) on Mueller-Hinton agar plates (Becton Dickinson), with the pH of the medium adjusted to 5.5 as previously described (2). The strains obtained in the cloning experiments were also tested for susceptibility to copper sulfate (0.125 to 16 mM), sodium arsenate (0.125 to 16 mM), cadmium acetate (0.125 to 16 mM), and silver nitrate (0.008 to 16 mM) using agar dilution at pH 7.4.

S. aureus S0385 (zinc and cadmium resistant), S. aureus C10682 (zinc and cadmium resistant), E. faecium A17sv1 HHAS210 (copper resistant), and S. aureus ATCC 29213 (susceptible quality control [QC] strain) were used on every test plate for the quality control of the susceptibility testing procedures.

The association between the zinc or cadmium susceptibility phenotype and the presence of czrC was investigated by the statistical analysis of the data in contingency tables using Fisher's exact test.

RESULTS

Cloning and metal susceptibility testing of the isogenic strains.

Isogenic strains of S. aureus RN4220, one containing the vector pAT18 alone and one with the cloned fragment containing the czrC gene, were compared. The presence of czrC conferred a 4-fold increase in the MIC of zinc chloride, from 2 mM in RN4220 with pAT18 to 8 mM in the isogenic strain containing pAT18 with czrC. Additionally, the resistance level of cadmium also was affected, and the strains carrying czrC showed an 8-fold increase in the MIC of cadmium acetate. On the contrary, the MICs of copper sulfate, silver nitrate, and sodium arsenate did not show any observable changes (Table 1).

TABLE 1.

Results of susceptibility testing against heavy-metal compounds of Staphylococcus aureus recipient strain and clones

Strain	czrC status	MICa (mM)
		ZnCl2	Cd(CH3COO)2	AgNo3	Na2HAsO4	CuSO4
RN4220	Negative	2	<0.125	0.125	0.5	8
RN4220 + pAT18	Negative	2	<0.125	0.125	0.5	8
RN4220 + pAT18 with czrC 1-1	Positive	8	1	0.125	0.5	8

^aAll MIC determinations were performed in agar dilution assays as described before (2), and a tentative breakpoint for resistance was defined as MIC > 2 mM.

Metal susceptibility testing.

A total of 74% (n = 23) of the MRSA from swine were resistant to zinc chloride (MIC > 2 mM), whereas all 28 MSSA strains were susceptible. All of the zinc-resistant MRSA strains harbored czrC, whereas all of the MSSA isolates and the eight MRSA isolates that displayed a zinc-susceptible phenotype tested negative for this gene. Among the human isolates, 48% (n = 24) of the MRSA strains tested were resistant to zinc chloride and harbored czrC, whereas 52% (n = 26) were susceptible. All 20 human MSSA isolates tested were susceptible to zinc chloride. All of the MSSA isolates and zinc-susceptible MRSA isolates were negative for the presence of czrC. These results show 100% correlation between phenotypic zinc chloride resistance and the presence of czrC. Additionally, the SCCmec typing data revealed that all of the MRSA strains positive for the czrC gene harbored a SCCmec cassette classified as type V according to the Kondo multiplex PCR method (11), whereas the negative strains harbored either a type IVa or V cassette in both strain collections.

DISCUSSION

MRSA CC398 isolates have emerged as a livestock-associated zoonotic pathogen in many countries, mostly in relation to occupational contact with animals (10, 14, 22, 25). Factors related to this emergence need to be understood before efficient counteractive measures and interventions can be designed.

MSSA CC398 isolates belong to one of three major clonal complexes that have been found to be associated with pigs (6). Among the MRSA population, mostly strains of CC398 have been isolated from pigs, although other CCs have been found in China (CC9) and some countries in Europe (CC1, CC8, CC9, CC30, and CC97) (3, 4, 16, 23, 26).

In our previous study, it was observed that most MRSA CC398 strains isolated from pigs in Denmark had a high prevalence of zinc resistance (74%), whereas the corresponding MSSA isolates were susceptible, and it was hypothesized that zinc resistance was a factor involved in the selection of MRSA of this clonal complex (1).

In the present study, a gene was cloned and expressed in S. aureus RN4220 and found to encode reduced susceptibility to zinc and cadmium but not to arsenic, copper, or silver. We therefore suggest renaming this gene czrC (for cadmium and zinc resistance gene C). The screenings performed in animal and human isolates confirmed that czrC is strongly associated with the resistance phenotype observed.

The MRSA CC398 strains from animals had a higher prevalence of zinc resistance than the strains from humans. This difference might be explained by differences between the SCCmec cassettes. It was observed that czrC was associated with SCCmec type V (5C2&5) in the two sequenced SCCmec cassettes (19a and Li et al., unpublished) and also with SCCmec type VIII (28). Although classified as type V SCCmec elements based on the identification of the ccrC and the class C mec element using the method described by Kondo and colleagues, there are variations within the cassette types that might not be detectable using the multiplex PCR strategy for typing (8). This also may explain the observation of both zinc-resistant isolates from pigs (n = 23) and humans (n = 24) but also zinc-susceptible MRSA isolates from pigs (n = 7) and humans (n = 15) carrying SCCmec elements classified as type V. Further information about the SCCmec elements circulating can be retrieved only by full sequencing, which may reveal more diversity of SCCmec elements (8).

The finding of a gene causing resistance to zinc that is linked to the SCCmec element is of concern, as compounds such as copper sulfate and zinc oxide are widely used in food-producing animals for the prevention of diarrheal diseases. Thus, the potential selective effect of heavy metals might affect the flora by providing a competitive advantage to resistant strains. The use of metallic compounds might be an important factor to consider in the future when investigating the selection dynamics of MRSA.

In conclusion, we cloned a resistance determinant, czrC, that is involved in zinc resistance in S. aureus. czrC was present in 74% of MRSA CC398 isolates from swine and 48% of MRSA CC398 isolates from humans in Denmark and was strongly related to zinc resistance. Further studies are needed to determine the occurrence of this gene among other populations and the importance for the coselection of MRSA.