INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

The trace heavy metal ions such as cobalt, zinc, copper, and nickel play important roles in bacteria. They regulate a wide array of metabolic functions as coenzymes or cofactors, as catalysts or Lewis acid in enzymes, and as structural stabilizers of enzymes and DNA-binding proteins (9, 18). However, these trace heavy metal ions are toxic in excess of normal physiological levels (28). Increasing environmental concentrations of these heavy metals pose a challenge to bacteria. Therefore, bacteria have evolved mechanisms to regulate the influx and efflux processes to maintain the relatively steady intracellular level of the heavy metal ions. Different molecular mechanisms have been reported that are responsible for resistance to various trace heavy metal ions in bacteria (2, 8, 13, 18, 22, 23, 27). The molecular mechanisms involve a number of proteins, such as ion transporters, reductases, glutathione-related cadystins and cysteine-rich metallothioneins, and low-molecular-weight cysteine-rich metal ligands (27). These protein molecules either export the metal ions out of cells or detoxify or sequester them so that the cells can grow in an environment containing high levels of toxic metals. However, there is no common mechanism of resistance to all heavy metal ions. In bacteria, the genes encoding resistance to heavy metals are located either on the bacterial chromosome, on the plasmids, or on both (18, 27).

Staphylococcus aureus is a common human pathogen associated with a number of diseases. Understanding of metal resistance in staphylococci has progressed rapidly in the past 10 years, with well-established cadmium, mercury, antimony, and arsenic resistance systems encoded by plasmids (20, 25, 27). However, staphylococcal strains without plasmids show resistance to heavy metal ions, such as zinc and cobalt. This implies that a plasmid-independent chromosomal determinant might encode resistance to heavy metals such as zinc and cobalt. Although operons encoding cobalt, zinc, and cadmium in Alcaligenes eutrophus (17) and zinc in Escherichia coli (2) have been investigated, relatively little is known about the transport of and resistance mechanisms to zinc and cobalt ions in S. aureus. Here we report the cloning, sequencing, and genetic analysis of a determinant located on the bacterial chromosome that codes for zinc and cobalt resistance in S. aureus.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Bacterial strains, plasmids, and growth conditions. The bacterial strains and plasmids used in this study are listed in Table 1. S. aureus strains were grown on tryptic soy agar or broth (TSA or TSB), whereas E. coli strains were grown on Luria-Bertani (LB) agar or broth at 37°C with shaking (200 rpm). When necessary for selection, ampicillin (50 µg/ml), kanamycin (30 µg/ml for E. coli, 500 µg/ml for S. aureus), and chloramphenicol (20 µg/ml for E. coli, 10 µg/ml for S. aureus) were added to the media. When required, an appropriate volume of filter-sterilized 0.5 M stock solutions of ZnCl₂ or CoCl₂ was added to TSB. The biomass in liquid cultures was estimated from the optical density at 580 nm (OD₅₈₀) with a DU-64 Beckman spectrophotometer. Cell yield was determined from a calibration curve relating OD₅₈₀ to cell dry weight.

DNA manipulation and sequencing. All standard methods of DNA manipulation were performed according to the protocols of Novick (19) and Sambrook et al. (26). Genomic DNA of S. aureus was isolated by using DNAzol kits (Molecular Research Center, Inc., Cincinnati, Ohio). Plasmid was purified with the QIAgen plasmid minipreparation kit (Qiagen, Inc., Chatsworth, Calif.). PCR-amplified products and DNA fragments from agarose gels were purified with QIAquick gel extraction kits. DNA probes were labeled by using the Rediprime DNA labeling system (Amersham Life Science, Arlington Heights, Ill.). All DNA restriction and modification enzymes were obtained from Promega (Madison, Wis.) and used according to the manufacturer's instructions. DNA sequences were determined with an ABI Prism 310 genetic analyzer system (Perkin-Elmer, Foster City, Calif.). Two pairs of oligonucleotide primers were used for PCR amplification: PCA1 and PCA2 (5'-TAAAGGCGGCGACACTTCACAC-3' and 5'-CTGGTGGTTTTTGCCCAAATTG-3') and CAF1 and CAB1 (5'-TTAGATGACATCCACGTAGCAACT-3' and 5'-GACCAAACAAGTCGCCATAAAGAC-3'). DNA sequences were analyzed by the MacVector (version 5.0) program, and multiple protein alignments were performed by the ClustalW program (29).

Construction of the zntA mutant and complementation. The 2.9-kb EcoRI fragment containing zntR and zntA was cloned into vector pTZ18R. The resulting plasmid pTZ18R-ZC (5.8 kb) was digested with SalI and SmaI to remove the AccI site from the multicloning site of the vector and then religated after end filling with Klenow DNA polymerase. The resulting plasmid was digested with AccI to insert a 1.5-kb kanamycin cassette from pOSTkan (11). The EcoRI fragment containing the kanamycin cassette within zntA was then subcloned into the pBT2 shuttle vector that contained a temperature-sensitive staphylococcal origin of replication (4). The resulting plasmid pBT2-ZCK was electroporated into competent S. aureus RN4220 cells. Selection for double-crossover events with the chromosome of S. aureus was carried out at 43°C as described previously (3, 4). One representative mutant was analyzed by Southern blotting in order to exclude possible rearrangement adjacent to the insertion sites or a single crossover event by using the 2.9-kb EcoRI fragment as a probe. The exact insertion site of the kanamycin cassette on the chromosome was confirmed by nucleotide sequencing of the PCR-amplified product with the primers PCA1 and PCA2. To complement the mutation in trans, the 2.9-kb EcoRI fragment containing zntR and zntA was cloned into the pCU1 shuttle vector (1). The resulting plasmid, pCU1-ZC, was electroporated into the mutant strain RN-MZ.

Analysis of zinc ion accumulation. Zinc concentration was measured as described by Beard et al. (2). Cultures grown overnight were transferred to 40 ml of fresh TSB to give an OD₅₈₀ of approximately 0.1. When the optical density of cultures came close to 1.0, appropriate amounts of ZnCl₂ were added to the cultures to give various final concentrations. The cultures were then incubated for an additional 4 h. Then, 25-ml aliquots of the cultures described above were centrifuged at 3,000 × g for 15 min. Cell pellets were washed with 4 ml of TSB and then with 4 ml of 0.1 N HNO₃ to remove the surface-bound zinc ions. The intracellular concentrations of zinc were determined with an atomic absorption spectrophotometer (Thermo Jarrell Ash Corp., Franklin, Mass.) (2).

Induction of znt transcription. The wild-type S. aureus strains were grown under the following conditions to evaluate the induction of znt operon in the presence of zinc ions: (i) in TSB without zinc, (ii) in TSB without zinc to mid-log phase and then subcultured (1:100) into TSB containing 1.5 mM zinc; and (iii) in TSB containing 0.5 and 2 mM zinc to mid-log phase and then subcultured (1:100) in TSB containing 1.5 mM zinc. The cells were harvested after 6 h, and the total RNA was isolated with the TRI reagent kit (Molecular Research Center). Then, 10 µg of total RNA from each sample was electrophoresed on formaldehyde agarose gels (1.0%) and transferred to nitrocellulose membrane. Membranes were prehybridized for 8 to 12 h and then hybridized overnight with [-³²P]dCTP-labeled znt or kanamycin gene under high-stringency conditions. The membrane was washed and subsequently autoradiographed.

Primer extension. The primer extension assay was performed by using an oligonucleotide primer of 19 bases (5'-GTAATCGCCTAATGCCTTG-3') specific to the zntR coding region. The primer (10 ng) and total RNA (10 µg) from wild-type S. aureus were combined in 3 µl of water and boiled for 1 min followed by quick cooling in ice water. The primer extension reaction was carried out in a total volume of 10 µl containing 2 µl of 5× AMV buffer; 2 µl of a 250-µM dATP, dTTP, and dGTP mix; 1 µl of [-³²P]dCTP; 12 U of avian myeloblastosis virus (AMV) reverse transcriptase; and 2 U of RNasin. The mixture was incubated at 37°C for 30 min. Chase solution (3 µl) containing 25-mM concentrations of dGTP, dCTP, dTTP, and dATP was then added, and the mixture was incubated for an additional 30 min at 37°C. The reaction was stopped by the addition of an equal volume of 98% formamide, 0.3% xylene cyanol, and 0.3% bromophenol blue. Denatured samples (5 µl) were electrophoresed on denaturing polyacrylamide gels. A sequence ladder generated by using the same primer on a 2.9-kb EcoRI fragment was coelectrophoresed and used to determine the position of the transcription start site.

Nucleotide sequence accession number. The nucleotide sequence reported here has been submitted to GenBank under accession number AF044951.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

Cloning and sequencing of S. aureus genes encoding zinc and cobalt resistance. The genomic library of S. aureus RN450 was constructed in the pCU1 shuttle vector (1). S. aureus RN450 was transformed with this pCU1-based S. aureus genomic library by electroporation (19). Approximately 2,000 S. aureus transformants were tested for zinc tolerance by replica plating on TSA medium containing 10 mM ZnCl₂. After incubating for 2 days at 37°C, the transformants that showed faster growth were further analyzed. The plasmids isolated from these clones were transformed into S. aureus. DNA from positive transformants yielded resistance to 10 mM ZnCl₂, suggesting that zinc resistance was due to a sequence contained on the plasmids. Further restriction analysis and zinc resistance studies indicated that the gene(s) encoding the zinc resistance were located within a 2.9-kb EcoRI fragment that was subsequently subcloned into the pTZ18R vector (16).

Analysis of predicted amino acid sequences of ZntR and ZntA. The predicted ZntR sequence showed 32% identity of amino acid residues with SmtB of Synechococcus (8), 32% identity with CadC of S. aureus (32), 30% identity with ArsR of S. xylosus (25) and S. aureus (10), and 28% identity with ArsR of E. coli (6). All of these homologs are members of the ArsR family (6, 10). ZntR appeared to be a hydrophilic protein, suggesting a cytoplasmic location. The secondary structure and domain analysis of ZntR suggested that there was a consensus DNA-binding helix-turn-helix motif (7), but it lacks the conserved CXC sequence. A putative regulatory region with an inverted repeat was located upstream of zntR, which represented the potential binding site of a regulatory protein (8).

Construction of a zntA null mutant and complementation analysis. The mutation of zntA was achieved to test our hypothesis whether this operon is responsible for zinc and cobalt resistance. The mutant strain RN-MZ was created by the insertion of the kanamycin cassette at the AccI site of the zntA gene (Fig. 1). Insertion of the ~1.5-kb kanamycin cassette increased the size of the hybridizing EcoRI fragment from ~4.3 kb in the parent to ~5.8 kb in the mutant RN-MZ (data not shown). The PCR analysis with primers PCA1 and PCA2 spanning the insertion site of the kanamycin cassette further proved that the size of the PCR product from the mutant strain was increased by ~1.5 kb compared to that of the wild type. Nucleotide sequence analyses of the PCR products from the mutant and the parent strains also confirmed the insertion of the kanamycin cassette at the AccI site of the zntA gene of the chromosome. As shown in Table 2, the MICs of zinc and cobalt for the mutant RN-MZ (zntA) were each 0.5 mM compared to 5 and 3 mM for the parent strain, respectively. Thus, the mutant RN-MZ (zntA) strain was sensitive to zinc and cobalt ions.

View larger version (19K): [in this window] [in a new window]   FIG. 1.   Construction of the zntA mutant. First, a 2.9-kb EcoRI fragment containing zntR and zntA was cloned into pTZ18R (SmaI-SalI) to construct the pTZ18R-ZC plasmid (5.8 kb). A 1.5-kb fragment containing the kanamycin resistance cassette from pOSTkan was introduced into the AccI site within zntA by blunt-end ligation. From the resulting plasmid, pTZ18R-ZCK (7.3 kb), the EcoRI fragment was subcloned into the pBT2 shuttle vector to generate the pBT2-ZCK plasmid (11 kb). ERI, EcoRI; A, AccI; X, XbaI; P, PstI; Amp, ampicillin; Kan, kanamycin; Cat, chloramphenicol; ori E, E. coli origin of replication; ori S(ts), temperature-sensitive S. aureus origin of replication.

Functional analysis of ZntA. The intracellular concentration of zinc was measured to determine whether ZntA is involved in the influx or efflux of zinc ions. S. aureus strains grown in TSB medium containing various concentrations of zinc were collected, washed, and digested (2) and were used to determine zinc concentration by atomic absorption spectrophotometry. As shown in Fig. 2, the mutant strain accumulated zinc ion to twice the level of the parent strain. In contrast, in the RN-ZC strain (containing multiple copies of znt) the intracellular zinc concentration was only one-half that of the parent strain. The accumulation of high concentrations of zinc in the zntA mutant is probably indicative of its inability to efflux zinc. The lower zinc concentration in RN-ZC further supports the involvement of ZntA in its transport. Again, the RN-ZC strain showed more-efficient efflux of zinc ions from cells compared to the parent strain due to an increased intracellular level of ZntA.

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Zinc accumulation by RN4220, RN-ZC and the zntA mutant (RN-MZ) strains of S. aureus. The bacteria grown in TSB medium containing various concentrations of zinc were collected and washed, and the levels of intracellular zinc were determined by atomic absorption spectrometry. Symbols: , RN4220; , RN-MZ; , RN-ZC.

Analysis of znt transcript. A preliminary study was conducted to evaluate the effect of various concentrations of zinc on the growth of wild-type and mutant S. aureus strains. Northern blot analysis was performed to evaluate the expression of znt during growth of S. aureus strains in the presence or in the absence of zinc. Total RNA was isolated from cultures exposed to different concentrations of zinc and then separated by electrophoresis in 1% agarose. Northern blot analysis revealed the presence of only one transcription unit of ~1.4 kb when fragments containing a portion of zntR and zntA were used as probes (Fig. 3A). This size seemed to be in good agreement with that of the predicted 1.3-kb operon. It also implied that zntR and zntA were cotranscribed. Interestingly, as shown in Fig. 3A (lane 4), the zntA mutant also expressed a ~1.4-kb transcript size, which was almost equal to the transcript size of zntA in the wild-type strain (lane 5). We assumed that although the insertion inactivated zntA, it read through the kanamycin resistance gene and produced a ~1.4-kb transcript. When the kanamycin resistance cassette was used as a probe, Northern blotting revealed two hybridizing bands (~1.4 and ~1.2 kb) in the mutant total RNA (Fig. 3B). The ~1.2-kb unit is the transcriptional product of the kanamycin resistance cassette.

View larger version (67K): [in this window] [in a new window]   FIG. 3.   Northern blot analysis of the zntR and zntA. Total RNA (10 µg) isolated from samples was separated electrophoretically and transferred by blotting onto a membrane. (A) Lanes: 1, S. aureus RN4220 grown with 2 mM zinc to mid-log phase and then diluted (1:100) in TSB medium containing 1.5 mM zinc; 2, S. aureus RN4220 grown with 0.5 mM zinc to mid-log phase and then diluted (1:100) in TSB medium containing 1.5 mM zinc; 3, S. aureus RN4220 grown without zinc to mid-log phase and then diluted (1:100) in TSB medium containing 1.5 mM zinc: 4, RN-MZ grown in TSB; 5, S. aureus RN4220 grown in TSB. The blot was probed with a radiolabeled DNA fragment encompassing zntR and zntA. The sizes of the rRNA are marked with arrows. (B) An RNA sample (10 µg) isolated from the mutant RN-MZ strain was separated electrophoretically and transferred by blotting onto a membrane. The blot was probed with a radiolabeled DNA fragment containing the kanamycin gene.

Mapping of znt transcription start site. To define the transcription unit more precisely, the 5' end of zntR transcript was mapped. A 19-base oligonucleotide specific to the coding region was annealed with total RNA and extended in a primer extension assay. The transcription start site was located in between the 10 region and an imperfect 8-2-8 hyphenated inverted repeat that represented a potential regulatory site (Fig. 4). The presence of a minor band in Fig. 4 (lane P) may be due to a minor transcription start site in the znt operon, to the degraded mRNA product, or to RNA secondary structures.

View larger version (54K): [in this window] [in a new window]   FIG. 4.   Mapping of the 5' end of zntR gene by primer extension analysis. Total RNA from the parent strain RN4220 was hybridized with an oligonucleotide complementary to the mRNA of zntR locus and extended by AMV reverse transcriptase (lane P). Lanes T, G, C, and A correspond to a dideoxy sequencing reaction performed with the same primer. The sequencing encompassing the initiation start (marked by an asterisk) is enlarged.

Localization of znt on the bacterial chromosome. Various known S. aureus gene clones available in our laboratory were used to explore the location of znt on the bacterial chromosome. These include lytI (15), lytM (24), atl (21), brnQ (30), and others. Dot blot studies showed hybridization of znt with the cosmid clones containing the lytI gene, thus mapping it within the SmaI-B fragment of S. aureus (14).