INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Copper is an essential cofactor for enzymes involved in diverse biological processes including respiration, destruction of free radicals, iron homeostasis, and neurological development. However, when in excess, copper is toxic as it generates reactive oxygen species via the Fenton reaction, disrupts metal ion binding and homeostasis, and binds macromolecules such as proteins inappropriately (25, 50).

Resistance to the toxicity of copper and other heavy metals is achieved through mechanisms such as reduced influx, facilitated efflux, sequestration, and modification. Eukaryotic organisms generally utilize small cysteine-rich proteins termed metallothioneins (MTs) to sequester metals (17). MTs bind heavy metals through thiolate bonds; the multiple cysteine residues are arranged in Cys-X-Cys or Cys-X-X-Cys repeats that coordinate metal binding. In contrast, prokaryotes generally achieve copper resistance by facilitated efflux utilizing transporters such as the P1-type ATPases (45).

The closely related yeasts Saccharomyces cerevisiae and Candida glabrata use MTs for copper homeostasis. The major copper resistance determinant in S. cerevisiae is CUP1, which encodes an MT (4, 10). CUP1 expression is induced to a high level when excess copper is present, and the gene can be amplified up to 20 times or more (5, 18, 52). These two factors combine to allow CUP1-amplified strains to grow in medium with high levels of copper (21). The S. cerevisiae CRS5 gene encodes a second MT, which is not amplified or expressed at a high level and which provides minimal resistance to copper ion toxicity (8). S. cerevisiae also utilizes the P1-type ATPases Ccc2p and Pca1p for copper homeostasis. Ccc2p is a homolog of the Menkes disease protein that is involved in exporting cytosolic copper to the secretory pathway and the extracytosolic domain of the copper-dependent oxidase Fet3p (54). In contrast, Pca1p is thought to play a role in resistance to copper ion toxicity as pca1 strains do not achieve wild-type growth levels in medium containing high concentrations of copper (39). C. glabrata contains three MT-encoding genes (28). The MTII_A/B genes are highly induced in the presence of copper and are most important in conferring resistance to copper (29, 30). While a few other fungal species have been shown to contain MTs, the fission yeast Schizosaccharomyces pombe utilizes an alternative mechanism to achieve resistance to copper ion toxicity. In S. pombe, (-Glu-Cys)_n-Gly peptides termed phytochelatins are used to sequester copper (7).

This report describes mechanisms of resistance to copper ion toxicity employed by the diploid yeast C. albicans, an opportunistic pathogen of humans (43). C. albicans is a common commensal that is found associated with the gastrointestinal tracts of various mammalian species. In immunocompromised individuals this generally harmless yeast can cause both superficial and life-threatening disseminated diseases (37). This study of copper homeostasis in C. albicans demonstrates that the major resistance locus encodes a copper-inducible P1-type ATPase. In addition to conferring high-level copper resistance, this ATPase contributes to silver resistance. It is also shown that C. albicans contains an MT gene that is not amplified and that provides minimal copper resistance. Since silver is widely used as an antimicrobial agent, possession of a pump that confers resistance to silver ion toxicity may enhance the spread of C. albicans as a nosocomial pathogen.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and plasmids. C. albicans strains are listed in Table 1. Parent strains were CAI4 (ura3/ura3) and SGY243 (ura3/ura3) (11, 23). To construct crd1 and crd2 null mutants, the parental strains were transformed with the disruption plasmids pPR300 and pPR400, respectively. The S. cerevisiae strain dRM110.18A (MATa leu2-3,112 ura3::HIS3 cup1::ura3::THR3) was kindly provided by R. Boumil and D. Dawson. For plasmid cloning, E. coli XL1-Blue (Stratagene) was used.

Growth media. C. albicans and S. cerevisiae strains used in this study were routinely cultured on YPD (1% yeast extract, 2% Bacto peptone, 2% glucose), CM lacking uracil or uridine (Urd), or SD (0.67% yeast nitrogen base, 2% glucose) at 30°C (2). For solid media 2% agar (Difco) was added. Urd auxotrophs were selected on medium containing 5-fluoroorotic acid (5-FOA) (2). Escherichia coli cells were cultured in L broth or on L plates (31). Ampicillin was added at a concentration of 100 µg/ml.

Metal sensitivity assays. C. albicans strains were grown to saturation at 30°C in liquid SD medium and diluted 100-fold into SD medium with various concentrations of CuSO₄, AgNO₃, or CdSO₄. Cell growth was monitored by determining the optical density at 600 nm (OD₆₀₀) of the cultures.

DNA manipulations and analysis. Plasmid isolation, PCR, restriction digestion, cloning, gel electrophoresis, and Southern hybridization analysis were performed by standard methods (2). C. albicans genomic DNA was isolated using a glass bead disruption procedure (2). Automated DNA sequencing was performed by Michael Berne and coworkers at the Tufts University Core Facility. DNA sequence analysis utilized the Lasergene sequence analysis software (DNAStar Inc., Madison, Wis.) or the Blast algorithm (National Center for Biotechnology Information).

RNA analysis. Cells for RNA isolation were grown in SD liquid medium at 30°C to an OD₆₀₀ of 1.0. Where appropriate CuSO₄ or CdSO₄ was added at 0.1 or 1.0 mM and the cultures were incubated for an additional 30 min. For RNA extraction, cells were lysed using glass beads and phenol (2). Ten micrograms of total RNA per sample was separated on a formaldehyde-agarose gel, transferred to a Nytran-plus membrane, and probed by standard Northern hybridization methods (2). Densitometric analysis was performed with a Molecular Dynamics computing densitometer and ImageQuant, version 3.3, software.

Transformation and screening of the C. albicans genomic library. S. cerevisiae cells were transformed by the lithium acetate method of Gietz et al. (13). A C. albicans genomic library (42) was used to transform S. cerevisiae strain dRM110.18A, and transformants were replicated to SD medium containing 125 µM CuSO₄. Approximately 50,000 transformants were screened, and 15 colonies that exhibited growth on 125 µM CuSO₄ were chosen for further study. Plasmids were isolated from the resistant transformants and reintroduced into the cup1 strain dRM110.18A to confirm the heterologous complementation.

Plasmid construction. Disruption constructs for CRD1 and CRD2 utilized the "URA blaster" plasmid pMB7 (11). In the primer sequences given below, underlined sequences are complementary to sequences from the CRD1 and CRD2 loci while sequences that are not underlined indicate restriction enzyme sites added to the primer. The CRD1 deletion construct primers NTRMDR (5'-GCCGAGCTCGAAGACGGTAATAGTTCAGTTGTTGG-3') and NTRMRV (5'-GCCGAGCTCACACGATGTTGACATTGG-3') yielded a 443-bp PCR fragment with SacI sites at the ends, and the primers CTRMDR (5'-GCCCTGCAGGTGAAAGCTCACTCTCG-3') and CTRMPL (5'-GCCCTGCAGTCGGCGGATGTTGGAATTG-3') yielded a 390-bp PCR fragment with PstI sites at the ends. The fragments were digested with SacI or PstI and ligated sequentially into the SacI and PstI sites flanking the URA blaster cassette in pMB7 to give the CRD1 disruption plasmid pPR300. The CRD2 deletion construct primers MTNDSR (5'-CGGGAGCTCGAAGACCTCCGCCTTTACTTTCAACG-3') and MT5'RVP (5'-GCCGAGCTCGAGCACAGACACATTGAGC-3') yielded a 318-bp fragment with SacI sites at the ends, and the primers MT3'UTRP (5'-GCCCTGCAGCTATACTAACCAACAACG-3') and 3'DSRPT (GCCCTGCAGGAAGACTGAAAGATTATCTGAGTAC-3') yielded a 549-bp fragment with PstI sites at the ends. The fragments were digested with SacI or PstI and ligated sequentially into the SacI and PstI sites flanking the URA blaster cassette in pMB7 to give the CRD2 disruption plasmid pPR400. To construct the CRD1-complementing plasmid pPR311, a 1.1-kb HindIII/SphI fragment containing the smaller C-terminal portion of the CRD1 open reading frame (ORF), some of the 3'-untranslated region, and 187 bp of the tet gene sequence (including the SphI site) from YEp13 was cloned into the HindIII/SphI-digested plasmid pCK66. Plasmid pCK66 is derived from pNEB193 (New England BioLabs) and contains the SacI/XbaI (blunted) fragment of the C. albicans URA3 gene cloned into the NdeI site (blunted). A 4.3-kbp HindIII fragment encompassing most of the CRD1 ORF and some of the 5' upstream sequence was cloned in the proper orientation into the HindIII site of the above construct, reconstituting the CRD1 ORF. For directed integration into the CRD1 locus, pPR311 was digested with the restriction enzyme ClaI.

Gene disruption of the CRD genes. The CRD genes were disrupted using the URA blaster method (1, 11). Prior to transformation, both the CRD1 and CRD2 disruption cassettes (from the disruption plasmids pPR300 and pPR400, respectively) were released from the pMB7 plasmid backbone by digestion with the restriction enzyme BbsI. The primer pairs were engineered so that digestion with the restriction enzyme BbsI would generate ends derived from the C. albicans sequence. To generate the CRD1 and CRD2 disruption strains, sequential transformations were done as described previously (1, 11). Strains that had resolved out the URA3 gene were selected on 5-FOA plates (2). Southern hybridization analysis was used to confirm all integration events. For analysis of CRD1 disruptions, chromosomal DNA was digested with EcoRI and XhoI and probed with an 1,126-bp XbaI/XhoI fragment from the CRD1 ORF. Sizes of the hybridizing bands were 2.4 kb for the CRD1 allele and 3.8 and 8.1 kb for crd1::hisG-URA3-hisG and crd1::hisG alleles, respectively. EcoRI- and XhoI-digested DNA from CAPR301 yielded fragments of the expected sizes of 2.4 and 3.8 kb for the wild-type and disrupted alleles, respectively. However, the fragment corresponding to the wild-type CRD1 allele was approximately twice as intense as that corresponding to the allele disrupted with the URA blaster cassette, suggesting that C. albicans contains three copies of the CRD1 gene. Several Urd strains were isolated by plating these first-round disruptants on 5-FOA medium and selecting strains that had undergone recombination between the direct hisG repeats of the URA blaster cassette. EcoRI- and XhoI-digested DNA from CAPR302 yielded a fragment of approximately 8.1 kb due to the loss of the URA3 gene and its accompanying EcoRI site. Again, the hybridization signal of the wild-type allele was about twofold greater than that of the hisG-disrupted allele. Four independently constructed 5-FOA^r heterozygotes were retransformed using the CRD1 disruption construct with URA3 as the selectable marker. EcoRI- and XhoI-digested DNA from several independent strains exhibited a banding pattern where the CRD1, crd1::hisG-URA3-hisG, and crd1::hisG alleles were all visible, indicating that C. albicans contains three copies of the CRD1 gene. These strains were plated on 5-FOA medium and screened for the desired intrachromosomal recombination event. Two of the 5-FOA^r strains were chosen for a third round of transformation with the CRD1 disruption construct. The DNA from several third-round transformants gave the pattern of hybridizing fragments expected for homozygous null mutants. For analysis of CRD2 disruptions, chromosomal DNA was digested with NdeI and SacI and probed with a 420-bp PCR fragment corresponding to sequence that starts 759 bp downstream of the CRD2 ORF. DNA for this probe was amplified using the primer pair MTCTP1F (5'-GCTGTTGTTATTGTCAGG) and MTCTP2R (5'-ATGGATTGGGTTTGCTAG). Sizes of the hybridizing bands were approximately 8 kbp for the CRD2 allele and 7 and 6 kbp for the crd2::hisG-URA3-hisG and crd2::hisG alleles, respectively.

Nucleotide sequence accession number. The sequences of the CRD1 and CRD2 genes have been deposited in GenBank under accession no. AF268098 and AF268099, respectively.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of C. albicans genomic sequences conferring copper resistance. S. cerevisiae strains with the CUP1 locus deleted (cup1) are extremely sensitive to exogenous copper ions (18). To isolate C. albicans genes that mediate resistance to copper ion toxicity, a 2 µm library containing C. albicans genomic fragments was screened for sequences that would heterologously complement the copper sensitivity phenotype of an S. cerevisiae cup1 strain (42). Fifteen plasmids that conferred the ability to grow on medium containing 125 µM copper sulfate were characterized. Restriction mapping and Southern hybridization analysis indicated that the plasmids were of two classes (six isolates for class I and nine isolates for class II), containing inserts of approximately 7.3 kb for class I and 4 kb for class II. The DNA sequences of subclones of the class I and class II plasmids were determined.

Class I clones encode a P1-type ATPase. The nucleotide sequences of the class I clones were analyzed using the Blast algorithm, and a 3,591 bp ORF whose product exhibited significant sequence similarity to copper-transporting P1-type (CPx-type) ATPases was found (Fig. 1A) (26, 47). An essentially identical sequence was independently determined and deposited in GenBank by Weissman et al. (accession no. AAF04593) (51). As can be seen in Fig. 1B, an alignment of the most conserved region of P1-type ATPases reveals high similarity to ATP-driven copper pumps from humans, plants, and prokaryotes. Five copies of CXXC motifs, characteristic of proteins that bind copper, are found in the N-terminal region (6). The three most C-terminal CXXC sequences are part of the larger motif GMXCXXC, termed the HMA (heavy-metal-associated) sequence, which is seen in the N-terminal regions of most prokaryotic and eukaryotic P1-type ATPases. Additional sequences of approximately 60 amino acids surrounding the HMA motif exhibit extended sequence similarity to the copper binding modules of copper-transporting P1-type ATPases (3). The two most N-terminal CXXC sequences are not part of the classical HMA sequence and may not represent copper binding modules. Other features characteristic of P1-type ATPases included eight transmembrane domains (predicted from hydropathy analysis), the conserved CPC motif in the sixth transmembrane domain, thought to be involved in ion translocation, and a conserved histidine-proline sequence 41 residues C-terminal to the CPC motif (41).

View larger version (27K): [in this window] [in a new window]   FIG. 1.   (A) Two-dimensional model of the Crd1p P1-type ATPase depicting key features and putative membrane topology. Crd1p is predicted to have five N-terminal, cytosolic metal binding domains (CXXC) thought to be involved in Cu(I) binding. Hydropathy analysis is consistent with eight transmembrane (TM) domains. The following conserved motifs and their function are shown: TGES, phosphatase domain; CPC, ion translocation; DKTGT, aspartyl kinase domain; SHHPLG, unknown function (conserved HP dipeptide); GDGINDAP, ATP binding. The diagram is not to scale. (B) Protein sequence alignments of conserved domains involved in ion translocation from related proteins involved in copper transport. The most conserved region of all P1-type ATPases, encompassing a stretch of approximately 70 amino acids which contains the membrane helix with the CPx motif and the phosphorylation domain DKTGT, is shown. Regions of sequence identity to the consensus sequence are in black. Dashes, gaps in the amino acid sequence compared to other sequences. The sequences were aligned using the MegAlign program (Clustal method) of the DNAStar software package. GenBank accession numbers: C. albicans Crd1p, AF268098; S. cerevisiae Pca1p, Z29332; S. cerevisiae Ccc2p, AAC37425.1; human/Menkes AT7A, AAA35580.1; A. thaliana RAN1, AAD29109; Bacillus subtilis P1-type ATPase, CAB15335; E. hirae CopA, AAA61385.1.

Class II clones encode an MT. Analysis of the DNA sequence of the class II clones indicated that they contained a small ORF exhibiting sequence similarity to higher eukaryotic MTs, and this gene has been termed CRD2 (17). The C. albicans MT is 76 amino acids in length and contains 12 cysteine residues with 10 of the Cys residues part of Cys-X-Cys motifs (Fig. 2). The CRD2 gene product contains 16% cysteine, 12% serine, 9% lysine, and only one aromatic amino acid residue, phenylalanine. These features are characteristic of MTs, which are typically 50 to 70 amino acid residues in length containing multiple Cys-X-Cys or Cys-Cys sequences and having 20 to 30% Cys, 20 to 30% Ser and Lys, and few aromatic amino acids. The C. albicans MT was quite similar to those of fish, such as carp and flounder, and to those of plants, such as Arabidopsis thaliana and tomato, but shared little sequence similarity with the S. cerevisiae Cup1p (data not shown).

View larger version (14K): [in this window] [in a new window]   FIG. 2.   Deduced amino acid sequence of the C. albicans MT encoded by CRD2. Cys-Xaa-Cys repeats thought to coordinate copper binding are underlined.

Expression of the CRD1 and CRD2 genes. The MT genes of eukaryotes and the resistance pumps of prokaryotes that are involved in preventing copper ion toxicity exhibit copper-specific induction. To determine whether the CRD1 and CRD2 genes were regulated in response to changing copper levels, C. albicans strains SGY243 (ura3/ura3) and CAI4 (ura3/ura3) were grown in various concentrations of copper or cadmium and CRD1 and CRD2 mRNA levels were examined by Northern hybridization analysis. CRD1 mRNA was induced approximately 14- and 32-fold by exposure to 0.1 and 1 mM CuSO₄, respectively (Fig. 3A). However, exposure of cells to 0.1 and 1 mM CdSO₄ led only to two- to threefold induction. In contrast to the results seen with the CRD1 gene, CRD2 transcription was maintained at a basal level, as exposure of cells to 0.1 and 1 mM CuSO₄ led to no significant induction of CRD2 mRNA over the levels seen with cells grown without added copper (Fig. 3B). Exposure to CdSO₄ at 0.1 and 1 mM led to approximately 4- and 12-fold decreases, respectively, in steady-state mRNA levels of the CRD2 gene (Fig. 3B). Results similar to those shown in Fig. 3 were obtained from Northern hybridization analysis of CRD1 and CRD2 transcription in strain CAI4 (data not shown). These results indicate that the presence of copper leads to high-level induction of the CRD1 gene while transcription of the CRD2 locus is not induced, suggesting that the CRD1 gene may play the major role in resistance to copper ion toxicity.

View larger version (30K): [in this window] [in a new window]   FIG. 3.   Analysis of CRD1 and CRD2 transcription in response to metals. Strain SGY243 was grown to an OD600 of 1.0 in SD plus Urd and then treated for 30 min with the indicated concentration of CuSO4 or CdSO4 prior to RNA isolation for Northern hybridization analysis. Results were quantitated by densitometry. (A) Autoradiogram of a blot that was sequentially hybridized to CRD1- and ACT1-containing probes (the ACT1 mRNA was used as a loading control). (B) Autoradiogram of a blot that was sequentially hybridized to CRD2- and ACT1-containing probes.

Disruption of the CRD1 and CRD2 genes. To determine the importance of the CRD1 and CRD2 genes in resistance to copper ion toxicity in C. albicans, the genes were sequentially disrupted using the URA blaster technique (1, 11). Recombination with the disruption fragment from pPR300 resulted in replacement of 740 bp of the CRD1 ORF with the URA blaster cassette, as illustrated in Fig. 4A. The deleted portion of the CRD1 ORF encoded the DKTGT (aspartyl kinase) and the GDGINDAP (ATP binding) motifs, which are expected to be essential for CRD1 function. C. albicans strain CAI4, a homozygous ura3 strain congenic to the clinical isolate SC5314, was transformed as described in Materials and Methods. Several independent transformants were selected and analyzed. Southern hybridization of chromosomal DNA from representative strains is shown in Fig. 4B.

View larger version (11K): [in this window] [in a new window]   FIG. 4.   Disruption of the C. albicans CRD1 gene. (A) A portion of the CRD1 ORF which encodes the conserved DKTGT (aspartyl kinase domain), HP, and GDGXNDXP (ATP binding domain) amino acid sequences was replaced by the hisG-URA3-hisG cassette. The 1,126-bp XhoI/XbaI restriction fragment (positions 464 and 1590 of the CRD1 sequence, respectively) was used as a probe. Shaded rectangles, CRD1 ORF; open rectangle, C. albicans URA3 gene that is part of the URA blaster cassette; unboxed arrows, Salmonella hisG repeat sequences that are part of the URA blaster cassette; heavy line, fragment used as probe. Only restriction sites relevant for analysis of the disruption strains are given. The sites are as follows: EcoRI (E), SacI (S), PstI (P), XbaI (Xb), and XhoI (Xh). Numbers in parentheses, positions in base pairs of relevant restriction sites with respect to the first nucleotide of the CRD1 ORF, unless indicated otherwise. (B) Southern hybridization analysis of the CRD1 disruption strains used in this study. Lane A, CAI4 (CRD1/CRD1/CRD1); lane B, CAPR301 (CRD1/CRD1/crd1::hisG-URA3-hisG); lane C, CAPR302 (CRD1/CRD1/crd1::hisG); lane D, CAP303 (CRD1/crd1::hisG-URA3-hisG/crd1::hisG); lane E, CAPR304 (CRD1/crd1::hisG/crd1::hisG); lane F, CAPR305 (crd1::hisG-URA3-hisG/crd1::hisG/crd1::hisG).

ura3/ura3

Phenotypic analysis of the crd1 and crd2 disruption strains. The crd1 homozygous null mutant CAPR305 was extremely sensitive to copper ions, exhibiting a complete lack of growth in SD medium containing 100 µM CuSO₄ regardless of the time of incubation (Fig. 5B). In contrast, the strain that contains two disrupted alleles and one wild-type CRD1 allele exhibited growth in the presence of CuSO₄, but its growth was clearly slower than that of wild-type strain CAI4 (Fig. 5A). After 24 h of growth in medium containing 1,000 µM CuSO₄, the mutant strain CAPR303 (crd1/crd1/CRD1) grew to a density that was 26% of that observed in the absence of CuSO₄. After 72 h, the level of growth seen with CAPR303 was close to that of the wild-type strain up to the highest levels of CuSO₄ tested (Fig. 5B). Additionally, complementation of the homozygous null mutant with a single, integrated copy of the CRD1 gene (crd1/crd1/crd1/p3011[CRD1]) led to growth in copper-containing medium that was equal or greater than that seen with CAPR303 (Fig. 5A and B). These results indicate that C. albicans employs a copper-inducible P1-type ATPase transporter as the primary mechanism of resistance to high levels of copper.

View larger version (20K): [in this window] [in a new window]   FIG. 5.   Metal ion resistance of wild-type and crd disruption strains. Cells were grown overnight to saturation at 30°C in SD and diluted 1:100 into SD with various concentrations of CuSO4, AgNO3, or CdSO4. Cell growth was monitored by determining the OD600 of the cultures, with 100% growth equal to total cell growth in SD medium without added metal ions. (A and B) Strains CAI4 (), CAPR303 (×), CAPR305 (), and CAPR307 (). Growth at 24 (A) and 72 h (B) is shown. (C) Growth of strains CAI4 (CRD1) and CAPR305 (crd1) in 200 µM AgNO3 at 40 h and in 200 µM CdSO4 at 60 h. (D) Growth of wild-type and crd2 disruption strains at 18 h. Strains: CAI4 (), CAPR401 (×), and CAPR403 (). For genotypes of strains, see Table 1.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Cells need trace amounts of copper; but when copper is in excess, it is highly toxic (25, 50). To prevent copper toxicity, eukaryotes generally use MTs to chelate copper (17). In contrast, prokaryotes generally reduce influx and/or utilize efflux mechanisms to control intracellular copper levels (44). This work demonstrates that C. albicans primarily utilizes a P1-type ATPase (copper transporter) to resist copper ion toxicity because deletion of all copies of CRD1 resulted in a strain that exhibited greatly reduced resistance to copper ion toxicity, while deletion of all copies of CRD2 had a minor effect. This is the first example of a eukaryote utilizing a P1-type ATPase as its primary copper resistance mechanism.

Database analysis shows that the product of the CRD1 gene is highly similar to other P1-type ATPases involved in copper homeostasis in mammals, yeast, and prokaryotes. Examples of such proteins from mammals include the Menkes disease and Wilson disease proteins (3). Menkes disease is an X-linked disorder in which copper export is defective in many cell types, including the intestine, leading to systemic copper deficiency and ultimately neurological dysfunction and death. The autosomal recessive Wilson disease is due to mutations in a highly sequence-similar ATPase pump that is defective in hepatocytes, leading to liver and neurological damage due to copper ion toxicity. The Menkes and Wilson disease proteins each contain six CXXC motifs in their N-terminal regions, while Crd1p contains five CXXC repeats. In contrast, with the exception of E. coli CopA, which has two CXXC N-terminal repeats, most prokaryotic Cu-transporting ATPases have a single N-terminal HMA metal binding motif (40). The S. cerevisiae homolog of the Menkes disease and Wilson disease proteins is Ccc2p, which is required for export of cytosolic copper to the extracytosolic domain of the copper-dependent oxidase Fet3p (54). Prokaryotic genes exhibiting high levels of sequence similarity to the C. albicans CRD1 gene include known copper homeostatic genes copA of Enterobacter hirae (34), involved in copper import, and pacS of Synechococcus sp. strain PCC7942 (20), involved in copper export and resistance.

Previously characterized P1-type ATPases from yeast and prokaryotes involved in copper homeostasis exhibit tightly controlled transcriptional regulation that is consistent with their physiological roles. For example, the PacS protein of Synechococcus is induced upon exposure of the cells to excess copper (20). In contrast, the S. cerevisiae genes CTR1, CTR3, FRE1, and CCC2, which are involved in importing copper into the cell or delivering copper to the secretory pathway, are highly expressed in conditions of copper limitation; conditions of copper excess cause transcription of these genes to be strongly repressed (14, 24, 53). The CUP1 and MTII_A/B genes, encoding the major copper resistance MTs of S. cerevisiae and C. glabrata, respectively, are induced by growth in the presence of copper (5, 29, 30). C. albicans CRD1 is also induced by growth in the presence of copper, consistent with the observation that CRD1 confers copper resistance.

The cellular function(s) of the C. albicans MT gene CRD2 is undefined. However, physiological analyses of crd2 null mutants indicate a minor role for CRD2 in resistance to copper ion toxicity. The crd2/crd2 homozygous null strains exhibited a kinetic growth defect in that they grew more slowly in the presence of copper but ultimately reached wild-type levels of growth. Expression of the CRD2 gene is not induced in the presence of copper. Since intracellular copper can cause damage to DNA, proteins, and lipids, the basal expression of CRD2 may permit immediate binding of excess intracellular copper until the CRD1 efflux pump can be induced. The kinetic growth defect seen in the crd2 disruption strains may reflect the time required to synthesize and localize sufficient Crd1p and/or repair oxidative damage. Under normal growth conditions and low environmental copper concentrations, C. albicans synthesizes the CRD2 MT and does not express CRD1 at appreciable levels. This may permit C. albicans to maintain an intracellular reservoir of the essential trace element copper, and Crd2p may function in the proper distribution of copper to the various enzymes which require this trace element (36). The lack of induction in the presence of copper exhibited by the C. albicans CRD2 gene and its lack of amplification (data not shown) are consistent with a minor role in overall copper resistance. In addition to its role in copper homeostasis, C. albicans Crd2p may perform other functions in cellular metabolism. It has been demonstrated that S. cerevisiae Cu-Cup1p exhibits antioxidant activity, and Cu-Crd2p may also perform this role (36, 49).

Recently, a copper binding protein with significant sequence similarity to mammalian MTs was isolated from C. albicans cells grown in copper-containing medium (35). This protein is not identical to Crd2p reported in this study. While this work was in review, there appeared a paper by Weissman et al. describing the cloning of CaCRP1 (encoding a P1-type ATPase) and CaCUP1 (encoding a 33-amino-acid MT) involved in resistance to copper ion toxicity in C. albicans (51). The CaCUP1 gene was induced by copper and confers low-level resistance to copper ion toxicity. The deduced amino acid sequence of CaCup1p corresponded to that of the copper binding protein isolated by Oh et al. (35). CaCRP1 is essentially identical to CRD1 described in this study, and Weissman et al. demonstrated localization of CaCrp1p/Crd1p to the plasma membrane and the role of this P1-ATPase in maintaining low intracellular copper, consistent with our hypothesis that this protein functions in copper ion export.

Jensen et al. pointed out the striking similarities in terms of gene structure and regulation between the S. cerevisiae CUP1 and C. glabrata MTII_A/B genes and the S. cerevisiae CRS5 and C. glabrata MTI genes (19). The CUP1 gene of S. cerevisiae, the MTII_A/B gene of C. glabrata, and the CRD1 gene of C. albicans are all involved in high-level copper resistance and show high-level induction in conditions of excess copper (4, 21, 28, 30). In contrast, the S. cerevisiae CRS5 gene, C. glabrata MTI gene, and the C. albicans CRD2 gene contribute minimally to copper resistance, are single copy with no amplification, exhibit basal transcription in the absence of copper, and exhibit minimal transcriptional induction compared to that seen with the primary copper resistance determinants (8, 28, 30). Although the mechanism whereby high-level copper ion resistance is achieved in C. albicans is quite different from that of S. cerevisiae or C. glabrata, we suggest that the physiological role of the Crd1p ATPase of C. albicans is to detoxify copper ions, similar to the role of the S. cerevisiae Cup1p and C. glabrata MTII_A/B MTs, while the C. albicans Crd2p plays a role in copper homeostasis similar to that of the S. cerevisiae Crs5p and C. glabrata MTI MTs.

This hypothesis raises the question of why C. albicans utilizes an ATPase transporter as opposed to an MT as its primary means of copper resistance. Interestingly, the use of a P1-type ATPase may have important implications for nosocomial infections. The crd1 null mutant exhibits increased sensitivity to silver nitrate (Fig. 5C). Several Cu-transporting ATPases have been shown to be induced by and to transport silver ions in addition to copper (20, 33, 46). Generally this has been thought to be of no biological significance and a fortuitous occurrence due to the structural similarity between copper and silver ions (20, 46, 48). However, silver has been used as a microbiocidal agent for over a century and silver products are commonly used in the medical and dental fields (15). Silver-impregnated cloth and topicals containing silver sulfadiazine are used in hospital burn wards, and catheters containing silver as an antiseptic are widely used (9, 12, 22, 27, 32, 38). C. albicans is a frequent nosocomial infectious agent in burn patients and patients with indwelling catheters. Crd1p, which contributes to the ability of C. albicans to resist silver ion toxicity, may enhance the prevalence of C. albicans as a nosocomial pathogen.

Recently, a silver-resistant strain of Salmonella isolated from a hospital burn unit was shown to harbor a plasmid that contained multiple silver resistance genes linked in an approximately 14-kb region (16). One of the ORFs in this region encoded a P1-type ATPase (SilP) that contributes to silver resistance and that shares high sequence similarity with Cu-transporting ATPases (16). Thus, silver resistance mediated by ATPase pumps appears to make an important contribution to the virulence of nosocomial pathogens.