Identification of a Candida albicans Ferrichrome Transporter and Its Characterization by Expression inSaccharomyces cerevisiae *

Saccharomyces cerevisiaecan accumulate iron through the uptake of siderophore-iron. Siderophore-iron uptake can occur through the reduction of the complex and the subsequent uptake of iron by the high affinity iron transporter Fet3p/Ftr1p. Alternatively, specific siderophore transporters can take up the siderophore-iron complex. The pathogenic fungus Candida albicans can also take up siderophore-iron. Here we identify aC. albicans siderophore transporter, CaArn1p, and characterize its activity. CaARN1 is transcriptionally regulated in response to iron. Through expression studies in S. cerevisiae strains lacking endogenous siderophore transporters, we demonstrate that CaArn1p specifically mediates the uptake of ferrichrome-iron. Iron-ferrichrome and gallium-ferrichrome, but not desferri-ferrichrome, could competitively inhibit the uptake of iron from ferrichrome. Uptake of siderophore-iron resulting from expression of CaARN1 under the control of theMET25-promoter in S. cerevisiae was independent of the iron status of the cells and of Aft1p, the iron-sensing transcription factor. These studies demonstrate that the expression of CaArn1p is both necessary and sufficient for the nonreductive uptake of ferrichrome-iron and suggests that the transporter may be the only required component of the siderophore uptake system that is regulated by iron and Aft1p.

Iron is an essential element for all eukaryotes and most prokaryotes. Its importance in biology lead to the evolution of multiple uptake mechanisms that would satisfy the requirements for the metal in both single cell and multicellular organisms. Under conditions of iron starvation most microorganisms secrete siderophores, low molecular weight organic molecules that bind extracellular iron (1). Siderophores are chemically heterogeneous and there are specific transport systems for different siderophores (2,3). In some instances a transport system may take up more than one siderophore. Many microorganisms secrete more than one siderophore, and in addition to utilizing their own siderophores they may take up iron complexes of siderophores secreted by other microorganisms.
The budding yeast, Saccharomyces cerevisiae, does not secrete siderophores, yet it can use siderophore-iron. Siderophore-iron uptake can be accomplished by either extracellular reduction and subsequent uptake of the iron by the high affinity iron transport system Fet3p/Ftr1p or by the uptake of siderophore-iron complexes by specific transporters belonging to the major facilitator super family (4,5). The hydroxamatetype siderophores ferrioxamine B, triacetyl fusarinine C, and ferrichrome are taken up by the S. cerevisiae siderophore-iron transporters Arn1p, Arn2p, and Arn3p (5)(6)(7)(8). Arn4p was found to be specific for the catecholate-type bacterial siderophore enterobactin (9).
Iron uptake mechanisms are highly conserved among yeast. Schizosaccharomyces pombe and Candida albicans have a high affinity uptake system that is homologous to the Fet3p/Ftr1p transport system of S. cerevisiae (10 -12). Unlike S. cerevisiae, the pathogenic fungus C. albicans can secrete siderophores (13). To date no specific siderophore uptake system in C. albicans has been identified. Based on studies of siderophore-iron utilization in S. cerevisiae, we hypothesized that C. albicans may have siderophore transporters. In this paper, we demonstrate the existence of a specific siderophore transporter in C. albicans, CaArn1p, encoded by the ORF 1 CaYHL040C, 2 which is orthologous to the ARN1 siderophore transporter in S. cerevisiae. Genetic analysis in C. albicans is difficult because it lacks a sexual cycle and exists as a diploid. S. cerevisiae has been used successfully as a tool to characterize C. albicans genes. The elemental iron transport systems from other yeast species retain function when expressed in S. cerevisiae (10 -12). Characterization of CaARN1 in S. cerevisiae shows that the C. albicans siderophore transporter has high specificity for ferrichrome and that expression of the transporter is necessary and sufficient for siderophore-iron transport.

EXPERIMENTAL PROCEDURES
Yeast Strains, Plasmids, and Media-Yeast strains used in this study are listed in Table I. The following primers were used to construct pmetCaARN1 (5Ј-CGGGATCCGCATGACATCTTACCAG-3Ј and 5Ј-CG-GAATTCCTATTAAACAGCTACTCTTTTCTTC-3Ј), pmet-flagCaARN1 (5Ј-CGGGATCCGCATGGATTACAAGGATGACGATGACAAGGGTGG-TACATCTTACCAG-3Ј and 5Ј-CGGAATTCCTATTAAACAGCTACTCT-TTTCTTC-3Ј), and the pmet-CaARN1-myc (5Ј-CGGGATCCGCATGAC-* This work was supported in part by National Institutes of Health Grant NDDK-DK 30534. Support for use of Core facilities for oligonucleotide synthesis and DNA sequencing was provided by a NCI, National Institutes of Health Grant NCI 5 P30 CA 42014. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  ATCTTACCAG-3Ј and 5Ј CGGAATTCCTATTAATTCAAGTCCTC-TTCAGAAATGAGCTTTTGCTCCATAACAGCTACTCTTTTCTTC-3Ј).
C. albicans (CAI-4) genomic DNA was used as a template to amplify ORF CaYHL040C by PCR using 92°C, 40 s, 55°C, 60 s, and 68°C for 3 min conditions. The PCR products were digested with BamHI and EcoRI and cloned behind a MET25 promoter in the vectors pTF62 (LEU2) or pTF63 (URA3). All constructs were verified by DNA sequencing. pARN1-HA and pmetARN1-HA were as described (6).
Cells were grown in YPD (1.0% yeast extract, 2.0% peptone, 2.0% glucose), CM (a synthetic medium of yeast nitrogen base, amino acids, and glucose), CM deficient in specific amino acids, CM made ironlimited by the addition of 50 M bathophenanthroline sulfonate (BPS) (14) or CM made iron replete by addition of 50 M FeCl 3 .
For plate assays, iron-limiting CM was used in which BPS was added to a final concentration of 50 M, 100 l of Fe-siderophore complexes were added to a final concentration of 0.5 M, and the plates allowed to dry at 30°C overnight. Yeast were grown to mid-log phase and 10-fold dilutions spotted onto plates and grown for 2 or 3 days at 30°C.
Iron Uptake Assay-Iron transport assays were performed as described (17) with the following modifications. Briefly, 2 ϫ 10 7 cells were mixed with 59 Fe supplied as 59 FeCl 3 or 59 Fe-siderophore. Cells were incubated at 30°C for 15 min, placed on filters (Whatman GF/C), and washed with EDTA-containing buffer to remove unincorporated iron. The filters were air-dried, and associated radioactivity determined. The uptake activity was expressed as pmol of 59 Fe/min/10 6 cells.
Western Blot Analysis-Cells were grown to mid-log phase, collected by centrifugation, and washed. Spheroplasts were made and Dounce homogenized, and a postnuclear supernatant obtained. A crude membrane fraction was isolated by centrifugation at 15,600 ϫ g for 30 min. Protein concentrations were determined, and equal amounts run on 10% SDS-polyacrylamide gel electrophoresis. Samples were transferred to nitrocellulose, probed with anti-FLAG (M2 Sigma), (1:10,000) or anti-HA (HA.11Covance) (1:2000) followed by goat anti-mouse horseradish peroxidase (Jackson Immuno Research Laboratories Inc, 1:10,000). Western blots were developed using the chemiluminescence detection reagent Renaissance (PerkinElmer Life Sciences) as per the manufacturer's instructions.
Northern Blot Analysis-Total RNA was isolated using standard techniques (18). All samples were isolated from mid-log phase cultures grown in defined iron media containing either 50 M FeCl 3 ("high iron") or 50 M BPS ("low iron"). A biotinylated CaARN1 probe was PCRgenerated using a C. albicans (CAI-4) genomic prep as a template, primers 5Ј-CGGGATCCGCATGACATCTTACCAG-3Ј and 5Ј-CGGAAT-TCCTATTAAACAGCTACTCTTTTCTTC-3Ј with PCR conditions 92°C for 40 s, 55°C for 60 s, and 68°C for 3 min. A biotinylated C. albicans actin probe was generated using 5Ј-TTGCCGGTGACGACGCTCC and 5Ј-GCTCTGAATCTTTCGTTACC. The blots were developed using DNA Detector kit (Kirkegaard & Perry Laboratories, Inc.). The size of the mRNAs on Northern blots correlated with the lengths expected from the calculated sequences.

RESULTS
C. albicans has been shown to use different hydroxamate siderophores for its growth on plates (19). To determine whether C. albicans has a specific siderophore transport system, we measured the uptake of 59 Fe-siderophore complexes in C. albicans. Wild type cells were able to take up 59 Fe from both 59 Fe-Fc and 59 Fe-FOB (Fig. 1a) in a concentration dependent manner. The apparent K m of uptake was 1.88 M and V max 3.20 pmol/min/10 6 cells for Fc and 5.71 M and 2.40 pmol/min/10 6 cells, respectively, for FOB, values that are similar to that seen in S. cerevisiae (6). Both siderophores were found to promote cell growth (data not shown).
In S. cerevisiae siderophore-iron is accumulated by two mechanisms (4,5). The first involves reduction of siderophoreiron complexes by cell surface reductases and uptake of iron by the high affinity iron transport system. A second mechanism for siderophore-iron utilization results in the uptake of siderophore-iron complexes by specific transport systems. To determine whether C. albicans also employs both mechanisms we examined the effect of a non-permeable Fe(II) chelator on siderophore-iron-mediated transport. Reduction of siderophoreiron complexes leads to the formation of Fe(II), which is the substrate for the high affinity iron transport system and chelation of Fe(II) prevents iron uptake. Incubation of C. albicans with the impermeable Fe(II) chelator BPS results in the complete inhibition of 59 Fe uptake when cells are incubated with 59 Fe(III)Cl 3 in citrate buffer (Fig. 1b). This same concentration only effects a 50% reduction in the uptake of 59 Fe(III)-Fc. Uptake of 59 Fe-FOB is completely inhibited by the addition of BPS. These results suggest that a significant component of Fe-Fc uptake does not occur by reduction at the cell surface, whereas uptake of Fe-FOB only occurs by reduction followed by transport of elemental iron.
S. cerevisiae has a family of siderophore transporters (5). We questioned whether C. albicans also had an orthologous siderophore transport system. Inspection of the C. albicans genome at 5ϫ coverage (sequence.stanford.edu/group/candida/) revealed one ORF that demonstrated a high homology to the S. cerevisiae siderophore transporters (ARN1-4) (Fig. 2). This ORF was on the 33,181 base pair contig 4 -3057 and was from base pairs 15,417-13,606 on the reverse strand. The amino acid identity between the putative C. albicans ARN, CaYHL040C, and the yeast ARNs ranged between 28 -46%, with ARN1 showing the highest identity and homology (46 and 63%, respectively). Primers were designed, and CaYHL040C was amplified using PCR from a C. albicans genomic DNA prepara- tion. Southern analysis of C. albicans genomic DNA, using CaYHL040C as a probe, was performed using non-stringent conditions (2ϫ saline/sodium phosphate/EDTA /0.1% SDS). A single band was detected, consistent with the genome search, strongly suggesting that there is just one ARN homologue in C. albicans.
The S. cerevisiae siderophore transporters are regulated by the iron-sensing transcription factor Aft1p (5). If the C. albicans gene CaYHL040C is a siderophore transporter, then we might expect that its transcription would be iron-regulated, as are other known components of the C. albicans iron transport system (12,20). A Northern analysis was performed on C. albicans RNA isolated from either iron-replete or iron-depleted cultures using CaYHL040C as a probe. A single band was detected in cells grown under iron-limiting conditions (Fig. 3). No band was observed under high iron conditions. This finding supports the hypothesis that the C. albicans ORF CaYHL040C, termed CaARN1, may have a role in iron metabolism in this fungus.
Expression and Characterization of CaARN1 in S. cerevisiae-To study the function of CaARN1, independent of iron conditions, we cloned CaARN1 under the regulation of the MET25 promoter and expressed it in a ⌬fet3⌬arn1-4 strain of S. cerevisiae. These cells are unable to take up siderophore iron either through the high affinity iron transport system or the ARN family of siderophore transporters and were used previously to characterize S. cerevisiae siderophore transporters (5). The ⌬fet3⌬arn1-4 cells can not grow on BPS plates supplemented with siderophores (5). Transformation of this strain with a plasmid containing a MET25-regulated CaARN1, permitted cell growth on media supplemented with Fe-Fc, but not supplemented with Fe-FOB (Fig. 4) or Fe-rhodotorulic acid (data not shown). Both N-terminal and C-terminal epitopetagged CaARN1 constructs also permitted growth on Fe-Fc.
To further characterize the iron uptake activity of cells transformed with CaARN1, uptake studies were conducted using 59 Fe-Fc. In ⌬fet3⌬arn1-4 cells, the uptake of 59 Fe-Fc was significantly higher in cells transformed with the methionineregulated CaARN1 constructs than in cells transformed with vector alone. The rate of iron accumulation in cells expressing CaArn1p (under the control of the MET25 promoter) was much greater than ⌬fet3 cells expressing the native ARNs under iron-limited conditions, which should result in their maximal induction (Fig. 5a). There were no significant differences in 59 Fe-Fc uptake between cells expressing the epitope-tagged protein (5ЈFLAG or 3ЈMyc) and the non-tagged CaArn1p. Addition of methionine to the growth media reduced the rate of 59 Fe-Fc uptake (Fig. 5b), although not to baseline levels. Western analysis of cells expressing CaArn1p showed regulation of CaARN1 expression (Fig. 5c). At high levels of methionine (560 g/ml), no CaArn1p was detected by Western blot, although 59 Fe-Fc uptake was still observed. This uptake may be due to the lack of complete repression of the MET25 promoter, particularly in high copy vectors.
We utilized pmetCaARN1-expressing cells to define the specificity of Fc recognition and transport. Addition of desferri-Fc to 59 Fe-Fc-containing media did not inhibit uptake of 59 Fe-Fc (Fig. 6). This result indicates that the transporter must recognize a structural difference in the Fc once it has bound iron. It has been reported that there is little conformational difference between Fe-and Ga-complexed hydroxamate siderophores (21). Both metals are bound in the trivalent state, although unlike Fe-Fc, Ga-Fc complexes cannot be reduced. Both Fe-Fc and Ga-Fc exhibited a similar concentration-dependent inhibition of 59 Fe-Fc uptake (Fig. 6). Mineral Ga, supplied as Ga(NO 3 ) 3 had no effect on the uptake of 59 Fe-Fc. The inhibition of 59 Fe-Fc uptake by Ga-Fc was transient, as cells washed free of Ga-Fc showed no subsequent inhibition of 59 Fe-Fc uptake (data not shown). These results suggest that Ga-Fc competes with Fe-Fc for a recognition site on CaArn1p. Ferrichrome synthetic analogues are recognized and taken up by the Ustilago maydis ferrichrome uptake system (15,16). These analogues were not taken up by pmet-CaARN1-transformed ⌬fet3⌬arn1-4 cells, as ascertained by growth, radiotracer experiments, and fluorescence microscopy (data not shown). These results demonstrate that the activity of CaArn1p is sensitive to alterations in siderophore structure.
Ferrichrome Uptake in Iron-depleted and Iron-replete Cells-The high affinity iron transport system, comprised of Fet3p and Ftr1p, is regulated at the level of transcription by Aft1p (22). Expression of FET3 and FTR1 by iron-independent promoters showed that these proteins mediated iron transport even when cells were iron-replete (10). These results indicate that the only two surface proteins required for high affinity iron transport are Fet3p and Ftr1p. Because the S. cerevisiae siderophore transporters are also regulated by Aft1p, we asked whether other proteins regulated by Aft1p are required for siderophore transport. The S. cerevisiae strain ⌬fet3⌬arn1-4 was transformed with plasmids containing S. cerevisiae ARN1 under its own promoter, pmetARN1 or pmetCaARN1. Cultures were grown under high iron or iron-limiting conditions for 6 h and then assayed for 59 Fe-Fc uptake (Fig. 7a). The 6-h incubation period is long enough to permit expression of Aft1p-regulated genes (14). As expected, cells expressing Arn1p under the control of its native promoter, showed Fe-Fc uptake when grown under iron-depleted conditions but not under iron-replete conditions. Expression of either ARN1 or CaARN1 under the control of the MET25 promoter led to a high rate of Fe-Fc uptake regardless of whether cells were grown in high or low iron media. The rate of iron uptake was similar in cells expressing either the S. cerevisiae ARN1 or the C. albicans ARN1. These results suggest that once expressed, the siderophore transporter can function independently of cellular iron or other genes whose transcription is dependent on the Aft1p transcription factor. This conclusion is further supported by measuring 59 Fe-Fc uptake in ⌬aft1 cells transformed with pARN1, pmetARN1, or pmetCaARN1 (Fig. 7b. There is little iron taken up by ⌬aft1 cells transformed with a plasmid that has ARN1 under the control of the Aft1p-sensitive promoter. The same cells transformed with the MET25-regulated ARN1 showed high rates of iron accumulation. When the ARNs are expressed from their endogenous promoters there is little measurable siderophore iron transport in the absence of AFT1. Siderophore iron, however, can support the growth of S. cerevisiae ⌬aft1 cells (Fig. 8). Recently, a homologue of AFT1, termed AFT2, has been identified, and it appears to share some functions with AFT1 (23). To determine whether AFT2 played a role in siderophore iron transport, we examined a ⌬aft2 strain for its ability to grow on siderophore iron. Deletion of both AFT1 and AFT2 results in a lack of growth on Fe-Fc. In the absence of the Arnp's or Fet3p there is no growth on Fe-Fc ( Fig. 8 and Ref. 6). These results suggest that in S. cerevisiae AFT2 may permit a low level of expression of ARN1.

DISCUSSION
Siderophore-mediated iron accumulation is an important mechanism of iron uptake for many organisms including bacteria, fungi, and plants (2,25). Numerous studies have shown that siderophore transport systems are virulence factors in microorganisms (26 -28). S. cerevisiae, which does not secrete siderophores, has multiple transport systems that can utilize siderophores secreted by other organisms (5-9, 29). C. albicans, a pathogenic fungus of major medical importance, can secrete siderophores of the hydroxamate-and possibly phenolate-type (30). We took advantage of the sequence of S. cerevisiae siderophore transporters to identify a siderophore transporter in C. albicans. A gene, homologous to the S. cerevisiae ARNs was shown to be iron-regulated in C. albicans, consistent with a role in iron homeostasis. Proof that this gene, CaARN1, could mediate iron transport was shown by its ability to complement a strain of S. cerevisiae that was unable to grow on siderophores due to the deletion of both the high affinity iron transport system and the native siderophore iron transporters. Through both growth assays and direct measurement of iron transport, we demonstrated that ectopic expression of the C. albicans gene permitted 59 Fe-Fc uptake.
A data base search of 5ϫ DNA sequence coverage of the C. albicans genome identified a single C. albicans homologue of the S. cerevisiae ARN gene family. Both Southern and PCR analysis also showed a single C. albicans ARN homologue. This gene, at least as expressed in S. cerevisiae, was highly specific for ferrichrome. This result leads to the suggestion that C. albicans may only have a transporter specific to the siderophores it secretes. S. cerevisiae, which does not secrete siderophores, has multiple transporters that permit utilization of a broad range of siderophores. Perhaps the ferrichrome transporter is an ancestral transporter, which under evolutionary pressure and gene duplication led to the generation of the other siderophore transporter genes. C. albicans, because of its ability to secrete siderophores, may not have been subject to the same pressure and consequently retained a single siderophore transporter gene. This hypothesis could be tested by examination of the fission yeast S. pombe. There is no evidence that this yeast secretes siderophores, yet it has three homologues of the ARN family. If these homologues can be shown to be functional siderophore transporters, then it would strengthen the argument that a lack of siderophore synthesis is the driving force behind the diversity in siderophore transporter gene evolution.
We took advantage of the expression of CaArn1p in S. cerevisiae to examine some of the characteristics of siderophore transport. Based on competition studies, CaArn1p recognized metallo-Fc but not desferri-Fc. Ga-Fc, which can not be reduced, blocked the uptake of 59 Fe-Fc at concentrations comparable with that of Fe-Fc. This competition was transient as removal of Ga-Fc alleviated the block in 59 Fe-Fc uptake. Analogues of ferrichrome were previously shown to be biologically active in the ferrichrome-producing fungus U. maydis (15,16). While it has been suggested that S. cerevisiae can accumulate these siderophores and their fluorescently labeled derivatives (31), we have not seen uptake or growth promotion by the fluorescent ferrichrome analogues in cells expressing either the CaARN1p or the endogenous S. cerevisiae ARN1p. This result suggests that there may be significant differences in the recognition of ferrichrome by different fungal siderophore transporters.
The ability to regulate the synthesis of siderophore transporters independent of cellular iron status also provided data on the role of AFT1 and iron in siderophore transporter expression and function. Deletion of AFT1 dramatically reduces the transcription of genes encoding the elemental iron transporters or proteins required for their assembly. Genes that encode siderophore transporters (ARNs) are also highly transcriptionally regulated (5). While the recently identified AFT2 (23) may share some functions with AFT1, it is clear that AFT1 is the major iron-sensing transcriptional factor in S. cerevisiae. Expression of either the endogenous ferrichrome transporter (ARN1) or the C. albicans ferrichrome transporter (CaARN1) by iron-independent promoters resulted in a functional siderophore transport system. Under high iron conditions and in the absence of AFT1, Fe-Fc uptake occurred. This result suggests that the transporter is the only iron-regulated gene that is a structural component of the siderophore uptake system. C. albicans is an important pathogen. The ability to study siderophore transport in a well defined system (i.e. genetically characterized S. cerevisiae) offers an opportunity to study the recognition properties of the pathogen's siderophore transporter. This may help in the development of siderophore analogs with antimicrobial properties (32,33). While this development may be done empirically by screening analogs, knowledge of the mechanism of siderophore-iron uptake would accelerate this approach. It is still unresolved whether the iron-siderophore complex is transported into the cell or whether iron is released from the siderophore as a consequence of binding to the transporter. Studies suggest that iron can be released from siderophores by either reduction through cell surface reductases, by intracellular-reducing agents, or as shown recently through ligand exchange with endogenous siderophores (34). A major question remaining is the mechanism of siderophore iron FIG. 7. Arn1p-mediated 59 Fe-Fc uptake in iron-replete or irondepleted conditions. S. cerevisiae ⌬fet3, ⌬fet3⌬arn1-4 (a) or S. cerevisiae ⌬aft1 (b) transformed with ARN1, pmetARN1, or pmetCaARN1 were cultured in methionine-deficient media with 50 M FeCl 3 (iron replete) or 50 M BPS (iron-depleted). Cells were grown to mid-log, and uptake of 59 Fe-Fc was measured.
FIG. 8. Siderophore iron-dependent growth in ⌬aft1, ⌬aft2, or ⌬aft1⌬aft2 strains of S. cerevisiae. Indicated strains were grown to mid-log and then plated in serial dilutions on CM, CM-BPS, or CM-BPS with added siderophore iron. Plates were incubated for 2 days at 30°C, and colonies examined for growth.
release. An answer to this question will affect the development of toxic-siderophores. A release of siderophore iron while the siderophore is bound to the surface of the transporter, as opposed to being transported into the cell prior to iron release, would affect the classes of toxic compounds that might be made. Our results suggest that whatever the mechanism of siderophore iron release, it is highly similar for both C. albicans and S. cerevisiae, and perhaps for other fungi as well.