Glutathione Utilization by Candida albicans Requires a Functional Glutathione Degradation (DUG) Pathway and OPT7, an Unusual Member of the Oligopeptide Transporter Family

Background: Glutathione biosynthesis is essential for pathogenesis in Candida albicans. Exogenous glutathione utilization is not known. Results: The DUG pathway of glutathione degradation and an unusual transporter, OPT7, function in glutathione utilization. Conclusion: Glutathione utilization is efficient but not essential for survival in vivo. Significance: It enables one to understand how this pathway contributes toward homeostasis of redox and sulfur in a yeast pathogen. Candida albicans lacks the ability to survive within its mammalian host in the absence of endogenous glutathione biosynthesis. To examine the ability of this yeast to utilize exogenous glutathione, we exploited the organic sulfur auxotrophy of C. albicans met15Δ strains. We observed that glutathione is utilized efficiently by the alternative pathway of glutathione degradation (DUG pathway). The major oligopeptide transporters OPT1–OPT5 of C. albicans that were most similar to the known yeast glutathione transporters were not found to contribute to glutathione transport to any significant extent. A genomic library approach to identify the glutathione transporter of C. albicans yielded OPT7 as the primary glutathione transporter. Biochemical studies on OPT7 using radiolabeled GSH uptake revealed a Km of 205 μm, indicating that it was a high affinity glutathione transporter. OPT7 is unusual in several aspects. It is the most remote member to known yeast glutathione transporters, lacks the two highly conserved cysteines in the family that are known to be crucial in trafficking, and also has the ability to take up tripeptides. The transporter was regulated by sulfur sources in the medium. OPT7 orthologues were prevalent among many pathogenic yeasts and fungi and formed a distinct cluster quite remote from the Saccharomyces cerevisiae HGT1 glutathione transporter cluster. In vivo experiments using a systemic model of candidiasis failed to detect expression of OPT7 in vivo, and strains disrupted either in the degradation (dug3Δ) or transport (opt7Δ) of glutathione failed to show a defect in virulence.

sources in the medium. OPT7 orthologues were prevalent among many pathogenic yeasts and fungi and formed a distinct cluster quite remote from the Saccharomyces cerevisiae HGT1 glutathione transporter cluster. In vivo experiments using a systemic model of candidiasis failed to detect expression of OPT7 in vivo, and strains disrupted either in the degradation (dug3⌬) or transport (opt7⌬) of glutathione failed to show a defect in virulence.
Candida albicans is the most important yeast pathogen to cause systemic fungal infections in humans (1)(2)(3). The ability of this yeast to effectively survive the harsh host environments suggests the presence of stress response capabilities that are likely to be unique to this yeast. Oxidative stress is one of the major stress conditions faced by yeast pathogens in vivo (4), and an effective response is required by the yeast pathogen to defend itself from the host immune system. C. albicans has both a thioredoxin-dependent and a glutathione-dependent pathway of redox homeostasis (5). Although the redox functions of these pathways have been shown to partially overlap in S. cerevisiae (6,7), disrupting either the thioredoxin pathway (Trx1 knockouts) (8) or the glutathione biosynthetic pathway alone in C. albicans (Gsh1 knockouts) leads to attenuated virulence (9).
The abundance of glutathione in the human host and the presence of apparent homologues of glutathione utilization pathways of S. cerevisiae in C. albicans have made the essentiality of glutathione biosynthesis for the survival of C. albicans in vivo surprising. Two possible explanations could be put forward to explain these observations. The first is that the high levels of intracellular glutathione that are required by these yeast pathogens in vivo might demand continuous glutathione biosynthesis. Glutathione levels in all eukaryotic cells reach up to 10 mM (10), and these levels might be difficult to reach only from extracellular sources. A second explanation for the essentiality of glutathione biosynthesis in vivo might be the absence of an efficient glutathione utilization pathway in C. albicans, as a consequence of which glutathione, though abundant in the environment, would be inefficiently utilized by the yeast. Exogenous glutathione utilization requires both a transporter and a degradation pathway. Glutathione transporters in several yeasts have now been identified and found to belong to the peptide transporter clade of the oligopeptide transporter superfamily (11)(12)(13). Glutathione degradation has been shown to require the alternative pathway (DUG pathway) of glutathione degradation (14) whose homologues are present in all yeasts and fungi barring Schizosaccharomyces pombe. The DUG pathway involves the proteins DUG1, DUG2, and DUG3. The DUG1 has been shown to be a Cys-Gly peptidase (15), whereas DUG2 and DUG3 form a complex to cleave the ␥-glutamyl linkage of glutathione. 8 The C. albicans genome has revealed the presence of several members of the oligopeptide transporter (OPT) 9 family, OPT1-OPT8 (16). Of these, OPT1-OPT7 belonged to the peptide transporter (PT) clade, and among these OPT1 is the closest in similarity to the Saccharomyces cerevisiae glutathione transporter, HGT1. Despite the strong similarity to HGT1, OPT1 was shown not to play a role in glutathione transport as seen from heterologous complementation experiments (13). Thus it was possible that C. albicans lacks the ability to efficiently transport and utilize glutathione. This would be akin to the situation in Candida glabrata that completely lacks the homologues of the oligopeptide transporter superfamily belonging to the PT clade and in fact fails to transport and utilize glutathione when provided from external sources (9).
Considering the importance of glutathione and redox pathways for yeast pathogens in vivo, we sought to resolve these issues, and in this study we have investigated glutathione transport and utilization in C. albicans. The construction of a met15⌬ strain in C. albicans that was an organic sulfur auxotroph revealed that C. albicans could utilize glutathione as a sulfur source efficiently, and we have used this phenotype for facilitating the further analysis of glutathione utilization in this yeast. We investigated the role of the OPTs as well as the DUG pathway, and our studies revealed that C. albicans can utilize glutathione efficiently, and the transport is dependent on an unusual member of the OPT family, OPT7, whereas degradation was dependent on the DUG pathway. The possible importance of these pathways in the virulence and survival of C. albicans in vivo was also investigated in a mouse model of systemic candidiasis.

EXPERIMENTAL PROCEDURES
Chemicals and Reagents-All of the chemicals used in this study were analytical grade and obtained from commercial sources. Media components were purchased from Difco Strains, Media, and Growth Conditions-The Escherichia coli strain DH5␣ was used as a cloning host. S. cerevisiae and C. albicans strains used in the study are described in (Table 1). S. cerevisiae and C. albicans strains were regularly maintained on yeast extract, peptone, and dextrose medium (YPD). Synthetic defined minimal medium (SD) contained yeast nitrogen base, ammonium sulfate, and dextrose supplemented with histidine, leucine, and lysine (as per requirement) at 50 mg/liter. Glutathione was added as required (17). For C. albicans, synthetic defined medium contained ammonium chloride instead of ammonium sulfate and was supplemented with either methionine, cysteine, cystine, or glutathione (as per requirement). Mycophenolic acid plates were prepared by adding 1 mg/ml of mycophenolic acid in YPD medium. The molecular techniques used in the study were according to standard protocols (18).
Construction of Disruption Cassettes of MET15 in C. albicans-The MET15 gene of C. albicans was cloned by amplifying the ORF including 100 bp upstream and downstream of the ORF by PCR with SmaI sites introduced into the primers and cloning into SmaI site of pBSK. For preparing the two disruption cassettes for each of the alleles, we excised the SAT1 flipper from pSFS2a by ApaI-SacI, blunted and cloned into the MscI and BsmI sites. A second disruption cassette was made by PCR amplifying an internal piece of the MET15 clone corresponding to, but internal to, the site and inserting the SAT1 flipper into the HpaI-BstEII site of this smaller MET15 fragment. To use the first cassette for disruption, we evicted the same with SmaI and transformed into the appropriate C. albicans strain. To create a disruption in the second allele, we first evicted the SAT1 flipper from the first strain by growth in maltose medium as described earlier (19). The second disruption cassette was excised with SmaI sites and then transformed into the strain containing a disruption in the first allele. The MET15 gene was disrupted in this manner in both alleles of C. albicans SC5314 to yield strain ABA 2490, in OPT7M4AB to yield strain ABA 3031, in OPT1M4AB to yield strain ABA 2546, in OPT12345M4AB to yield strain ABA 2548, and in ABA 2558 to yield strain ABA 2592. The disruptions were confirmed by diagnostic PCR, and the absence of a third allele was also confirmed by diagnostic PCR. The SAT1 flipper was evicted using maltose-containing medium as described previously.

Construction of Disruption Cassettes of DUG3 in C. albicans-
The DUG3 gene of C. albicans was cloned by PCR amplifying the ORF including 100 bp upstream and downstream of the ORF with SmaI sites introduced into the primers and cloning into the SmaI site of pBSK. For preparing the two disruption cassettes for each of the alleles, we excised the SAT1 flipper from pSFS2a by ApaI-SacI, blunted and cloned into the NdeI-HpaI sites. A second disruption cassette was made by PCR amplifying an internal piece of the DUG3 clone corresponding to, but internal to, the NdeI-HpaI site and inserting the SAT1 flipper into the BclI-MscI site of this smaller DUG3 fragment. To use the first cassette for disruption, we evicted the same with SmaI and transformed into the appropriate C. albicans strain. To create a disruption in the second allele, we first evicted the SAT1 flipper from the first strain by growth in maltose medium as described earlier. The second disruption cassette was excised with SmaI sites and then transformed into the strain containing a disruption in the first allele. In this way both alleles of DUG3 of C. albicans SC5314 were disrupted to yield strain ABA 2558. The disruptions were confirmed by diagnostic PCR, and the absence of a third allele was also confirmed by diagnostic PCR. The SAT1 flipper was evicted using maltose-containing medium.
Cloning of OPT7 in S. cerevisiae Expression Vector-The ORF OPT7 was PCR-amplified from C. albicans genomic DNA using the primer pairs OPT7F and OPT7R (supplemental Table S3). The 2.3-kb PCR product obtained was digested with Xba1 and BamHI and cloned downstream of the TEF promoter in the single copy, URA3-based expression vector to construct p416TEF-OPT7 (20). OPT7 was tagged with HA epitope by PCR mutagenesis. The reverse primer had the HA tag before the stop codon.
Growth Assay by Dilution Spotting-For growth assays, S. cerevisiae strains were grown overnight in minimal medium without uracil, whereas C. albicans strains were grown in SD minimal medium. The cells were reinoculated in fresh medium to an A 600 of 0.1 and grown up to exponential phase. The exponential phase cells were harvested washed with water and resuspended in water to an A 600 of 0.2. These were serially diluted to 1:10, 1:100, and 1:1000. 10 l of these cell resuspensions were spotted on minimal medium containing different concentration of either glutathione, ␥-Glu-Cys, ␣-Glu-Cys, Cys-Gly, Leu-Tyr-Met-Arg, GSSG, or methionine. The plates were incubated at 30°C for 2-3 days, and photographs were taken.
Glutathione Transport Assay-The S. cerevisiae ABC 817 strain (met15⌬hgt1⌬) transformed with appropriate plasmids was grown in minimal medium containing methionine and other supplements, without uracil. C. albicans strains met15⌬, met15⌬-opt7⌬, and cbf1⌬ were grown in SD minimal medium containing 20 M to 40 M methionine (nonrepressive sulfur conditions) or 1 mM methionine (repressive sulfur conditions) overnight. These cultures were reinoculated in the same medium and allowed to grow till they reached exponential phase. The cells were harvested and washed and put on ice in a pH 5.5 MES-buffered medium, until the transport was initiated. Transport experiments were carried out with [ 35 S]GSH as  (21).
For saturation kinetics, the initial rate of glutathione uptake was measured at different concentrations of glutathione with specific activity kept constant at each concentration. The initial rate of glutathione uptake was determined by measuring the radioactive glutathione in the cell. After subtracting background uptake observed with vector alone from uptake with P416TEF-OPT7, the data were plotted. The Lineweaver-Burke best fit line was plotted and was used to calculate the kinetic parameters K m and V max . The experiment was repeated a minimum of two times in duplicate at each glutathione concentration.
Inhibition studies were done with different peptides, Glu-Cys, Cys-Gly, Leu-Gly, Met-Gly, Gly-Gly-Gly, GSH, and GSSG. The inhibitors were added at 10-fold excess along with the assay medium, and the initial rates of glutathione were measured. For the studies done with the metabolic inhibitors, the cells were preincubated with the metabolic inhibitor for 15 min at 30°C before measuring the initial rate of glutathione uptake.
For determining the pH optima, the initial rates of glutathione uptake in the OPT7 were determined at different pH levels ranging from 3.5 to 7.5. The different pH levels were maintained using different buffers: 20 mM acetate-sodium acetate (pH 3.5, 4.0, and 4.5), 20 mM MES/KOH (pH 5.0, 5.5, and 6.0) and 20 mM HEPES/KOH (pH 6.5, 7.0, and 7.5). In each case, the buffer also contained 0.5 mM CaCl 2 , 0.25 mM MgCl 2 , and 2% glucose.
Indirect Immunofluorescence of the HA-tagged OPT7 by Confocal Microscopy-For localization studies, the P416TEF-OPT7HA construct was transformed in the S. cerevisiae met15⌬hgt1⌬ strain, and indirect immunofluorescence was performed using a previously published protocol (21).
Multiple Sequence Analysis and Phylogenetic Analysis-The OPT sequences were retrieved from Entrez. The multiple sequence alignment of the protein sequences was generated using the ClustalX program using default parameters (22). The MEGA 3.1 software (23) was used to visualize the phylogenetic tree of the family.
Experimental Infection in Mice-Experimentation with mice was done under the health and safety executive guidelines for level 2 biohazard containment. Male BALB/c mice aged 6 -8 weeks were used for the survival and IVET experiment. To investigate the virulence of C. albicans opt7⌬ and dug3⌬ strains in comparison with C. albcians WT (SC5314), both of the strains were grown on minimal medium for 16 h; further, the cells were washed twice with 1ϫ PBS and resuspended in the same buffer. To correlate the number of cells to 1 A 600 , hemocytometer cell counting was done. The mice were intravenously injected with 5 ϫ 10 5 cells of C. albicans WT SC5314, opt7⌬, and dug3⌬ and individually and observed for survival. For the in vivo expression technology experiments, two mice were sacrificed after every 2 days, the kidneys were isolated and homogenized in buffer, and an aliquot (after dilutions) was plated on selective plates.

C. albicans met15⌬ Are Organic Sulfur Organic Auxotrophs and Can Use Glutathione as a Source of Sulfur-The MET15
gene of S. cerevisiae codes for O-acetyl homoserine thiolase, an enzyme in the sulfur assimilation pathway. Deletion of this gene has been found to lead to organic sulfur auxotrophy in both S. cerevisiae and C. glabrata. met15⌬ strains have been previously created in C. albicans, and these disruptants were shown to be methionine auxotrophs, although prolonged growth on minimal medium permitted growth even on inorganic sulfate (24). However, the growth on other sulfur compounds was never examined. We decided to recreate the met15⌬ in a WT background to re-evaluate the earlier findings, and further to examine whether the strain would be useful to evaluate growth on glutathione. The met15⌬ strain in SC5314 background was created as described under "Experimental Procedures." These strains were not only methionine auxotrophs as described previously, but organic sulfur auxotrophs because the strain did not grow on inorganic sulfate but were able to grow on cysteine, cystine, and glutathione as a source of sulfur ( Fig. 1). They could also grow on homocysteine (data not shown). However, prolonged growth of met15⌬ led to growth even in the absence of any organic sulfur sources in accordance with the previous observations that it was not a strict auxotrophy (24). Despite this, and considering that the growth on organic sulfur sources was much faster, they have proved useful in investigating growth on sulfur sources. The ability to grow on other sulfur sources was shared with S. cerevisiae and C. glabrata. The only difference was that the C. glabrata met15⌬ strain could not utilize glutathione as a sulfur source, but C. albicans met15⌬ strain could efficiently utilize glutathione as a sulfur source. We used this glutathione-dependent growth of met15⌬ strains of C. albicans to identify and investigate the glutathione transport and degradation pathways.
The Oligopeptide Transporters (OPT1-OPT5) Are Not Involved in Glutathione Transport in C. albicans-The first step in the glutathione utilization pathway of yeasts is transport. In S. cerevisiae glutathione transport is mediated by the glutathione transporter Hgt1p, which belongs to the family of oligopeptide transporters (11). The absence of glutathione utilization in C. glabrata was shown to be due to the absence of homologues of HGT1 (9). Analysis of the C. albicans genome FIGURE 1. C. albicans met15⌬ are organic sulfur organic auxotrophs and can use glutathione, cysteine, and cystine as sulfur sources. The C. albicans WT and met15⌬ strains were grown to exponential phase in minimal medium containing methionine, harvested, washed, resuspended in water, and serially diluted to give 0.2, 0.02, 0.002, and 0.0002 A 600 of cells. 10 l of these dilutions were spotted on minimal ammonium chloride medium containing methionine, glutathione, cysteine, and cystine. The photographs were taken after 2 days of incubation at 30°C.
has revealed the presence of eight members of the oligopeptide transporter family. Seven of these belong to the peptide transporter clade (OPT1-OPT7), whereas OPT8 belonged to the more remote YSL (yellow stripe-like) clade. The PT clade members include the previously identified ORFs: OPT1, OPT2, OPT3, OPT4, and OPT5, which were found to be induced during nitrogen limiting conditions and shown to transport different oligopeptides of different sequence and sizes, as well as two other hypothetical ORFs, OPT6 and OPT7 (16). The OPT1 of C. albicans, which shows maximum similarity and identity to HGT1 (25), did not display significant capability to transport glutathione during heterologous complementation in S. cerevisiae (13). To evaluate the contribution of OPT1 and the combined contribution of OPT1 to OPT5 in glutathione uptake in its native host, we created met15⌬ knockouts in previously deleted opt1⌬ and opt1-5⌬ mutant strains, the latter being deleted for five OPTs (16). We observed that both opt1⌬met15⌬ and opt1-5⌬ met15⌬ strains showed a very marginal (almost negligible) growth defect on glutathione as compared with met15⌬ strains grown on glutathione (Fig. 2). This indicated that OPT1, as observed previously by heterologous complementation in S. cerevisiae met15⌬hgt1⌬, was not contributing to glutathione transport in C. albicans in any significant manner. Furthermore, none of the other four OPTs, OPT2-OPT5, were contributing to glutathione transport either. We corroborated these findings biochemically by carrying out glutathione uptake assays in different strains carrying the different OPT genes under control of the ADH1 promoter that had been integrated into an opt123⌬ mutant strain (16). We compared the accumulation of radiolabeled glutathione ([ 35 S]GSH) in the opt12345⌬ delete strain with the OPT1-OPT5 genes expressed under constitutive ADH1 promoter. Although OPT1, OPT2, and OPT5 led to a slight accumulation of radiolabeled glutathione as compared with the deletion strain opt12345⌬, the contributions did not appear to be significant (data not shown). These observations further confirm that OPT1 to OPT5 family members of C. albicans do not function as glutathione transporters of the cell.
The DUG Pathway Is Essential for Glutathione Degradation in C. albicans-The ability of C. albicans met15⌬ to grow on glutathione as a sulfur source in the absence of involvement of the OPTs OPT1-OPT5 was surprising. The immediate question was whether glutathione needed to be transported inside to be utilized or whether it was being degraded extracellularly following the transport of the constituents inside. We addressed this issue by arguing that if glutathione was being transported inside, then it would most likely involve the alternative degradation pathway, the DUG pathway that required the DUG1/DUG2/DUG3 proteins. In S. cerevisiae exogenous utilization of glutathione as a sulfur source has been shown to be mediated by the alternative pathway of glutathione degradation (DUG pathway) (14).
Putative orthologues of the Dug proteins (Dug1p, Dug2p, and Dug3p) were identified in C. albicans, and to investigate the role of the Dug pathway in glutathione degradation in C. albicans, the DUG3 gene (19.5159) was deleted in a met15⌬ background. This strain was severely compromised in its ability to utilize exogenous glutathione in C. albicans, clearly demonstrating that the Dug pathway is critical for utilization of extracellular glutathione in C. albicans (Fig. 3).
Isolation of OPT7 as a Glutathione Transporter in C. albicans-The absence of involvement of OPT1-OPT5 with the requirement for the DUG pathway suggested the presence of an unidentified transporter enabling efficient glutathione transport in C. albicans. To identify a possible transporter enabling growth on glutathione, we transformed a S. cerevisiae met15⌬hgt1⌬ strain that lacked the ability to grow on glutathione as a sulfur source with a C. albicans genomic library made in an S. cerevisiae expression vector Yep13. The transformants were selected on glutathione-containing plates (100 and 150 M), and seven clones appeared on these plates after 4 days. Four of these were confirmed as genuinely complementing the glutathione transport defect in S. cerevisiae met15⌬hgt1⌬ strains. Plasmids were isolated and retransformed into S. cerevisiae met15⌬hgt1⌬ strains, and after reconfirmation of complementation, the inserts were sequenced at the ends. All four clones contained one common ORF, OPT7 (ORF 19.5673). Sequencing of the clone revealed that it had four polymorphic  albicans. The C. albicans WT, met15⌬, and met15⌬-dug3⌬ strains were grown to exponential phase in minimal medium containing methionine, harvested, washed, and resuspended to an A 600 ϭ 0.2. Serial dilutions were spotted on minimal ammonium chloride medium containing methionine and glutathione as described under "Experimental Procedures" and in the legend to Fig. 1. DECEMBER 2, 2011 • VOLUME 286 • NUMBER 48 changes that differed from the sequence deposited in the Candida Genome Database, but corresponded to polymorphic changes in an allele of this ORF reported earlier (GenBank TM accession number DQ372680). To confirm whether the OPT7 ORF was indeed responsible for complementing the defect of glutathione transport in S. cerevisiae met15⌬hgt1⌬, we amplified this ORF from C. albicans genomic DNA by PCR, cloned it downstream of the TEF promoter, and checked for complementation in S. cerevisiae. As shown in Fig. 4A, TEF-OPT7 complemented the glutathione transport defect of the S. cerevisiae met15⌬hgt1⌬ strain at low glutathione concentration.

Glutathione Transport and Utilization in Candida albicans
To gain more direct information on the role of the OPT7 encoded protein in glutathione uptake, we measured the accumulation of [ 35 S]GSH in S. cerevisiae met15⌬hgt1⌬ cells. OPT7 expressed from the TEF promoter led to a 2-3-fold higher accumulation of [ 35 S]GSH in cells as compared with the vector control (Fig. 4C). These observations further demonstrated the ability of OPT7 to transport glutathione.
To evaluate the involvement of OPT7 in GSH utilization in its native host C. albicans, we deleted OPT7 in a met15⌬ background and examined the ability of the C. albicans met15⌬opt7⌬ to utilize glutathione as a sulfur source. The C. albicans met15⌬ and met15⌬opt7⌬ grew comparably on methionine as a sulfur source. However, a clear growth defect was observed for the met15⌬opt7⌬ mutant at low concentrations of glutathione (20, 50, 100, and 200 M), as compared with the met15⌬ (Fig. 4B). At higher glutathione concentrations, both strains could grow and could be a consequence of multiple OPTs contributing marginally to the transport that masked the phenotype at the higher glutathione concentrations. These results confirmed that OPT7 is the major glutathione transporter in C. albicans.
The ability of OPT7 to take up glutathione from the extracellular environment also suggested that it was a plasma membrane transporter. To confirm this, we tagged OPT7 with a nine-amino acid HA tag at the C terminus and examined the localization in S. cerevisiae. The tagged construct was functional as seen from the complementation in the S. cerevisiae met15⌬hgt1⌬ (data not shown). Indirect immunofluorescence in using an anti-HA monoclonal antibody followed by the Alexa Fluor 488-conjugated secondary antibody images captured by confocal microscopy revealed a prominent signal on the plasma membrane with few intracellular patches (supplemental Fig.  S1). These studies suggested that OPT7 localizes primarily to the plasma membrane. Plasmid bearing OPT7, OPT1, and HGT1 under the TEF promoter and the corresponding vector p416TEF were transformed into met15⌬hgt1⌬ (ABC 817), were grown to exponential phase in minimal medium containing methionine, harvested, washed, resuspended in water, and serially diluted to give A 600 . Dilutions were spotted on minimal medium containing methionine and glutathione as described under "Experimental Procedures." B, disruption of OPT7 in C. albicans met15⌬ background results in growth defect at glutathione. The met15⌬ and met15⌬-opt7 strains were grown to exponential phase in minimal medium containing methionine, harvested, washed, and resuspended to an A 600 ϭ 0.2. Serial dilutions were spotted on minimal ammonium chloride medium containing glutathione as described under "Experimental Procedures." C, OPT7 expressed in S. cerevisiae met15⌬ hgt1⌬ strain led to uptake of [ 35 S]glutathione. met15⌬hgt1⌬ strain of S. cerevisiae was transformed with TEF-OPT7, TEF-OPT1, and TEF-HGT1 and with the corresponding vector control p416TEF. Transformants were grown in SD medium containing 200 M methionine, and log phase grown cells were then harvested and used for the transport experiment as described under "Experimental Procedures." The data shown are the means of one experiment done in duplicate (each of the experiment repeated twice).
OPT7 has a single CUG codon at position 575 predicted to code for serine in C. albicans and leucine in S. cerevisiae. We also mutated this codon so that it codes for serine in S. cerevisiae by site-directed mutagenesis. The Leu to Ser clone was functional and complemented the S. cerevisiae met15⌬hgt1⌬ similar to the clone bearing the CUG codon with leucine at the position (data not shown).
Biochemical Characterization of OPT7 Reveals That It Is a Glutathione Transporter, with the Ability to Transport Tripeptides-To biochemically characterize OPT7, we exploited the S. cerevisiae met15⌬hgt1⌬ strain as the host strain (11), transforming it with TEF-OPT7 (and the corresponding vector) and thus eliminating any interference of other C. albicans OPTs contributing toward glutathione transport.
To determine the kinetic parameters of OPT7, the initial rate of glutathione transport of the transformants was measured over a range of glutathione concentrations. After deduction of the background glutathione uptake activity (associated with the p416TEF strain), a Lineweaver-Burke plot was obtained. The plot was linear (Fig. 5A), and determination of the kinetic parameters revealed a K m of 205.5 Ϯ 30.5 M and V max of 18.2 Ϯ 4.9 nmol of glutathione⅐mg⅐protein Ϫ1 min Ϫ1 . We also determined the K m without changing CUG codon (Leu to Ser) and obtained a value close but slightly higher at 314.5 Ϯ 12.4 M.
To examine the substrate specificity of the transporter, competitive transport studies were undertaken in which the initial rate of glutathione uptake was measured in the presence of a 10-fold excess of unlabeled competing ligand. The effect of a dipeptide, tripeptide, GSH, and GSSG was evaluated on the glutathione uptake by OPT7. Among the various compounds used, unlabeled glutathione was found to be the most effective competitor, blocking nearly 60% uptake with 10-fold and 70% with 20-fold excess, followed by GSSG and tripeptide (Gly-Gly-Gly) (Fig. 5B). The ability of GSSG to inhibit uptake mediated by OPT7 was similar to what has been observed with other high affinity glutathione transporters (11,12) and is a reflection of the ability of the transporters to also transport glutathione conjugates. However, the significant inhibition by the tripeptide suggested that although OPT7 was principally a glutathione transporter, it might also be able to transport tripeptides as well. This possibility was examined by a growth assay in which we evaluated the ability of the S. cerevisiae met15⌬ hgt1⌬ strain (transformed with p416TEF-OPT7) to grow on the cysteinecontaining tripeptide Glu-Cys-Gly and the methionine-containing tetrapeptide Leu-Tyr-Met-Arg in addition to GSSG as a source of sulfur for growth. We observed that Glu-Cys-Gly and GSSG substrates were efficiently used as a sulfur source by OPT7, but not by OPT1. The latter transformants were not able to grow on these peptides. In contrast, although we did not observe any growth of the OPT7-transformed cells on the tetrapeptide Leu-Trp-Met-Arg, OPT1 grew efficiently on this tetrapeptide as expected from previous observations on OPT1 (16). The observations with OPT7 are also in contrast to the S. cerevisiae glutathione transporter, Hgt1p, which did not show any growth on Glu-Cys-Gly, unlike OPT7 (Fig. 5C). Put together, these observations, along with the inhibition studies, suggest that OPT7 mediates uptake of tripeptides in addition to glutathione and its conjugates and thus, despite being a glutathione transporter, has a broader specificity.
The pH dependence studies indicated that GSH uptake mediated by OPT7 was optimal at pH 5.0, and strongly decreases at higher pH and lower pH (data not shown). Preincubation with the metabolic inhibitors, such as sodium azide and carbonyl cyanide p-chlorophenylhydrazone, resulted in a considerable loss in glutathione uptake (Fig. 5B), suggesting that OPT7-mediated glutathione uptake is likely to be an energy-dependent process.
Glutathione Transport by OPT7 Is Significantly Repressed When Cells Are Pregrown in Methionine-Glutathione transporters of S. cerevisiae and S. pombe have been shown to be regulated by the sulfur regulatory network, a further demonstration that their true function is in glutathione transport. Considering the ability of OPT7 to take up tripeptides as well, it was important to obtain information on its regulation so as to throw light on its true physiological function.
Genome wide expression profiling studies have revealed that OPT7 is induced 5.6-fold by cadmium (26). Cadmium causes depletion of glutathione levels and has been shown to cause sulfur pathway derepression in S. cerevisiae (27). To investigate whether OPT7 regulation is under sulfur limitation, glutathione uptake was measured in both met15⌬ and met15⌬opt7⌬ strains after growing it in sulfur-starved and sulfur-rich medium. In a met15⌬ strain, uptake of GSH was maximum in nonrepressing sulfur conditions and considerably repressed in the presence of sulfur-rich medium. Moreover, the fold induction in the OPT7 expression between the poor methionine and the rich methionine medium was 3-fold, whereas it was just 1.5-fold in the met15⌬opt7⌬ background (Fig. 6A). These studies suggest that OPT7 is under sulfur regulation. Interestingly, when we examined the promoter of OPT7, we observed a CBF1 binding motif, TCACGTG, at position Ϫ145. Importantly, this motif was not present in any of the other C. albicans OPTs but was interestingly also observed in the DUG3 promoter at position Ϫ166. CBF1 disruptions in C. albicans have been shown to lead to cysteine and methionine auxotrophy (28), and a regulatory role in sulfur metabolism is likely to be played by this protein in C. albicans, similar to the role of the orthologous protein in S. cerevisiae.
To investigate a possible role for CBF1 in the glutathione utilization pathway, we initially examined the growth of C. albicans cbf1⌬ in different sulfur sources. We observed that the strain could grow on cysteine and methionine (as previously reported (28)), but it does not grow on glutathione (Fig. 6B), clearly implicating a role in glutathione utilization. We further measured the transport capabilities of glutathione uptake in the cbf1⌬, met15⌬, and met15⌬opt7⌬ strains after growing it in sulfur-starved medium. Significantly reduced glutathione transport activity was observed in the cbf1⌬ strain of C. albicans and was comparable with the transport in the met15⌬opt7⌬ (Fig. 6C). These results strongly suggest that the sulfur regulatory factor CBF1 is essential for expression of OPT7 in glutathione utilization.
Glutathione Transport and Glutathione Utilization Are Not Essential for Virulence of C. albicans in a Mouse Model of Systemic Candidiasis-To determine whether the ability to transport and utilize exogenous glutathione was essential for the survival of C. albicans in vivo, we examined the virulence of C. albicans strains deleted either in OPT7 or DUG3 for their survival in vivo in a systemic murine model of candidiasis in BALB/c mice. The mice were infected intravenously with an inoculum of yeast (5 ϫ 10 5 cells), and the survival of mice was monitored on a daily basis. The majority of mice infected with wild-type C. albicans could not survive beyond 11 days, with death onset seen as early as day 4. All of the mice infected with the C. albicans opt7⌬ and dug3⌬ strains behave similar to the wild-type strain (Fig. 7). These results revealed that glutathione transport and utilization are not required for virulence in C. albicans.
To examine whether the OPT7 gene was expressed in vivo, we employed the IVET developed for C. albicans (29,30). Based on this technique, the OPT7 promoter was fused to FLP recombinase, and expression of the FLP recombinase would lead to eviction of a mycophelonic resistance cassette flanked by FRT sites that was integrated into the same strain (29). The expression of OPT7 in vivo would lead to expression of FLP recombinase and eviction of the MPA resistance cassette that would be seen as appearance of small colonies (sensitive colonies) on MPA plates. In contrast, a lack of induction would lead to a lack of FLP recombinase expression and consequently the appearance of only large colonies reflecting the MPA resistance cassette being intact.
As controls, we used strains carrying the SAP5 fusions to FLP recombinase described earlier (29). These are known to be expressed in vivo. The mice were injected with each of these fusion constructs along with control strains S2UF1A (which contains the mycophenolic acid cassette and gives rise to large sized colonies on MPA-containing plates) and S2F1 s (which lacks the mycophenolic acid cassette and gives rise to small sized colonies on MPA-containing plates). Two mice were sac-rificed at 2, 4, and 6 days, and the right and left kidneys were separately homogenized; aliquots of the extract were plated on YPD medium containing mycophenolic acid (1 mg/ml). We observed that in the case of SAP5, we could observe numerous small colonies appearing at different time points, indicating in vivo expression even after 2 days post-infection. In contrast, in the case of OPT7 only large-sized colonies, and no small colonies were observed even after 6 days post-infection ( Table 2). The results suggested that OPT7 was not being induced in vivo in the infection model used in this study.

DISCUSSION
The work described in this study demonstrates the presence of an efficient transport and degradation pathway for glutathione in C. albicans. The pathway involves OPT7, a member of the oligopeptide transporter family, and DUG3, a component of the DUG pathway of glutathione degradation (14). The K m of OPT7 for glutathione was 205 M and reflects a reasonably high affinity glutathione transporter. The study seems to suggest that the essentiality of glutathione biosynthesis for the survival of this yeast in vivo may not be a consequence of an inefficient pathway of glutathione utilization. A possible explanation is that it is a consequence of the demand of high glutathione levels for the yeast pathogen in vivo that cannot be met from external sources. FIGURE 6. A, expression of OPT7 is under sulfur regulation. Strains met15⌬ and met15⌬opt7⌬ were grown in minimal ammonia medium containing 20 M methionine (nonrepressive sulfur) and 1 mM methionine (repressive sulfur), and the initial rate of [ 35 S]glutathione uptake was calculated by measuring the radiolabeled glutathione accumulated in the cells at 30-and 180-s time intervals. B, C. albicans cbf1⌬ can use cysteine and methionine as sulfur sources but fail to grow on glutathione. The C. albicans WT and cbf1⌬ strains were grown to exponential phase in minimal medium containing methionine, harvested, washed, resuspended in water, and serially diluted to give 0.2, 0.02, 0.002, and 0.0002 A 600 of cells. 10 l of these dilutions were spotted on minimal ammonium chloride medium containing methionine, cysteine, and glutathione. The photographs were taken after 4 days of incubation at 30°C. C, C. albicans cbf1⌬ strain defective in glutathione uptake. C. albicans met15⌬, met15⌬opt7⌬, and cbf1⌬ were grown in minimal ammonia medium containing 40 M methionine (nonrepressive sulfur). The initial rate of [ 35 S]glutathione uptake was calculated by measuring the radiolabeled glutathione accumulated in the cells at 30-and 180-s time intervals.
The observation that among the seven OPTs (of the peptide transporter clade of the OPT superfamily), it was OPT7, the most remote OPT to the known glutathione transporters (HGT1 of S. cerevisiae and PGT1 of S. pombe), that was the glutathione transporter in C. albicans was completely unexpected and intriguing. OPT7 is unusual not only in its remoteness to HGT1/PGT1 but in that it belongs to a different cluster that lacks the conserved disulfide bonding cysteine pairs (21) seen in the larger OPT family, including other C. albicans OPTs, OPT1-OPT6. Because disulfide-bonding cysteine residues are more conserved even than tryptophan and are rarely lost in evolution, it would be interesting to know the basis of this peculiar evolution. Previous studies on the oligopeptide transporter superfamily have demonstrated that in the PT clade almost all members contain two conserved cysteines at positions corresponding to Cys-622 and Cys-632 of Hgt1p and were indispensable for trafficking to the cell surface and its functionality (21). The only members of the PT clade that did not contain these two cysteines were a small cluster of members that  . albicans opt7⌬ (A), dug3⌬ (B), and WT in mice. Survival (%) for mice injected with C. albicans is shown. Ten BALB¢ male mice were injected with 5 ϫ 10 5 cells of opt7⌬, dug3⌬, and WT.

TABLE 2 OPT7 expression in vivo by IVET
The mice were injected with 5 ϫ 10 5 cells/mice of C. albicans with the strains S2U11A,S2F11A S , S5F12A, and OPT7FL1AB. The mice were sacrificed at 2, 4, and 6 days post-infection, the right and left kidneys were isolated and crushed in tissue homogenizer, and different dilutions were plated on indicator plates containing 1 mg⅐ml Ϫ1 MPA. A percentage of small colonies (identified in expression) are shown (two mice/time point); however, for the S5FI2A strain, at 4 days, only one mouse was sacrificed.  (21). Multiple sequence alignment of all the OPTs, from OPT1 to OPT7 of C. albicans, revealed that in these seven members also, it was only OPT7 that did not contain these two conserved cysteines (data not shown). A phylogenetic tree of all the Candida OPTs along with the known glutathione transporters ( Fig. 8 and supplemental Fig. S2 and Table S2) reveals that OPT7 transporters form a distinct cluster that was clearly quite remote from the others. It is interesting that although C. albicans, Candida tropicalis, and Candida lusitaniae have OPT7 transporters, C. lusitaniae also has a member in the HGT1 cluster. Because the C. lusitaniae member in the HGT1 cluster also contains the conserved Phe and Gln residues in transmembrane domain 9, it is likely to be a glutathione transporter (13,31), and it would be interesting to see whether the C. lusitaniae OPT7 is also a glutathione transporter. The phylogenetic analysis also allows us to cluster the different OPTs, and it is interesting to observe a large number of members of OPT4 and OPT5 among C. tropicalis. Another unusual feature of OPT7 as a glutathione transporter is that unlike the high affinity glutathione transporters of S. cerevisiae and S. pombe, which only transport glutathione (and glutathione conjugates), OPT7 also displays the ability to transport tripeptides as shown above. Interestingly, whereas orthologues of the glutathione transporters (HGT1 of S. cerevisiae/PGT1 of S. pombe) are present in almost all organisms in the fungal/plant kingdom, OPT7 orthologues were clearly absent in S. cerevisiae or S. pombe and seemed to be more prevalent in pathogenic organisms, being found in all the Candida spp. (barring C. glabrata, which lacks any members of the family), and also fungi belonging to the Aspergilli family.

Strains
The observation that neither OPT7 nor DUG3 was required for virulence was interesting. External glutathione can meet two requirements of the yeast pathogen. The first is the requirement of glutathione as a redox compound and in iron metabolism. The second use of glutathione is as a sulfur source. The lack of requirement of DUG3 clearly reflects that glutathione is not an important sulfur source in vivo (despite its abundance). The lack of requirement of OPT7 reinforces this aspect but additionally suggests that it cannot meet the high glutathione requirements per se.
Although OPT7 was regulated by sulfur sources, we could not observe any expression in vivo using the IVET approach. This observation is consistent with the observation that OPT7 disruptants had no effect on virulence in the mouse model of systemic candidiasis and may also explain the need for glutathione biosynthesis in vivo. However, the observation also suggests that the cells were not experiencing a sulfur starvation in vivo. The other sulfur-regulated promoter that has been extensively used in C. albicans research is the MET3 promoter. This  Table S1).
promoter is derepressed many-fold in the absence of both cysteine and methionine. In vivo studies carried out with gene fused to this promoter have also revealed that the promoter is not induced to any significant extent in vivo (32,33). This suggests that the yeasts do not face a sulfur starvation in vivo, at least in the mouse model of systemic candidiasis. The lack of OPT7 expression in vivo (or at least below detection levels of the assay) is in keeping with these results, but considering the higher prevalence of OPT7 among pathogenic yeasts, it will be interesting to see whether in other models of infection, similar or different observations would be observed.