Aft2, a Novel Transcription Regulator, Is Required for Iron Metabolism, Oxidative Stress, Surface Adhesion and Hyphal Development in Candida albicans

Morphological transition and iron metabolism are closely relevant to Candida albicans pathogenicity and virulence. In our previous study, we demonstrated that C. albicans Aft2 plays an important role in ferric reductase activity and virulence. Here, we further explored the roles of C. albicans Aft2 in numerous cellular processes. We found that C. albicans Aft2 exhibited an important role in iron metabolism through bi-directional regulation effects on iron-regulon expression. Deletion of AFT2 reduced cellular iron accumulation under iron-deficient conditions. Furthermore, both reactive oxygen species (ROS) generation and superoxide dismutase (SOD) activity were remarkably increased in the aft2Δ/Δ mutant, which were thought to be responsible for the defective responses to oxidative stress. However, we found that over-expression of C. albicans AFT2 under the regulation of the strong PGK1 promoter could not effectively rescue Saccharomyces cerevisiae aft1Δ mutant defects in some cellular processes, such as cell-wall assembly, ion homeostasis and alkaline resistance, suggesting a possibility that C. albicans Aft2 weakened its functional role of regulating some cellular metabolism during the evolutionary process. Interestingly, deletion of AFT2 in C. albicans increased cell surface hydrophobicity, cell flocculation and the ability of adhesion to polystyrene surfaces. In addition, our results also revealed that C. albicans Aft2 played a dual role in regulating hypha-specific genes under solid and liquid hyphal inducing conditions. Deletion of AFT2 caused an impaired invasive growth in solid medium, but an increased filamentous aggregation and growth in liquid conditions. Moreover, iron deficiency and environmental cues induced nuclear import of Aft2, providing additional evidence for the roles of Aft2 in transcriptional regulation.


Introduction
Candida albicans, a common opportunistic human fungal pathogen, can cause superficial mucosal infection as well as lifethreatening system diseases in immunocompromised individuals, such as organ transplant recipients, cancer patients and people with HIV/AIDS. In the last few years, C. albicans infections occur more frequently with high mortality rates, and are considered as the major sources causing hospital-acquired fungal diseases. Therefore, a better understanding of C. albicans pathogenicity will be beneficial to the identification of new antifungal targets and the treatment of C. albicans infections. Numerous studies have reported that the pathogenicity of C. albicans is relevant to its feature of morphological transition, oxidative stress, as well as iron acquisition and metabolism [1,2].
Iron is an essential nutrient, which is required for the growth and metabolism in most organisms, including the budding yeast Saccharomyces cerevisiae and human fungal pathogen C. albicans [3,4]. Iron cofactors, such as heme and iron-sulfur clusters, are implicated in many major cellular processes, including the tricarboxylic acid cycle, chromatin remodeling and metabolite biosynthesis [4,5]. However, excess iron is potentially toxic because of the formation of highly toxic radicals by the Fenton reaction [3,6,7]. Therefore, it poses a ubiquitous challenge for the organisms to thrive in iron fluctuating environments. It has been widely reported that S. cerevisiae is able to employ transcriptional and metabolic remodeling in response to iron fluctuations [4,8,9,10]. The Aft1-Aft2 dependent regulation plays a central role in maintaining iron homeostasis [9,11,12]. Aft1 and Aft2 are functionally similar, and have partially overlapping roles in the control of iron-regulated pathways [13,14]. In iron deficiency, Aft1 activates numerous iron-regulon genes involved in iron binding/ acquisition from the environment, the mobilization of intracellular iron stores, and the metabolic adjustments from iron-dependent to iron-independent pathways. Aft2 regulates the transcription of genes involved in intracellular iron homeostasis in the absence of Aft1, including the vacuolar iron transporter SMF3 and the mitochondrial iron transporter MRS4. In addition, previous studies suggested that Aft1 shuttles from cytoplasm into nucleus under iron-deficient conditions, which is crucial for the Aft1 regulation in response to iron availability [15,16]. In the past several years, numerous attempts have been made to elaborate iron homeostasis regulation in C. albicans [17,18,19,20]. To acquire iron from the host environment, C. albicans has evolved at least three independent systems, including reductive iron uptake pathway, siderophore-iron uptake pathway and hemoglobin-iron uptake pathway. Moreover, Chen et al dissected an iron homeostasis regulatory circuit among transcriptional activator Sef1 and two other transcriptional repressors (Sfu1 and Hap43), providing a novel insight into the molecular mechanisms of iron metabolism [18]. However, little is known about the role of C. albicans Aft-type transcription factor in iron metabolism.
Functional genomics analyses and phenotypic screening experiments reveal that fungal Aft-type transcription factor is implicated in the diverse range of cellular metabolism in S. cerevisiae, including iron homeostasis, the RIM101 pH pathway, cell-wall stability, DNA damage, chromosome stability, and mitochondrial function [21]. Notably, the roles of Aft1 in some cellular processes are mediated through an iron-independent mechanism. In addition, a previous report suggested that Aft1 and Aft2 have redundant roles in response to oxidative stress [13]. The aft1D mutant, particularly the aft1Daft2D double mutant, shows hypersensitivity to hydrogen peroxide (H 2 O 2 ). Further research revealed that oxidative stress is implicated in the stability of Aft1 regulon mRNAs, and causes metabolic adjustment from the reductive to the non-reductive iron uptake pathways to minimize oxidative damage by the ferrous ions [22].
C. albicans has the ability to undergo reversible morphogenetic transitions between budding yeast form and filamentous form. The reversible transition is important for colonization, survival and the establishment of infections in the hostile environment, which is closely associated with pathogenesis and virulence [2,23]. Multiple environmental sensing and signal transduction pathways involved in morphogenesis and pathogenesis have been extensively characterized in C. albicans [24], including Cph1-mediated MAPK pathway, Efg1-mediated cAMP pathway, Rim101-mediated pH response pathway and Czf1-mediated matrix embedding pathway. Some negative regulation pathways are mainly mediated by Tup1 repression with Nrg1, Mig1, and Rfg1 [25,26]. Cell-cell adhesion (flocculation) and cell-surface hydrophobic interactions are indicative of adhesion properties, which are important for the penetration into internal organs/tissues and establishment of indigenous microbial flora. Many yeast cells, including C. albicans, possess the ability to adhere to abiotic surfaces, other cells and host tissues [27,28,29]. Adherence is often found to be highly regulated by many transcription factors, such as Bcr1, Als1 and Zap1 [30]. Several signaling cascades are shown to control the adhesion regulation, including the Ras-cAMP-PKA and MAPK-dependent pathways [27].
Our previous study identified a C. albicans Aft-type functional homologue C. albicans Aft2, an ortholog of S. cerevisiae Aft1/Aft2 regulators, and demonstrated its important role in ferric reductase activity and virulence [31]. In this study, we further elucidated the mechanism by which C. albicans Aft2 regulated iron acquisition and utilization. Our results suggested that C. albicans Aft2 functioned as both a positive and a negative transcription factor in the regulation of different iron-responsive genes. In addition, we also found that the aft2D/D mutant exhibited hypersensitivity to oxidative stress. Deletion of AFT2 in C. albicans increased adherence ability to polystyrene, cell surface hydrophobicity and flocculation. Here, we firstly demonstrated that C. albicans Aft2 functioned as a transcription repressor in morphogenesis through regulating the expression of hypha-specific genes in liquid inducing conditions. Furthermore, we provided the direct evidence that iron deficiency and environmental cues induced nuclear localization of Aft2, which was usually a prerequisite for transcriptional control.

Strains and Growth Conditions
All strains used in this study are listed in Table 1. Strains were routinely cultivated in YPD medium (1% yeast extract, 2% peptone, 2% glucose) supplemented with 80 mg/ml uridine or synthetic drop-out medium (0.67% yeast nitrogen base without amino acids, 2% glucose, 0.2% complete amino acid mixture lacking specific amino acids). For dot assay experiments under various conditions, YPD medium containing calcofluor white (CFW), sodium dodecyl sulfate (SDS), Na + , Zn 2+ , Cd 2+ and Co 2+ was achieved, respectively. YPD medium supplemented with 10% fetal bovine serum and Spider medium were used for hyphal induction. 100 mM bathophenanthroline disulfonate (BPS, Sigma) was added to YPD or M199 medium (Gibco) to achieve irondeficient conditions. For oxidative stress assay, hydrogen peroxide (H 2 O 2 ) at the indicated concentrations was added to YPD medium.

Strains and Plasmids Construction
Primers used in this study are listed in Table 2. The pNKSit1 promoter fusion, which included the native CaSIT1 promoter and LacZ reporter, was generated as described previously [32]. Briefly, a ,1 kb of SIT1 promoter was amplified from the BWP17 genome (primers 1/2), and ligated into pGEM-Teasy vector (Promega) to generate pGEM-T-P SIT1 . Then, a ,3.5 kb AseI-MluI fragment containing the LacZ gene digested from pDDB211 plasmid was cloned into the NdeI-MluI sites of pGEM-T-P SIT1 to yield pGEM-P SIT1 -LacZ. Both a ,5.5 kb DrdI fragment digested from pGEM-P SIT1 -LacZ and a ,7.3 kb NotI-EcoRI product containing TRP1 marker digested from pDDB78 plasmid were transformed into a trpD S. cerevisiae by in vivo recombination to generate pNKSit1 plasmid, which was recovered into Escherichia coli by electroporation. The pNKFrp1 plasmid was constructed by the same strategy (primers 3/4). The resulting plasmid was digested with NruI and transformed into C. albicans strains for bgalactosidase assays.
A ,3.0 kb BglII/SalI PCR fragment (primers 5/6) was subcloned into the pBES116-P ADH1 plasmid to yield the AFT2 complemented construction under the strong ADH1 promoter. A ,2.4 kb BamHI/XhoI AFT2 ORF product (primers 7/8) was subcloned into pFLAG-GFP to yield the Aft2-GFP fusion plasmid. StuI-linearized fusion plasmid was integrated into the RP10 locus to express the GFP fusion protein under the ACT1 promoter in C. albicans. All the constructs were verified by DNA sequencing. The Aft2-HA fusion protein was achieved by PCR-mediated homologous recombination. The aft2::URA3 cassette containing a flanking homology region was amplified according to the 9/10 primers and the pFA-HA-URA3 template. Then, the AFT2 heterozygous mutant was transformed by the aft2::URA3 cassette to generate correct Aft2-HA recombinants.

RNA Isolation and Quantitative Real-time PCR
Total RNA was extracted by the phenol-chloroform method as previously described [19]. The overall quality of RNA was measured by A260/A280 and analyzed by agarose gel electrophoresis. Quantitative real-time PCR was performed in triplicate and repeated in three independent experiments using an iQ5 Real-Time PCR system (Bio-Rad). Primers  used in quantitative real-time PCR are listed in Table 2. Independent reaction mixtures were carried out by the same cDNA for both the genes of interest and the ACT1 gene using the SYBR Green qPCR SuperMix (TransGen Biotech) according to the instructions. Thermal cycler conditions included initial denaturation at 95uC for 3 min, followed by 40 cycles of denaturation at 95uC for 20 sec, annealing at 58uC for 20 sec and elongation at 72uC for 20 sec. The relative fold changes in gene expression were determined by the 2 2ggCT method [33].

b-galactosidase Assays
Overnight cultures of strains in M199 (pH 4) medium were washed and re-cultivated in 35 ml fresh M199 medium buffered with 150 mM HEPES to pH 4 and pH 8 with or without 100 mM BPS to mid-exponential phase at 30uC. Then, b-galactosidase assays were performed as described previously [32]. The activity was calculated in Miller units according to the following formula: (A420)/(OD 600 6volume assayed6time).

Reactive Oxygen Species (ROS) and Superoxide Dismutase Activity (SOD) Assay
Determination and detection of ROS were performed as described previously [34,35]. The indicated strains were cultivated to mid-exponential phase, and stained with 10 mM 2', 7'dichlorofluorescein-diacetate (DCFH-DA) for 30 min. Then, cells were washed two times with phosphate-buffered saline (PBS) buffer, and incubated for 90 min in YPD medium supplemented with 8 mM H 2 O 2. The fluorescence signals, an indicator of the degree of general oxidative stress, were measured by a Zeiss Axio Imager Z1 fluorescence microscopy. SOD activity was determined by the Riboflavin/NitroBlue Tetrazolium (RF/NBT) assay as described previously with slight modifications [36,37]. Superoxide dismutase catalyses the inhibition of NBT reduction, and the extent of which could be monitored by spectrophotometry. Mid-exponential cells were suffered to 8 mM H 2 O 2 for 90 min, washed for three times and weighted. Crude cell extracts were prepared by breakage with glass beads in 0.1 M ice-cold Na 2 HPO 4 -NaH 2 PO 4 buffer containing protease inhibitors. After centrifugation, culture supernatants were collected. NBT reduction was measured in five reaction mixtures containing 0 ml, 40 ml,80 ml,120 ml and 300 ml cell extract. The 1 ml reaction mixtures contained 50 mM Na 2 HPO 4 -NaH 2 PO 4 buffer, 130 mM methionine, 20 mM riboflavin, 750 mM NBT and 1 mM EDTA. The samples were incubated in the dark for 10 min, and subsequently illuminated by 4 fluorescence tubes (Philips TLD/18 W, 4500 lux) for 20 min at the room temperature. Afterwards, absorbance was measured at 560 nm by spectrophotometry. The sample amount of 50% inhibition was calculated by regression using the linear part of a natural semi-log curve, and regarded as one unit of SOD enzyme activity (U). The SOD activity was expressed as U/(g6h) reagent, and calculated by following formula: (volume of assayed sample61000660)/(volume of 50% inhibition6weight of assayed sample6reaction time).

Adhesion, Flocculation and Cell Surface Hydrophobicity
The crystal violet assay was used to quantify adhesion ability [38]. Briefly, Overnight cultures were washed and resuspended to an optical density of 5610 6 cells/ml in RPMI 1640 medium (buffered to pH 7.4 with 165 mM morpholinepropanesulfonic acid). Suspensions of each strain (500 ml) were added into a separate 24-well polystyrene plate, and incubated statically at 37uC. After 2 h, 4 h and 12 h incubation respectively, each plate was gently washed with PBS buffer for three times to remove any non-adherent cells. The adherent cells were fixed with 1 ml of methanol for 15 min, stained with 1% crystal violet for 2 min, washed gently and submerged with 10% acetic acid for 15 min. The A595 of the supernatant was measured to determine the biomass of each sample. After 12 h incubation, structure of adhered cells was examined by Scanning Electron Microscopy (SEM) as previously described [39]. Samples were visualized at a magnification of 62000 by using a Lecia Cambridge S-360 scanning electron microscope.
Cell flocculation assay was examined as follows: Firstly, C. albicans strains were grown in YPD medium for 24 h at 37uC. Cells were harvested and resuspended in RPMI 1640 medium with OD 600 of 0.1, then cultivated to mid-exponential phase for 12 h at 37uC with shaking. The tubes were vortexed for 30 sec and allowed to settle for 10 min before photographed.
Cell surface hydrophobicity was determined as described in the previous study [40]. Cells were cultivated to mid-exponential phase, washed twice with PBS buffer, and resuspended to an OD 600 of 1.0. For each strain tested, 200 ml aliquots of xylene or hexadecane were added to glass tubes containing 3 ml of cell suspension. After 10 min incubation at 37uC water bath, the mixtures were subjected to a vigorous vortexing. Before the OD 600 of the aqueous phase was measured, the tubes were allowed to incubate for a further 20 min to separate the hydrocarbon from the aqueous phase. The percentage of cells removed from the aqueous phase was expressed as cell surface hydrophobicity. Each strain was tested in triplicate and two independent situations.

Fluorescence Microscopy
To observe the Aft2 subcellular localization in response to environmental cues, wild-type strains producing Aft2-GFP fusion protein were cultivated to mid-exponential phase in YPD, YPD+100 mM BPS, YPD+10% serum and Spider medium, respectively. For GFP microscopy, cells were fixed with 3.7% formaldehyde in the culture medium for 30 min, washed, and stained with 1 mg/ml 4',6-diamidino-2-phenylindole (DAPI, Sigma) for 10 min. Cells were washed for three times, examined and photographed with a Leica TCS SP5 confocal laser scanning microscope.

C. albicans Aft2 Plays an Important Role in Iron Metabolism under Iron-deficient Conditions
Our previous study revealed that deletion of AFT2 affects cell surface ferric reductase activity [31]. To further investigate the role of AFT2 in iron metabolism, cellular iron content was determined by atomic absorption spectroscopy. Our results demonstrated that deletion of AFT2 had no significant effect on cellular iron content under iron-adequate conditions ( Figure S1A). However, in comparison with the wild-type (BWP17) and AFT2 complemented (NKF46) strains, cellular iron levels in the aft2D/D mutant (NKF25) were reduced approximately 50% under irondeficient conditions ( Figure 1A). These results indicated that deletion of AFT2 reduced cellular iron accumulation, suggesting a potential role of AFT2 in iron metabolism. To explore the mechanism by which AFT2 affected iron metabolism, quantitative real-time PCR was performed to examine the expression of ironregulon genes. When grown under iron-adequate conditions, the transcript levels of iron-regulon genes were similar in the wild-type and aft2D/D mutant cells ( Figure S1B). However, C. albicans Aft2 had a dual role in regulation of iron-dependent genes in response to iron deficiency ( Figure 1B). Aft2 acted as a negative regulator to govern the expression of some iron-regulon genes, such as SIT1, MRS4, SMF3 and HAP43. On the other hand, it also functioned as a positive regulator to induce the expression of some other ironrelated genes, such as FRP1, CFL1, FET3, FET34 and FTR1. Taken together, our results indicated that C. albicans Aft2 is closely associated with the regulation of iron-responsive genes under irondeficient conditions.

C. albicans Aft2 is Involved in Iron-responsive Transcriptional Regulation in Different pH and Ironlimited Conditions
To further understand the regulatory mechanism of Aft2 transcription factor, both the P FRP1 -LacZ (pNKFrp1) and P SIT1 -LacZ (pNKSit1) promoter fusions were constructed and transformed into the wild-type and aft2D/D mutant strains, respectively. The expression of LacZ reporter, as an index of promoter activity, was examined in different pH and iron-limited conditions. In aft2D/D cells, the P FRP1 -LacZ expression was respectively reduced , 46% at pH 4 with 100 mM BPS, , 51% at pH 8 and , 65% at pH 8 with 100 mM BPS, compared to the results for wild-type cells (Figure 2A). These results demonstrated that C. albicans Aft2, as a transcription activator, was related to FRP1 induction under both alkaline environment and iron deficiency. On the contrary, the P SIT1 -LacZ expression was induced , 1.8fold at pH 4 with 100 mM BPS, , 2.0-fold at pH 8 and , 2.2-fold at pH 8 with 100 mM BPS, respectively ( Figure 2B). These data suggested that SIT1 repression was also dependent on C. albicans Aft2 under both alkaline environment and iron deficiency. Notably, both genes had low expression under acidic condition (pH 4). In acidic pH condition, there was no significant difference of FRP1 expression between wild-type and aft2D/D cells. However, SIT1 expression still retained , 4.5-fold induction in aft2D/D cells (Figure 2A, B). These results appeared to raise the question that Aft2 might play no or only a minor role in ironresponsive gene expression, especially in acaid pH condition. Although the present data could not sort out direct and indirect iron-regualtory events, these results at least partially suggested that C. albicans Aft2 has an effect on the expression of FRP1 and SIT1 under alkaline environment and iron deficiency.

Deletion of AFT2 Results in Hypersensitivity to Oxidative Stress
Phenotypic analyses have indicated that S. cerevisiae aft1D mutant exhibits several phenotypes related to oxidative stress, including hydrogen peroxide (H 2 O 2 ) hypersensitivity and oxygen-dependent copper toxicity [13]. To explore the function of C. albicans Aft2 in the oxidative stress response, we investigated the sensitivity of the aft2D/D mutant to H 2 O 2 treatment. There was no significant growth difference between the wild-type and aft2D/D mutant strains at low concentrations of H 2 O 2 ( Figure 3A, left panel). However, when exposed to a higher concentration of H 2 O 2 , deletion of AFT2 resulted in a severe growth defect in C. albicans ( Figure 3A, middle panel). The growth defect of the aft2D/D mutant could be rescued by integrating a wild-type AFT2 fragment under the control of ADH1 promoter, and the complemented strain showed similar phenotypes in comparison with the aft2D/ AFT2 heterozygote and wild-type strains ( Figure 3A, middle panel). To explore the possible correlation between Aft2-mediated iron regulation and resistance to oxidative stress, 1 mM exogenous iron was added to the medium. As shown in Figure 3A, exogenous iron could increase growth ability of all strains in response to oxidative stress. However, compared with the phenotype of the wild-type strain, the aft2D/D mutant still showed the hypersensitivity to H 2 O 2 treatment after the addition of exogenous iron ( Figure 3A, right panel). These results suggested that the enhancement of growth is independent of AFT2 status and the growth defect of the aft2D/D mutant might be mediated in an iron-independent fashion.
To expound the nature of growth defects of the aft2D/D mutant in response to H 2 O 2 treatment, we measured the levels of reactive oxygen species (ROS) production and superoxide dismutase (SOD) activity. As compared to the complemented and wild-type strains, the aft2D/D mutant exhibited a dramatically elevated level of SOD activity ( Figure 3B). Deletion of AFT2 highly increased generation and accumulation of ROS. Fluorescence assays suggested that ,72% of the aft2D/D cells harbored high levels of ROS production, whereas ,18% of wild-type cells had increased ROS accumulation ( Figure 3C). These results indicated that C. albicans AFT2 is closely associated with the oxidative stress response.
Regarding the above conclusion that Aft2 had a role in regulation of iron-regulon genes, further research was performed to evaluate the effect of oxidative stress on the expression of ironregulon genes in the aft2D/D mutant. Quantitative real-time PCR analyses revealed that, when exposed to oxidative stress, deletion of AFT2 had no significant effect on the transcript levels of ironregulon genes ( Figure S2). These results demonstrated that the effect of AFT2 deletion on iron-regulon expression might not be due to secondary effects of the sensitivity to oxidative stress.

C. albicans Aft2 does not Effectively Rescue the Non-iron Related Metabolic Defects of the S. cerevisiae aft1D Mutant
In our previous study, we confirmed that C. albicans Aft2 is able to rescue S. cerevisiae aft1D mutant growth defects under ironlimited conditions [31]. To further determine whether C. albicans Aft2 remained the functional diversity of Aft-type regulator in cellular metabolism, phenotypic analyses of the derivative strains were examined. The results indicated that, under the control of either native or S. cerevisiae strong PGK1 promoter, C. albicans Aft2 could not effectively rescue the Scaft1D defective phenotypes in cell-wall integrity (CFW and SDS), metal ion sensitivity (Zn 2+ , Na + , Co 2+ , and Cd 2+ ) and alkaline pH resistance (pH 7.5) (Figure 4). Interestingly, there were no significant growth differences between C. albicans the wild-type (BWP17) and aft2D/ D mutant strains under the same tested conditions, suggesting that C. albicans Aft2 might not exert the essential role in the tested metabolic processes (data not shown).

Aft2 is Required for Invasive Growth in C. albicans
In the previous study, we found that deletion of AFT2 affects colony morphology in solid inducing medium [31]. To further investigate the role of C. albicans Aft2 in hyphal development, we tested the invasive growth. 5 ul of the indicated strains was spotted on solid M199 and Spider medium, respectively. After 5 days incubation at 37uC, the plates were observed before and after wash. Our results suggested that deletion of AFT2 severely impaired invasive growth, almost all the mutant cells on the surface were washed off ( Figure 5A). On the contrary, a great many wild-type and complemented cells still remained on the Figure 1. Effect of AFT2 deletion on iron content and iron-regulon expression under iron-deficient conditions. (A) Cells were pre-grown in YPD medium for 24 h at 30uC, and re-cultivated in 100 ml fresh YPD+100 mM BPS medium for 12 h, respectively. Then, cells were washed twice with sterilized PBS buffer, twice with 1 mM EDTA, and followed by twice with distilled deionized water to remove all exogenous iron. Measurement of cellular iron levels was performed by a Hitachi 180-80 polarized Zeeman atomic absorption spectrometry. Note: ppm = mg/kg. (B) Overnight cultures of the wild-type and aft2D/D mutant strains were cultivated to mid-exponential phase in YPD+100 mM BPS medium, and used for RNA isolation. Quantitative real-time PCR was performed to determine the relative expression changes of iron-responsive genes. Data indicate mean values 6 standard deviations from three independent experiments performed in triplicates. doi:10.1371/journal.pone.0062367.g001 Aft2 Transcription Regulator in Candida albicans PLOS ONE | www.plosone.org surface, suggesting a strong invasive growth. Microscopy-based assay was performed to further determine the degree of invasion ( Figure 5B). The wild-type strain exhibited long filaments surrounded by numerous yeast form cells in the septa, which was similar to the observation in the complemented strain. However, the aft2D/D strain showed multiple branches of short and stunted filaments. To understand the nature of invasive growth defect of the aft2D/D mutant, total RNA was extracted to analyze the transcript levels of hypha-specific genes. Surface proteins, as pro-adhesive and pro-invasive factors, are closely related to the morphogenesis and pathogenicity of C. albicans [24]. HWP1 and HYR1 encode GPI-anchored cell wall proteins, which are required for adherence and phenotype switching [41,42]. ECE1 encodes hypha-specific cell wall-linked component, which is in association with the extent of hyphal cell elongation [43]. ALS3, as a major member of the ALS family, encodes the cell surface glycoprotein. It is known as an essential adhesion and implicated in the adherence to host surfaces and biofilm development [44,45,46]. Our data revealed that deletion of AFT2 reduced the expression of hypha-specific genes in solid inducing conditions, which was consistent with the colony morphology and invasive growth defects in the aft2D/D mutant ( Figure 5C). Taken together, these findings suggested that C. albicans Aft2 is required for invasive growth.

C. albicans Aft2 Functions as a Repressor in Flocculation, Plastic Adhesion, and Surface Hydrophobicity
In view of the above-mentioned result that Aft2 was implicated in agar invasion, we further explored the possible role of Aft2 in surface adhesion properties. Cell flocculation ability was measured after growth in RPMI 1640 medium ( Figure 6A). Deletion of AFT2 resulted in a small but significant increase in flocculation. However, ectopically expression of C. albicans Aft2 suppressed the flocculent phenotype of the aft2D/D mutant. These results indicated that Aft2 acted as a repressor in cell flocculation. In addition, cell surface hydrophobicity was also evaluated by the hydrocarbon partition test. Consonantly, the aft2D/D strain    showed a higher cell surface hydrophobicity in comparison with the results for the wild-type strain ( Figure 6B). The ability of adhesion to polystyrene was calculated through the measurement of total biomass by crystal violent staining in vitro system. After a 2 h, 4 h and 12 h incubation under static condition at 37uC, the total biomass of each sample was examined. Interestingly, the aft2D/D strain had more total biomass in comparison with the wild-type strain, suggesting its stronger adherence ability to polystyrene ( Figure 6C). As expected, adhesion phenotypic defect could be rescued by re-expressing AFT2 fragment under the control of its native promoter. The SEM results also revealed that the number of adhered cells in the aft2D/D mutant was much more than that of the wild-type strain ( Figure 6D). Taken together, although the alternations might be attributed to indirect effects caused by AFT2 deletion, these results still revealed that C. albicans Aft2 potentially is responsible for the flocculation, plastic adhesion, and surface hydrophobicity.

Deletion of AFT2 Enhances Hyphal Development under Liquid Inducing Conditions
To further understand the role of Aft2 in morphogenesis, we examined the phenotypes of the wide-type and aft2D/D mutant strains at 37uC under different liquid hyphal inducing conditions. Interestingly, both in liquid Spider and serum medium, the aft2D/ D mutant exhibited an enhanced filamentous aggregation and growth ability in comparison with the wild-type and AFT2overexpression complemented strains ( Figure 7A, left). To further understand the role of Aft2 in the formation of hyphae, hyphal lengths in the wild-type and mutant strains were measured. Our data suggested that the hyphal length in the aft2D/D mutant was increased compared with the wild-type and complemented strains ( Figure 7A, right). These results supported the conclusion that Aft2 had a role in repressing hyphal development under liquid inducing conditions. To explore the possible mechanism under liquid inducing conditions, quantitative real-time PCR was performed to determine the transcript levels of hypha-specific genes in liquid inducing medium. Consistent with the phenotypes, deletion of AFT2 in C. albicans highly induced the expression of hypha-specific genes, indicating that C. albicans Aft2 changed its transcription activity to a repressor under liquid inducing conditions ( Figure 7B). In addition, C. albicans Aft2 mRNA and protein expression levels were induced under both serum and Spider inducing conditions ( Figure 7C). After incubation for 1 h in YPD medium containing 10% serum, the levels of Aft2 mRNA were approximately 5-fold higher than those in YPD medium (marked as 0 h). When the incubation period was extended to 2 h, Aft2 mRNA abundance remained relatively constant in comparison with those for 1 h incubation. Consistent with mRNA levels, the expression levels of Aft2 protein were also elevated during the yeast-to-hypha transition. Similar results were obtained for incubation in liquid Spider medium. In conclusion, our results for the first time demonstrated that C. albicans Aft2 has a distinct mechanism in hyphal development by altering its transcription activity under different inducing conditions. Iron Deficiency and Environmental Cues Induce Nuclear Import of Aft2 in C. albicans Sequence analyses reveal that C. albicans Aft2 harbors two nuclear localization signals in the N-terminal region [31]. To determine whether Aft2 could shuttle between the nucleus and cytoplasm in response to iron levels and morphogenetic signals, we examined the subcellular localization of Aft2 by confocal microscopy. As a control, C. albicans strains transformed with empty vector showed no GFP fluorescence signal (data not shown). Our results suggested that Aft2 was localized to the whole cytoplasm in normal condition. However, once iron deficiency occurred, Aft2 shuttled from the cytoplasm into the nucleus, and accumulated in the cell nucleus (Figure 8, second panel). These results verified a hypothesis that iron deficiency induced the accumulation of Aft2 transcription factor in the nucleus, in which the activated Aft2 directly or indirectly regulated the expression of iron-regulon genes in response to iron fluctuations. Likewise, various morphogenetic signals, such as serum and nutrient starvation, affected Aft2 cellular localization. Under hyphal inducing conditions, C. albicans Aft2 was also constitutively localized in the nucleus of hyphal cells, indicating the role of Aft2 as a transcription factor in hyphal development (Figure 8, bottom panel). In summary, environmental-induced nucleocytoplasmic shuttling of C. albicans Aft2 further confirmed our conclusion that Aft2 plays an important role in iron metabolism and hyphal development by regulating the expression of specific genes.

Discussion
Evolutional analyses of iron-responsive transcriptional regulators reveal that Zn-finger GATA-type transcription repressors are widespread in ascomycete fungi, while Aft-type transcription activators are generally restricted to the Saccharomycotina fungi [4,18,47,48]. Iron homeostasis is mainly mediated by Aft-type (AFT1/AFT2) positive iron regulatory mechanisms in S. cerevisiae [12]. However, GATA-type iron-responsive transcriptional regulator has not been identified in S. cerevisiae. Interestingly, numerous iron regulatory pathways have been characterized in C. albicans, including Sef1 transcription activator, GATA-type transcription repressor Sfu1 and Hap43-associated CCAAT-binding complex, which are closely associated with the adaptation of iron availability in a hostile environment [17,18]. Our previous study identified and characterized one Aft-type homologous gene CaAFT2, and found its important role in regulating cell ferric reductase activity in C. albicans [31]. Here, we further explored the possible functional roles of C. albicans Aft2 in numerous cellular processes.
Deletion of AFT2 severely decreased whole iron content under iron-deficient conditions, which might be attributed to the reduced expression of iron-regulon genes involved in plasma membrane high-affinity iron uptake system, including the ferrous-specific transport complex composed of a multicopper ferroxidase Fet3 and an iron permease Ftr1. In addition, some members of the FRE metalloreductase family, including FRP1, CFL1 and FRE10, were also down-regulated in the aft2D/D mutant, indicating a positive regulatory role of Aft2 in activating iron uptake system. On the other hand, C. albicans Aft2 could also function as a negative regulator to repress some cellular iron-responsive genes, such as the mitochondrial iron transporter MRS4, the vacuolar iron transporter SMF3, the siderophore transporter SIT1 and ironresponsive regulator HAP43. In addition, our present studies also revealed that C. albicans Aft2 directly regulated MRS4 expression via the conserved core CACCC site [49]. These results suggested that C. albicans Aft2 was closely related to cellular iron mobilization and metabolism. Notably, Hap43 is essential for the growth under iron-limited conditions and virulence of C. albicans [19,20,32]. Hap43 exerts its functional role in response to iron deprivation by a mechanism involving transcriptional repression of iron utilizing genes, resulting in a transcriptional remodeling shift to ironindependent metabolic pathways. In this study, the aft2D/D mutant displayed a ,1.9-fold higher expression level of HAP43 in comparison with the results from wild-type cells, suggesting that C.
albicans Aft2 might partially regulate iron homeostasis via ironresponsive repressor Hap43 under iron-limited environment. Disruption of AFT2 resulted in disturbances of iron homeostasis and followed by the activation of iron-responsive regulator Hap43. This process provoked the rearrangement of metabolism and biosynthesis to decrease the utilization of iron-dependent enzymes/pathways, which was believed to facilitate the survival of C. albicans in iron-deficient environment. The potential relationship between Aft2 and Hap43 will shed a new insight into the understanding of iron-regulatory mechanisms in C. albicans.
The yeast S. cerevisiae Aft-type regulator is involved in its oxidative stress resistance. The aft1D, and particularly aft1Daft2D, mutants show hypersensitivity to hydrogen peroxide [13]. H 2 O 2 has the capacity of causing oxidative damage to cellular components, including DNA, proteins and lipids. Most organisms possess both enzymatic (e.g. superoxide dismutase, catalase, methionine reductase) and non-enzymatic (e.g. glutathione, metallothioneins, glutaredoxin) defense systems to protect cells against oxidative stress [50,51]. In this study, we demonstrated that C. albicans Aft2 was also required for oxidative stress resistance in a concentration-dependent fashion. In addition, we also found that the addition of exogenous iron could not effectively rescue the defective phenotype of the aft2D/D mutant in response to hydrogen peroxide. Further research revealed that the aft2D/D mutant had similar expression levels of iron-regulon genes with the wild-type strain under oxidative stress. Taken together, these data suggested that the role of Aft2 in iron metabolism might be independent of its role in oxidative stress tolerance. Deletion of AFT2 resulted in hypersensitivity to oxidative stress, which was attributed in part to the increased generation of ROS in the aft2D/ D mutant cells. These results also indicated that there was a significant correlation among H 2 O 2 hypersensitivity of the aft2D/ D mutant, the levels of ROS generation and SOD activities. In our model, deletion of AFT2 triggered a dramatic increase in ROS production in the presence of high H 2 O 2 concentrations, and subsequently induced the expression levels of SOD activity. The growth defect of the aft2D/D mutant under H 2 O 2 treatment was caused by increased accumulation of cellular ROS that could not be eliminated efficiently by antioxidant enzymes.
During the long and tedious evolutionary processes, C. albicans acquires Aft-type regulatory function before the whole-genome duplication event (WGD), while the S. cerevisiae lineage obtains two Aft-type regulators after the WGD [47]. Recent studies reveal that S. cerevisiae Aft-type transcription factor Aft1 plays an essential role in remodeling cellular metabolism in addition to iron homeostasis [21,22,52]. We have identified a novel Aft-type factor CaAft2 in C. albicans, and confirmed that CaAft2 is able to rescue S. cerevisiae aft1D mutant growth defects under iron-limited conditions [31]. In this study, our results indicated that C. albicans Aft2 exhibited an important role in iron metabolism through bi-directional regulation effects on iron-regulon expression, suggesting that the Aft-type regulator in C. albicans not only retains but also expands its major conserved function in iron metabolism. However, C. albicans AFT2 ORF fragment under the control of its native promoter, even in the strong S. cerevisiae PGK1 promoter, did not effectively restore the Scaft1D growth defects, suggesting the subdued functional roles of C. albicans Aft2 in metal ion sensitivity, alkaline response and other cellular metabolism. The comparison of Aft-type regulator in many cellular processes between C. albicans and S. cerevisiae provided a potential insight into the evolutional and functional divergence of Aft-type family. In addition, our previous results demonstrated that C. albicans Aft2 regulator evolves an important role in morphogenesis and virulence. The ability of adhesion to other cells and surfaces is considered as a remarkable feature of bacteria to offer protection from hostile environment [27,53,54]. Adhesion-mediated surface phenotypes are usually marked by cellcell interactions (flocculation), cell-substrate recognition and adhesion (including plastic adherence and cell surface hydrophobicity) [27]. Cell surface hydrophobicity is considered as an important factor in the ability of adherence to surfaces, which plays a major role in promoting hydrophobic interactions between the cells and abiotic surfaces, and increasing the resistance against the host immune response [27,55]. Here, we further revealed that C. albicans Aft2 was required for invasive growth, flocculation, plastic adhesion, surface hydrophobicity and hyphal development. The aft2D/D mutant exhibited an increased flocculation, a higher surface hydrophobicity and a stronger adherence to abiotic substrates, while the capacity to invade the agar plates was severely attenuated. The inconsistency between invasion and adhesion probably reflected the ability of adhesive specificity towards the appropriate substrates. In contrast to wild-type cells, the aft2D/D cells formed the aberrant and stunted filaments, and failed to penetrate the agar effectively. Additionally, deletion of AFT2 reduced the expression of hypha-specific genes in solid conditions, including the crucial adhesin ALS3. Therefore, the failure of normal filamentous growth and decrease in ALS3 expression level might be responsible for the attenuated agar invasion.
The ability of the yeast-to-hypha transition is a striking feature of C. albicans, which is highly relevant to its pathogenesis and virulence [56]. Though C. albicans Aft2 played an important role in colony morphology and hyphal development, there might be two distinct mechanisms by which Aft2 participated in the response to various morphogenetic signals. Our results suggested that C. albicans Aft2 functioned as an activator to regulate hypha-specific genes in solid inducing medium, whereas, it switched to a repressor in liquid inducing conditions. Interestingly, C. albicans Aft2 mRNA and protein expression levels were also elevated after the incubation in liquid hyphal inducing conditions, suggesting that Aft2-mediated hyphal repression had a lasting effect during the yeast-to-hypha transition. Aft2-mediated hyphal repression could not abolish hyphal development, which might be due to the compromising effects with multiple positive morphogenetic signaling pathways, such as the major morphogenetic processes mediated by Efg1 regulator [24,57]. In addition, over-expression of AFT2 under the control of ADH1 promoter in aft2D/D mutant background could remarkably reduce filamentous aggregation and hyphal elongation, indicating a potential role of Aft2 in repressing hyphal development. Furthermore, environmental cues, including serum and nutrient starvation, affected cellular localization of C. albicans Aft2 and its accumulation into the nucleus. The nucleocytoplasmic shuttling of C. albicans Aft2 was generally considered as a prerequisite to exert its regulatory activity. Thus, it will be an interesting and exciting challenge for future studies to elucidate the mechanisms by which Aft2 functions as both an activator and a repressor, and define its relative position in the network of signal transduction pathway. Figure S1 Deletion of AFT2 has little impact on iron content and iron-regulon expression under iron-adequate conditions. (A) Overnight cultures of the indicated strains were re-cultivated in 100 ml fresh YPD+200 mM Fe 3+ medium for 12 h, respectively. Cells were harvested and washed distilled deionized water. Cellular iron content was measured by atomic absorption spectrometry. (B) Overnight cultures of the wild-type and aft2D/D mutant strains were cultivated to mid-exponential phase in YPD+200 mM Fe 3+ medium, and used for RNA isolation. Quantitative real-time PCR was performed to determine the relative expression changes of iron-responsive genes. Data indicate mean values 6 standard deviations from three independent experiments performed in triplicates. (TIF) Figure S2 Quantitative real-time PCR analysis of ironregulon expression levels in wild type and aft2D/D Figure 8. Environmental stimuli induce C. albicans Aft2 nuclear accumulation. Overnight cultures of the strain transformed with Aft2-GFP fusion plasmid were regrown to mid-exponential phase in YPD, YPD+100 mM BPS, YPD+10% serum and Spider medium, respectively. Cells were fixed with 3.7% formaldehyde, and then stained with 1 mg/ml 4',6-diamidino-2-phenylindole (DAPI) for 10 min. The subcellular localization of Aft2 was determined by confocal laser scanning microscopy. DIC, Differential interference contrast. doi:10.1371/journal.pone.0062367.g008 mutant cells in response to oxidative stress. Overnight cultures of the wild-type and aft2D/D mutant strains were cultivated to mid-exponential phase. Then cells were incubated for another 90 min in YPD medium supplemented with 8 mM H 2 O 2 before harvesting for RNA isolation. (TIF)