Characterization of the Candida albicans Amino Acid Permease Family: Gap2 Is the Only General Amino Acid Permease and Gap4 Is an S-Adenosylmethionine (SAM) Transporter Required for SAM-Induced Morphogenesis

Candida albicans is a commensal organism that can thrive in many niches in its human host. The environmental conditions at these different niches differ quite a bit, and this fungus must be able to sense these changes and adapt its metabolism to them. Apart from glucose and other sugars, the uptake of amino acids is very important. This is underscored by the fact that the C. albicans genome encodes 6 orthologues of the Saccharomyces. cerevisiae general amino acid permease Gap1 and many other amino acid transporters. In this work, we characterize these six permeases and we show that C. albicans Gap2 is the functional orthologue of ScGap1 and that C. albicans Gap4 is an orthologue of ScSam3, an S-adenosylmethionine (SAM) transporter. Furthermore, we show that Gap4 is required for SAM-induced morphogenesis, an important virulence factor of C. albicans.

healthy population (depending on the method of sample collection and the body site) (1,2). However, C. albicans frequently causes oral and vaginal infections (thrush) if the balance of the normal microflora is disrupted by antibiotic usage or when the immune defenses are compromised, such as in HIV patients or in neonates (3,4). At the most serious level, severely immunocompromised individuals can develop life-threatening systemic infections, such as in the kidney, liver, spleen, or brain (5). Despite the availability of specialized antifungal drugs, such as azoles, polyenes, and echinocandins, mortality rates remain high (6). C. albicans is able to infect nearly every site of the human body and has evolved with quick adaptation to the changing availability of nutrients. C. albicans is able to utilize various fermentable and nonfermentable carbon sources. In the absence of glucose, this fungus can even use fatty acids or amino acids as a source of carbon (7,8). Moreover, a diverse range of nitrogen sources can be used by this fungus. Some of them are used preferentially (ammonia, glutamine, asparagine, and glutamate); however, when these primary nitrogen sources are limited or not available, C. albicans can also use less-preferred nitrogen sources, like isoleucine, tyrosine, and tryptophan (9). In the host, the largest proportion of amino acids is fixed in host proteins, requiring the production of proteolytic enzymes. C. albicans produces several proteases, all belonging to the secreted aspartyl proteinase (SAP) gene family (10). They are independently regulated and functionally distinct and allow the fungus to break down or decompose almost every tissue of the host into suitable nutrients (11,12).
In the model yeast Saccharomyces cerevisiae, amino acid uptake is mediated by a family of 20 amino acid permeases. Most of them are specific for one or a few related L-amino acids. One exception is ScGap1, which has broad substrate specificity, high capacity, and mediates the uptake of most L-and D-amino acids, nonproteic amino acids including citrulline and ornithine, and a number of toxic amino acid analogues (13). The induction of its expression as well as the activity of ScGap1 are tightly regulated according to both the quality and quantity of the nitrogen sources present in the medium. The presence of preferred nitrogen sources (ammonium, glutamate, and glutamine) or of high amino acid levels represses ScGAP1 expression and inactivates the transporter by ubiquitination and internalization for targeting to the vacuole and subsequent degradation (14)(15)(16). The transcription of ScGAP1 is regulated by the nitrogen catabolite repression (NCR) pathway (17). When cells are growing in medium with poor nitrogen sources, e.g., proline or urea, ScGAP1 expression is activated by two GATA-type transcription factors, and neosynthesized ScGap1 accumulates at the plasma membrane in an active, stable form (18).
In addition to its transport function, ScGap1 was found to play a role as an amino acid sensor for the rapid activation of protein kinase A (PKA) upon addition of amino acids to nitrogen-starved cells (19,20). Little is known about transport and signaling properties of Gap permeases in C. albicans or about their regulation. Phylogenetic analysis shows that the C. albicans genome encodes six orthologues of ScGAP1 (21). Based on findings with complementation experiments in S. cerevisiae amino acid transporter mutants, we previously showed that Gap2 is the functional ScGap1 orthologue and that the others function as more specific permeases (21). We also showed that Gap1, Gap2, and Gap6 function not only as amino acid transporters but also, similarly to ScGap1, as amino acid sensors for activation of the PKA pathway. Nevertheless, all these results were obtained by expressing the C. albicans genes in S. cerevisiae mutants lacking their own amino acid transporters.
Apart from ScGap1, another plasma membrane sensor for extracellular amino acids has been described in S. cerevisiae, Ssy1. It is a unique member of the amino acid permease family which is not able to transport amino acids but controls amino acid uptake in exponentially growing cells by inducing the transcription of genes encoding several amino acid permeases, such as ScDip5 and ScAgp1, but not ScGap1 (22)(23)(24). In C. albicans, it was shown that the expression of GAP1 and GAP2 is under control of the amino acid sensor Csy1, which is an orthologue of Ssy1 (25). Whereas this is different from the situation with S. cerevisiae, another level of GAP2 regulation which entails upregulation during nitrogen starvation and repression upon subsequent addition of arginine to the cells is similar to what is observed in S. cerevisiae for ScGAP1 (26).
In this study, we systematically analyzed all six members of the C. albicans general amino acid permease family, and we confirmed Gap2 as the only and real ScGap1 general amino acid permease orthologue, based on its broad substrate specificity and for its similar transcriptional regulation. Moreover, we show that all GAP genes are under control of the Csy1 amino acid sensor, which is different from the situation in S. cerevisiae. In addition to this, we also show that Gap4 is the orthologue of the S. cerevisiae Sam3 transporter, as Gap4 is required for S-adenosylmethionine (SAM) transport and SAM-induced morphogenesis.

RESULTS
The C. albicans genome encodes six ScGap1 orthologues. A BLAST analysis with the ScGap1 protein sequence as a query against the two available C. albicans databases, CGD (27) and CandidaDB (28), resulted in the identification of six putative orthologues (21). A reciprocal BLAST analysis of these six putative Gap protein sequences against the S. cerevisiae SGD database (29) always resulted in the ScGap1 permease as the closest orthologue (21). Phylogenetic analysis of the six C. albicans Gap protein sequences as well as all the other C. albicans amino acid permease protein sequences with all the S. cerevisiae amino acid permease protein sequences showed that all Gap permeases belong to the same cluster as ScGap1, a cluster that also includes ScHip1 (histidine permease) and ScTat2 (tryptophan/tyrosine permease) (Fig. 1). Gap2 constitutes the closest ScGap1 orthologue, while Gap4 is the most distantly related permease in the phylogenetic tree. Using Gap4 as a query, we found ScSam3 was the second best orthologue. ScSam3 is the high-affinity permease that mediates transport of SAM across the plasma membrane of S. cerevisiae (30). While we have characterized all six genes from the GAP gene family, these in silico results prompted us to focus mainly on Gap2 and Gap4. Gene names and systematic names of all the CaGAP genes are listed in Table 1.
To determine whether any of the mutant strains had a defect in the uptake of amino acids, we tested growth for all of them on yeast nitrogen base (YNB) agar plates containing 300 g/ml of one single specific amino acid. As a positive control, we used wild-type strain SN87 that had been made prototrophic for leucine and histidine (strain BSC1). We tested all amino acids that showed a growth phenotype for the S. cerevisiae mutant strains expressing single GAP genes (21). Only the gap2⌬/gap2⌬ strain showed a growth defect on medium containing phenylalanine ( Fig. 2A, left panels) or valine, leucine, or methionine (see Fig. S1 in the supplemental material) as sole nitrogen source. As a general control for growth, we used YNB plus 300 g/ml of ammonium sulfate plates ( Fig. 2A, right panels).
Despite the clear growth defect of the gap2⌬/gap2⌬ strain, growth was not completely abolished on medium with phenylalanine as the N source. This finding may point to the ability of other transporters to take up the tested amino acid. We therefore tested the involvement of Gap1 and Gap6, which we previously showed that, upon expression in S. cerevisiae, are able to take up methionine, leucine, and phenylalanine (21). However, additional deletion of GAP1 or/and GAP6 did not further decrease the capacity to grow on single amino acids as the sole source of nitrogen ( Fig. 2A, left panels for phenylalanine; see Fig. S1 for results with valine, leucine, and methionine).
Similar results were obtained in liquid media (Fig. 2B). Growth assays in the presence of YNB supplemented with specific amino acids showed that Gap2 is required for growth on phenylalanine (Fig. 2B, left panels) and on methionine, valine, leucine,   tyrosine, tryptophan or lysine (Fig. S1) as sole source of nitrogen. On the other hand, Gap2 is not required for growth on medium containing ammonium sulfate (Fig. 2B, right panels), glutamine, arginine, or proline ( Fig. S1) as the sole source of nitrogen. These results showed that Gap2 is the only amino acid permease with very broad substrate specificity. None of the other Gap permeases was able to transport (based on growth assays) such a variety of amino acids (21). Reintegration of GAP2 in the gap2⌬/gap2⌬ strain restored growth on media containing specific amino acids (data not shown), hence linking the phenotype to the absence of Gap2.
(ii) In contrast to S. cerevisiae, in which citrulline is only transported by ScGap1, several amino acid permeases exist in C. albicans that are able to transport citrulline. Citrulline is an amino acid that has been used frequently in studies of S. cerevisiae's transport and signaling capacities via ScGap1. At concentrations lower than 5 mM, citrulline can only be transported appreciably by ScGap1 (13,35). It was previously described that deletion of GAP1 has an effect on citrulline uptake in C. albicans under nitrogen-poor conditions (36). In our hands, however, a single deletion of the GAP1 gene, or single deletions of either GAP3, GAP4, GAP5, or GAP6, had no effect on citrulline uptake under nitrogen-poor conditions, whereas citrulline uptake was significantly lower in the gap2⌬/gap2⌬ strain than in the wild type (P Ͻ 0.001) (Fig. 3A). Citrulline transport was even more reduced in the triple-deletion mutant, gap1⌬/gap1⌬ gap2⌬/gap2⌬ gap6⌬/gap6⌬, than in the single-deletion gap2/gap2⌬ mutant. Since 20% of citrulline transport remained in the triple-deletion strain, other transporters for citrulline must be present in C. albicans.
To confirm the growth experiment results with different amino acids, we also measured the rate of uptake of 2 mM leucine, phenylalanine, and glutamine. Whereas uptake of leucine and phenylalanine showed similar trends as uptake of citrulline (significant differences between the wild-type and gap2⌬/gap2⌬ or gap1⌬/gap1⌬ gap2⌬/gap2⌬ gap6⌬/gap6⌬ strains), none of the tested strains showed differences in glutamine uptake, which was in accordance with the results of the growth experiments ( Fig. S1).
In S. cerevisiae, the general amino acid permease system is strongly repressed under nitrogen-rich conditions (37). To investigate this in C. albicans, we determined citrulline, leucine, phenylalanine, and glutamine uptake in cells grown in the presence of ammonium sulfate. Gap2 was clearly not involved in amino acid uptake under these conditions, pointing to repression under these conditions (see below). Surprisingly, the gap1⌬/gap1⌬ strain and the gap1⌬/gap1⌬ gap2⌬/gap2⌬ gap6⌬/gap6⌬ triple mutant exhibited mildly decreased uptake of citrulline, leucine, and phenylalanine ( Fig. 3B). For glutamine, we could not observe any difference between the BSC1 control strain and any of the tested strains. In summary, the uptake measurements confirmed the results obtained in the growth assays and clearly showed that only Gap2 can be considered a functional orthologue of ScGap1. It is also interesting that the repression of Gap2 transport activity in nitrogen-containing medium almost mimics the effect of a deletion of this gene.
(iii) Expression of all CaGAP genes is under control of the Csy1 amino acid sensor. S. cerevisiae and C. albicans each express an amino acid transporter homologue, Ssy1 and Csy1, respectively, which in both cases lost the capacity to transport. However, in the presence of histidine, they function as sensors to detect amino acids in the medium, resulting in the induction of specific amino acid transporter genes (23,25). The specific transporter genes controlled by this sensor are different between S. cerevisiae and C. albicans. ScGAP1 expression is not regulated by Ssy1. Previous gene expression analysis in C. albicans showed that both GAP1 and GAP2 are under control of Csy1 (25). Here, we confirmed this and showed that the expression of all GAP genes is under control of Csy1 (Fig. 4A). This is one of the main regulatory differences between GAP2 and ScGAP1. We observed slightly different regulation for GAP4 and GAP5 compared to the others, as the fold change difference was bigger after 15 min than after 30 min, whereas this was clearly not the case for the other GAP genes.
(iv) Expression of CaGAP2 is regulated by nitrogen sources, similar to ScGAP1. We performed quantitative real-time PCR (qRT-PCR) to quantify gene expression of the different GAP genes when cells were shifted from medium without any source of nitrogen (nitrogen starvation conditions) to yeast extract-peptose-dextrose (YPD) medium or to YNB medium with ammonium sulfate as nitrogen source (Fig. 4B). Expression of the GAP2 gene was rapidly and strongly repressed when cells were shifted to YPD, whereas the expression of the other GAP genes (mainly GAP1, GAP3, and GAP6) was induced (Fig. 4B). Transfer of the cells from nitrogen starvation medium to YNB with ammonium sulfate led to similar results, only the values were lower for all investigated genes (data not shown).
When performing the opposite shift (from YPD to nitrogen starvation), we observed a very strong induction of the GAP2 gene (Fig. 4C), whereas the other GAP genes were  transiently upregulated to a very low extent, except for GAP6, which was more strongly upregulated upon a shift to nitrogen starvation (Fig. 4C).
Apart from transcriptional regulation, ScGap1 is also posttranslationally regulated. When ScGap1 is in the plasma membrane (during nitrogen starvation or during growth on nitrogen-poor sources), addition of ammonium, glutamine, or glutamate causes internalization of the ScGap1 permease (37). To determine the cellular localization and to follow the fate of the different Gap permeases under the different nitrogen source conditions, we generated strains where one of the endogenous GAP genes was tagged with the GFP. This allowed us to determine the presence of the different Gap proteins and their location in the cell during growth under nitrogen-poor or nitrogen-rich conditions (Fig. 5). The strong induction of GAP2 when cells were shifted from YPD to nitrogen starvation medium was confirmed with the GFP construct, as clear plasma membrane-localized expression was observed. At later time points (12 h), the permease was no longer in the membrane but seemed to be degraded in the vacuole (Fig. 5A). This was a second major difference with the situation in S. cerevisiae, where ScGap1 remained in the plasma membrane under nitrogen starvation conditions. To follow the expression of the other GAP genes, the cells were shifted from YNB with ammonium sulfate to nitrogen starvation medium. Similar to the qRT-PCR data, these permeases were expressed at low levels, and they were all rapidly degraded. We also performed GFP expression analysis for the opposite shift. Cells from nitrogen starvation conditions were transferred to YPD or YNB with ammonium sulfate, and pictures were taken after 6 h and 18 h (Fig. 5B). Gap2 was rapidly degraded, whereas all other Gap proteins (except Gap5, for which expression was not changed) were stably expressed in the plasma membrane, in accordance with the qRT-PCR experimental findings (Fig. 5B). We then studied the expression of the Gap proteins after shifting the cells from nitrogen starvation medium to YNB medium supplemented with a single amino acid as the sole source of nitrogen (Fig. 5C). As was expected from the previous results, the Gap2 protein was present in the plasma membrane under all tested conditions except YNB with ammonium sulfate or glutamine, similar to the findings with ScGap1. The other investigated proteins were all present in the plasma membrane in medium with ammonium sulfate, arginine, or glutamine, with minor differences in expression patterns. For Gap1, the presence at the plasma membrane was observed also when methionine, phenylalanine, or proline was present in the medium as the sole nitrogen source. For Gap4, this was the case when histidine, leucine, methionine, or proline was present in the medium as the sole source of nitrogen. In addition, while Gap5 was also localized at the plasma membrane when leucine was present in the medium as the sole source of nitrogen, Gap6 was detected when methionine or proline was the sole source of nitrogen.

Gap4 is involved in SAM transport and is required for SAM-induced morphogenesis.
We tested the different gap mutants for defects/changes in colony morphology and hypha formation. Strains were grown on several solid media known to induce the morphological switch (Spider, SLAD, LEE medium, and medium containing N-acetylglucosamine) (38)(39)(40)(41)(42) and on SLD and YNB media containing NH 4 as a control. In addition, we tested the influence of methionine and SAM on colony morphology. On the control media, all our investigated strains showed smooth colonies ( Fig. 6A and B). Under hypha-inducing conditions, all strains clearly showed comparable hyphal morphogenesis ( Fig. 6A to D), including the gap1⌬/gap1⌬ mutant, which was previously shown to have a defect in morphogenesis in Spider medium (36). Only the GAP3 deletion strain was slightly defective in hypha formation in Spider medium; hyphae started to grow 2 days later than in all other deletion and wild-type strains (Fig. 6A). The same phenotype was shown for three independent gap3⌬/gap3⌬ mutants. The gap4⌬/ gap4⌬ strain showed a strong defect in morphogenesis on SLD plus SAM medium. Colonies of this strain were smooth, and they did not start to create hyphae even after prolonged incubation (7 days) (Fig. 6B).
As our phylogenetic tree showed that Gap4 is homologous not only to ScGap1 but also to ScSam3, the S. cerevisiae high-affinity SAM transporter (K m of 3.3 M), we hypothesized that Gap4 may be responsible for the transport of this metabolite in C. albicans (30). To determine whether Gap4 is a SAM transporter, we expressed the GAP4 gene in an S. cerevisiae sam3⌬ strain. The expression was under control of the SAM3 promoter and terminator sequences, and expression of SAM3 itself was used as a positive control. SAM transport assays clearly showed that Gap4 is a SAM transporter ( Fig. 7; Fig. S2).

DISCUSSION
C. albicans lives in the human body as a commensal and under specific conditions as a pathogen. Under both conditions, the cells must be able to sense and take up the available nutrients present in a specific niche. To allow the uptake of amino acids, C. albicans cells secrete a large number of hydrolase enzymes (mainly belonging to the secreted aspartyl proteinase gene family) that convert proteins into peptides and amino acids that can be taken up by peptide transporters or amino acid permeases, respectively (26, 43). Previously, it was shown that the capacity to take up amino acids is strongly linked with virulence. Deletion of the endoplasmic reticulum-localized packaging chaperone Csh3 results in a defect in the expression of amino acid permeases, and a strain with such a deletion showed virulence defects (44). C. albicans expresses 28 putative amino acid transporters and 1 amino acid sensor (Csy1) (21,25,27,45,46). Phylogenetic analysis of the C. albicans permeases indicates that there are six orthologues of the S. cerevisiae Gap1, which is the general amino acid permease and is able to transport most amino acids. The fact that there are six ScGAP1 orthologues underscores the importance of amino acid uptake for this human fungal pathogen. Several more specific amino acid permeases are also expressed under different conditions. We have previously characterized the transport capacities of these six permeases by expressing them in yeast mutant permease strains (21). Gap2 turned out to be the permease with the broadest transport capacities, resembling that of ScGap1. Here we confirmed, based on growth and transport assay results, that Gap2 can be considered the functional orthologue of ScGap1. Deletion of only GAP2 showed a growth defect when single amino acids were present in the medium, such as leucine, phenylalanine, valine, or methionine, and the gap2⌬/gap2⌬ mutant showed a lower transport capacity for certain amino acids. The other single mutants did not show any major defect in growth or transport capacities, indicating that they probably share redundant substrate specificities or that these transporters are not expressed or are not localized to the plasma membrane under the experimental conditions. The results obtained with the GFP-tagged genes (Fig. 5) confirmed that these genes are differentially expressed and/or localized, depending on the environmental conditions. This suggests that in the human body, the different GAP genes may be differentially expressed depending on the location. This would not be surprising, as Candida colonizes many different niches in and on the body and can live in several niches both commensally and as a pathogen.
In S. cerevisiae, ScGap1 is not only an amino acid transporter but also a sensor for rapid activation of the PKA pathway when nitrogen-starved cells are provided with amino acids (19,47). We previously showed that expression of three C. albicans Gap permeases (Gap1, Gap2, and Gap6) in S. cerevisiae resulted in activation of trehalase, the PKA signaling pathway target, upon addition of specific amino acids to nitrogenstarved S. cerevisiae cells (21). However, in C. albicans we are not able to show an effect of specific amino acid transporters on the activation of trehalase. We previously showed that addition of glucose to glucose-starved cells also did not result in activation of trehalase (48). It is still possible that some of the Gap proteins in Candida activate the PKA pathway, but we are currently unable to prove this.
Real-time PCR analysis and GFP-protein fusion constructs showed a clearly distinct expression pattern between GAP2 and the five other GAP permease genes. GAP2 is not expressed in rich medium, but upon transfer of cells into nitrogen starvation medium expression is strongly induced, and the protein is plasma membrane localized. Maximal membrane localization persists only for a few hours, as longer incubation periods showed the GFP signal in intracellular vesicles and no longer localized to the membrane. This is clearly different from S. cerevisiae. The other Gap permeases are expressed and plasma membrane localized when cells are growing in rich medium. Transfer to nitrogen starvation medium results in their relocalization to the vacuole or intracellular vesicles. They may undergo a complicated regulation under different specific conditions in the host.
In C. albicans, there is a direct connection between the sensing of amino acids by the Csy1 sensor and the expression of SAP2, which encodes a secreted aspartyl proteinase, and the expression of Gap2. This is because the transcription factors Stp1 and Stp2 are proteolytically activated upon sensing of amino acids by Csy1 and they subsequently induce the expression of SAP2 and GAP2, respectively (49). The GATA transcription factors Gln3 and Gat1, which are involved in nitrogen catabolite repression, are required for the expression of STP1 when proteins are the only nitrogen source available (50,51). Possibly, SAP2 must first be expressed before the GAP genes are induced, and that is why they are only transiently localized at the plasma membrane. Whereas S. cerevisiae lives in environments with fluctuating concentrations of nitrogen sources, C. albicans lives in a more stable environment in the host, whose cells also require nitrogen, so the rapid signaling pathways reacting to the presence of nitrogen sources might not be mediated by activation of the PKA pathway but rather as the result of normal metabolism.
An interesting finding was the role of Gap4 in the uptake of SAM. Based on the heterologous expression of GAP4 in a yeast sam3 mutant, we found that Gap4 is a SAM transporter. Moreover, deletion of GAP4 results in a complete defect in SAM-induced morphogenesis. Both results suggest that Gap4 transports SAM, which is subsequently used inside the cells to trigger morphogenesis. When we deleted GAP4, SAM transport was only reduced by about 20% compared to that of the wild-type strain (data not shown). This suggests that other transporters may be the reason why transport is only reduced by 20%. Alternatively, it is possible that a small reduction in SAM transport strongly affects a metabolic or signal transduction pathway involved in morphogenesis. We are currently further investigating SAM-induced morphogenesis in C. albicans. SAM is known to function as a methyl donor, and therefore it may affect methylation of DNA, histones, or other proteins that may result in the yeast-to-hypha transition. Whereas for C. albicans there are few data, for S. cerevisiae results have been obtained that clearly link histone methylation with pseudohyphal growth (52). Previously, we showed that methionine is a strong inducer of morphogenesis (38). SAM is also known as an acyl acceptor, where an acyl group is placed on the methionyl amine of SAM, a reaction that results in the production of homoserine lactones, which are important quorum-sensing molecules in some bacteria (53). So far, a role of SAM regarding quorum sensing in C. albicans or, for example, during mixed bacterium-fungus biofilm infections, has not been studied. As SAM is a bacterial product and SAM induces morphogenesis in C. albicans, it is possible that Gap4 may be an interesting protein that is involved in mixed infections. It will also be interesting to see in the future what the role of the different Gap permeases may be under in vivo conditions. We already generated a triple deletion strain, but it would be interesting to have strains where only one of the six permeases is present. With the use of CRISPR-Cas9, generating such strains may be more convenient compared to the current classical gene disruption strategies.

MATERIALS AND METHODS
Sequence analysis. For BLAST analysis, protein sequences from the C. albicans database (CGD; http:// www.candidagenome.org/) and the S. cerevisiae database (SGD; http://www.yeastgenome.org/) were used. Phylogenetic analysis of all known amino acid permeases in S. cerevisiae, the six-member CaGap family, and all other known amino acid permeases in C. albicans was performed by using of the MEGA software (Molecular Evolutionary Genetics Analysis; version 4), using the neighbor-joining method (54) with 500 bootstrap replicates (the bootstrap consensus tree was inferred from 500 replicates). Distances between the permease names are related to the degree of divergence between sequences.
Strains and growth conditions. All strains used in this work are listed in Table 2. All strains were routinely grown in YPD medium (1% yeast extract, 2% peptone, 2% glucose) or in YNB medium (0.17% This study YNB without amino acids and without ammonium sulfate) supplemented when indicated with BSM (Brent supplement mixture) and/or with 5 mg/ml of ammonium sulfate, 300 g/ml of ammonium sulfate, or single amino acids as the sole source of nitrogen. For transport experiments or for GFP fluorescence microscopy experiments, strains were starved of nitrogen (NS; 0.17% YNB without amino acids and without ammonium sulfate). All media were supplemented when indicated with 2% glucose, except that nitrogen starvation medium was supplemented with 4% glucose and SLD medium was supplemented with 0.1% glucose. Solid media were supplemented with 2% agar. For induction of morphology, different hypha-inducing media were used: Spider medium (1% nutrient broth, 0.2% K 2 HPO 4 , 1.35% agar, and 1% mannitol as a carbon source; pH 7.2) (42); Lee medium (described previously [41]); YNB with GlcNAc (0.17% YNB without amino acids or ammonium sulfate and with 100 g/ml N-acetylglucosamine); synthetic low ammonium (SLAD) medium (0.17% YNB without amino acids or ammonium sulfate and with 50 M ammonium sulfate and 2% glucose) (55); synthetic low-dextrose (SLD) medium (0.17% YNB without amino acids or ammonium sulfate and with 5 mg/ml of ammonium sulfate and 0.1% glucose; pH adjusted to 6.5) (38). Deletion and reintegration. We generated single GAP deletion strains by using the fusion PCR method (31, 32) (see Table 2 for the different strains that were obtained). The SN87 strain was used as a starting strain. We also generated double and triple GAP deletion strains: gap1⌬/gap1⌬ gap2⌬/gap2⌬, gap2⌬/gap2⌬ gap6⌬/gap6⌬, gap1⌬/gap1⌬ gap6⌬/gap6⌬, and gap1⌬/gap1⌬ gap2⌬/gap2⌬ gap6⌬/ gap6⌬, using the SAT1 flipper cassette (32,33). Single GAP deletion strains created with the fusion PCR method were used as starting strains. For creation of CaGAP2 and CaGAP4 reintegrants, the pSFS2A plasmid with the SAT1 cassette was used (34), starting with the single GAP deletion strains gap2⌬/gap2⌬ and gap4⌬/gap4⌬. The CaGAP2 or the CaGAP4 gene with approximately 500 bp of promoter sequence and approximately 500 bp of terminator sequence was integrated between the KpnI and the XhoI restriction sites of the pSFS2A plasmid, and the HIS1 marker was integrated between the NotI and the SacII restriction sites of the pSFS2A plasmid. The reintegration cassette was cut out of the plasmid with KpnI and SacII and used for homologous recombination into the gap2⌬/gap2⌬ and gap4⌬/gap4⌬ strains (Candida maltosa LEU2 and Candida dubliniensis HIS1 are present at the loci of the GAP alleles). Reintegrants were selected on YPD medium plus 200 g/ml nourseotricin, and afterwards the SAT1 cassette was removed as described elsewhere (34). The reintegrants were verified by PCR. All primers used for construction and diagnosis of reintegrants are listed in Table S1 in the supplemental material.
Growth curve measurements. To test the growth capacities of our strains, a microplate reader technique was applied (56). Freshly grown cells (on YPD plates) were suspended in sterile PBS (phosphate-buffered saline: 0.8 g/liter NaCl, 0.2 g/liter KCl, 1.44 g/liter Na 2 HPO 4 , 0.24 g/liter KH 2 PO 4 ; pH 7.4) to obtain an optical density at 600 nm (OD 600 ) of 1 and used to inoculate liquid YNB medium (supplemented with 300 g/ml of ammonium sulfate or single amino acids as sole sources of nitrogen, unless otherwise indicated) and grown to an OD 600 of 0.002. Aliquots (100 l) were transferred to a 96-well plate. The cell cultures in the plate were incubated at 30°C in an ELx808 absorbance microplate reader, and the optical density in each well was measured at 595 nm over a period of 24 h under continuous shaking. Growth curve measurements were repeated at least twice, each time with four identical parallel cultures for each strain and each set of tested conditions. Growth assays. The growth rate in the presence of different nitrogen sources was assayed by a classical drop test. Freshly grown cells (on YPD plates) of each tested strain were suspended in sterile PBS and adjusted to the same initial OD 600 of 1. Tenfold serial dilutions were prepared, and 3-l aliquots were spotted on a series of plates containing YNB with 300 g/ml of ammonium sulfate or a single amino acid as sole source of nitrogen (unless otherwise indicated). Plates were incubated at 30°C for 3 days, and digital gray-scale images of growing colonies were obtained using a Nicon Coolpix4500 digital camera. Representative results are reported.
Measurement of transport activity. Candida cells for amino acid uptake measurements were grown to exponential phase in liquid YNB medium supplemented with BSM and ammonium sulfate, harvested by centrifugation, resuspended in nitrogen starvation medium, and further cultivated either for 24 h for nitrogen-poor condition transport experiments or for only 12 h and then harvested again by centrifugation and resuspended in YNB supplemented with BSM and 5 mg/ml of ammonium sulfate. The second type of cultures were grown for another 5 to 6 h for nitrogen-rich condition transport experiments. S. cerevisiae cells for SAM uptake measurements were grown in complete synthetic medium (MP Biomedicals) at 30°C overnight, then diluted in YNB medium (5 mg/ml ammonium sulfate) and grown at 30°C overnight. Cells for amino acid or SAM uptake were cooled on ice for 20 min, harvested by centrifugation, washed with 5 ml 10 mM morpholineethanesulfonic acid (MES) buffer (pH 6.5; containing 0.1 mM EDTA for transport of Candida), and resuspended at a cell density of 80 mg (wet weight)/ml in corresponding fresh medium. Thereafter, 40 l of the cell suspension was preincubated at 30°C in a shaking water bath for 10 min, then 10 l of a mixture of unlabeled (2 mM final concentration for amino acid transport and 0.1 mM for SAM transport) and 14 C-labeled L-amino acid (2 mM unlabeled amino acid [final concentration] plus 14 C-labeled L-amino acid, to obtain a specific activity of 3,000 cpm/nmol) or 3 H-labeled S-adenosyl methionine (0.1 mM unlabeled SAM [final concentration] plus 3 H-labeled SAM, to obtain a specific activity of 3,000 cpm/nmol) was added to the cells, and the resulting mixture was incubated at 30°C for 1 min. The reaction was stopped by cooling via addition of 5 ml ice-cold water. The obtained suspension was filtered through a glass microfiber filter (Whatman GF/C; retention particle size, 1.2 m) prewet with 2 mM unlabeled amino acid or 0.1 mM unlabeled SAM and immediately washed twice with 4 ml ice-cold water. Afterwards, the filter was deposited in a scintillation tube which contained 4 ml scintillation liquid (Lumagel Safe; PerkinElmer). Radioactivity was measured in a liquid scintillation counter (model LS6500; Beckman Coulter, Inc.). For each determination, three samples and two blanks were taken (addition of ice-cold water followed by addition of the labeled amino acid). In addition, 500 l of the cell suspension was used to determine the protein content based on the Bradford method (57). The labeled amino acid or SAM solution in a total volume of 10 l was used for the determination of specific activity. The transport activity was calculated from the measured radioactivity and estimated protein content in the samples, and it was expressed as nanomoles transported per minute per milligram of protein. Three independent experiments were performed, and either the standard deviation was calculated between these experiments (for amino acid transport assay) or between different transformants in one representative experiment (for SAM transport in S. cerevisiae assay). A two-sided t test was used to determine statistical significance between transport rates of different strains.
RNA extraction and quantitative real-time PCR. Cells for RNA extraction were grown to exponential phase in liquid YPD medium, harvested by centrifugation, and either (i) resuspended in nitrogen starvation medium and samples taken after 15 or 30 min or (ii) resuspended in nitrogen starvation medium, cultivated for a further 12 h, harvested by centrifugation, and resuspended in YPD or YNB with 5 mg/ml of ammonium sulfate supplemented with 100 g/ml single amino acid (for csy1 qRT-PCR experiments), and samples were taken after 15 or 30 min. The samples were cooled by addition of ice-cold water, centrifuged at 4°C, and washed once with 1 ml ice-cold water, and the pellet was flash-frozen in liquid nitrogen and stored at Ϫ80°C. Total RNA was extracted from yeast cells with Trizol reagent (Invitrogen). RNA purity was determined with a NanoDrop Technologies ND-1000 spectrophotometer. After digestion with RNase-free DNase (Promega) to eliminate genomic contamination, cDNA was synthesized from 1 g of total RNA with the Bio-Rad iScript cDNA synthesis kit according to the manufacturer's instructions. The relative gene expression was determined by quantitative PCR on a StepOnePlus real-time PCR system (Applied Biosystems) using the KAPA fast kit (KAPA Biosystems). This kit uses SYBR green, which is a fluorogenic dye that shows little fluorescence in solution but emits a strong fluorescent signal upon binding to double-stranded DNA, resulting in increased fluorescence as the PCR product accumulates. The fold change was calculated using the comparative threshold cycle (C T ) method, and error bars showing the range of possible relative quantity values were defined by the standard error of the ΔΔC T values (see http://www.science.smith.edu/cmbs/wp-content/uploads/sites/ 36/2015/09/Analyzing-your-QRT-for-relative-2%5E-%E2%88%86%E2%88%86Ct.pdf). (58,59). All data were normalized to TEF1, which was used as the reference housekeeping gene. The data presented in Fig. 4 are the summary of three independent experiments. The sequences of the primers used in this assay are provided in Table S1 in the supplemental material.
GAPx-GFP strain construction. To tag GAP genes with GFP (optimized for C. albicans [60]), GAP gene-specific sequences (approximately 70 nucleotides in length) were added to the universal primer sequences (Table S1; lowercase and uppercase letters, respectively) in the pGFP-HIS1 plasmid (61). CaGAPx-GFP strains were created from gapx/GAPx His Ϫ heterozygous strains (where only one copy of CaGAP is still present) by homologous recombination of GFP sequences into the 3= end of the CaGAPs open reading frames. CaGAP-GFP mutants were selected by growth on medium without histidine and verified by PCR afterwards (see Table S1 for a list of the primers used).
Fluorescence microscopy. Cells for fluorescence microscopy were grown to exponential phase in YNB medium supplemented with 5 mg/ml of ammonium sulfate, harvested by centrifugation, and either (i) resuspended in nitrogen starvation medium and further cultivated for 1, 3, or 12 h, or (ii) resuspended in nitrogen starvation medium, further cultivated for 12 h, harvested by centrifugation, and resuspended in YPD or YNB supplemented with ammonium sulfate (5 mg/ml), grown for another 6 or 18 h, resuspended in YNB medium supplemented with a single amino acid (100 g/ml) as the nitrogen source, and grown for another 5 to 6 h. Nitrogen-starved cells grown for 1, 3, 5 to 6, 12, or 18 h were collected by centrifugation and used directly without fixation, visualized by fluorescence microscopy (the fluorescence signal in the cells was observed under an Olympus AX70 microscope using a U-MWB cube with a 450-to 480-nm excitation filter and 15-nm barrier filter) and differential interference contrast (DIC) optics (Nomarski). Various exposure times for different media were used (as indicated in the figure legends). Multiple images of representative cells (almost the whole population behaved similarly under our conditions) were captured with a DP70 digital camera using the program DP Controller. Subsequent processing was done using Adobe Photoshop.
Colony morphology. To determine the ability to undergo the yeast-to-hypha transition, freshly grown C. albicans cells (on YPD plates) were suspended in sterile PBS to obtain an OD 600 of 0.1. This suspension was diluted 1/1,0000 in four steps, leading to a cell suspension containing approximately 100 cells/ml. Afterwards, 100 l of the suspensions was plated on the indicated media. In some cases, SLD and SLAD media were supplemented with 5 g/ml of methionine or SAM. Plates were incubated at 30°C or 37°C. Pictures of representative colonies were taken using a binocular microscope.
Complementation of ScSAM3 with GAP4. The ScSAM3 construct with restriction sites (PstI and BamHI) was made through PCR of 1,000 bp of promoter sequence, the ScSAM3 gene, and 500 bp of terminator sequence by using the primers SAM3promF-PstI and SAM3terR-BamHI.
The CaGAP4 construct was made by fusion PCR between the 1,000-bp ScSAM3 promoter, the CaGAP4 gene, and the 500-bp ScSAM3 terminator. To make this construct, the following primers were used: SAM3promF-PstI, SAM3promR-fusion, GAP4F-fusion, GAP4Revfusion, SAM3terR-BamHI, and SAM3terFfusion. Both constructs were ligated into the multiple cloning site of the YCplac33 vector upon restriction digestion with PstI and BamHI. The S. cerevisiae sam3⌬ strain GV230 was transformed with the newly created plasmids and empty YCplac33 vector, using the Gietz transformation method (62). Three of the transformants and the wild-type strain GV287 were grown in 50 ml of YNB plus 5 mg/ml ammonium sulfate medium at 30°C overnight. Afterwards, the cultures were centrifuged and cells were washed with 5 ml MES buffer (pH 6). The pellet was resuspended in YNB plus ammonium sulfate medium to achieve a concentration of 80 mg/ml. The transport experiment was carried out as stated above (SAM concentration, 0.5 mM).
Reproducibility of results. All experiments were performed at least three times with independent deletion strains and/or cultures. The results always showed consistent trends, i.e., differences between strains and mutants were highly reproducible.
The results of single time point measurements (e.g., uptake rate of amino acid, mRNA expression levels) are based on three independent measurements, and standard deviations are indicated by the error bars in the graphs in the figures.