Dal81 Regulates Expression of Arginine Metabolism Genes in Candida parapsilosis.

Fungi can use a wide variety of nitrogen sources. In the absence of preferred sources such as ammonium, glutamate, and glutamine, secondary sources, including most other amino acids, are used. Expression of the nitrogen utilization pathways is very strongly controlled at the transcriptional level. Here, we investigated the regulation of nitrogen utilization in the pathogenic yeast Candida parapsilosis. We found that the functions of many regulators are conserved with respect to Saccharomyces cerevisiae and other fungi. For example, the core GATA activators GAT1 and GLN3 have a conserved role in nitrogen catabolite repression (NCR). There is one ortholog of GZF3 and DAL80, which represses expression of genes in preferred nitrogen sources. The regulators PUT3 and UGA3 are required for metabolism of proline and γ-aminobutyric acid (GABA), respectively. However, the role of the Dal81 transcription factor is distinctly different. In S. cerevisiae, Dal81 is a positive regulator of acquisition of nitrogen from GABA, allantoin, urea, and leucine, and it is required for maximal induction of expression of the relevant pathway genes. In C. parapsilosis, induction of GABA genes is independent of Dal81, and deleting DAL81 has no effect on acquisition of nitrogen from GABA or allantoin. Instead, Dal81 represses arginine synthesis during growth under preferred nitrogen conditions. IMPORTANCE Utilization of nitrogen by fungi is controlled by nitrogen catabolite repression (NCR). Expression of many genes is switched off during growth on nonpreferred nitrogen sources. Gene expression is regulated through a combination of activation and repression. Nitrogen regulation has been studied best in the model yeast Saccharomyces cerevisiae. We found that although many nitrogen regulators have a conserved function in Saccharomyces species, some do not. The Dal81 transcriptional regulator has distinctly different functions in S. cerevisiae and C. parapsilosis. In the former, it regulates utilization of nitrogen from GABA and allantoin, whereas in the latter, it regulates expression of arginine synthesis genes. Our findings make an important contribution to our understanding of nitrogen regulation in a human-pathogenic fungus.

INTRODUCTION

Nitrogen is a key component of all proteins, and fungi can use a variety of compounds as a sole source. Preferred nitrogen sources include glutamate, glutamine, ammonium, and peptones (1). When these sources are available, expression of genes associated with the utilization of poor nitrogen sources is repressed, in a process called nitrogen catabolite repression (NCR) (2–5). Regulation of nitrogen metabolism has been particularly well characterized in Saccharomyces cerevisiae, where expression of many nitrogen-responsive genes is controlled by four GATA-type transcription factors (2). Two (Gzf3 and Dal80) are repressors that switch off expression of target genes when preferred nitrogen sources are available (6, 7). The other two (Gat1 and Gln3) are activators and are required for expression of genes involved in acquisition of nitrogen from poor sources. During growth on preferred nitrogen sources, Gat1 and Gln3 are phosphorylated and are sequestered in the cytoplasm by the activity of Ure2 (8–11). When only poor nitrogen sources are present, Gat1p and Gln3p are dephosphorylated and translocated to the nucleus, where they drive expression of genes involved in assimilating and catalyzing degradation of nonpreferred nitrogen sources (1, 12–14). Nitrogen-rich cellular processes, such as protein translation and amino acid biosynthesis, are downregulated during nitrogen starvation (2). Many proteins, including membrane transporters that import ammonium (15, 16), amino acid transporters (17), and secreted proteases that degrade environmental proteins to release amino acids (4), are subject to NCR at both the mRNA and protein levels.

Several other proteins regulate expression of nitrogen metabolism genes. For example, Gcn4 regulates expression of many amino acid biosynthesis genes (18). In S. cerevisiae, Gcn4 is strongly regulated at the translational level through competition with small upstream open reading frames (uORFs) (19), which are conserved in Candida albicans, though they may not function in exactly the same manner (20, 21). In S. cerevisiae, translation of Gcn4 is increased during amino acid starvation, leading to derepression of most amino biosynthetic pathways (22). Growth on some amino acids as the sole nitrogen source also induces expression of specific transcription factors that are required for utilization of these sources. For example, growth of S. cerevisiae on proline induces expression of Put3, which drives expression of genes involved in proline import and utilization (23, 24), and induction of arginine utilization genes requires Arg80 and Arg81 (25, 26). In S. cerevisiae, Dal81 is required for the expression of genes involved in the metabolism of allantoin, γ-aminobutyric acid (GABA), leucine, and urea (27, 28). During growth on GABA, Dal81 acts with Uga3 to drive expression of GABA-responsive genes, including the UGA regulon (29). When allantoin is the sole nitrogen source, Dal81, together with Dal82, controls expression of the DUR and DAL genes (30). In addition, Dal81 acts with Stp1 to regulate expression of amino acid permease genes, especially those that import leucine (31).

NCR has been characterized in several other fungi, including Yarrowia lipolytica, Aspergillus nidulans, Neurospora crassa, Cryptococcus neoformans, and Candida albicans (3, 32–34). The roles of the GATA activators are generally well conserved. For example, in C. albicans, Gat1 and Gln3 are required for derepression of expression of the ammonium permease MEP2 and the secreted aspartyl protease SAP2 as well as the urea, allantoin, and GABA metabolism genes DUR1, DUR2, DAL5, and UGA4 during growth on preferred nitrogen sources (34–38). Loss of either GAT1 or GLN3 leads to attenuated virulence of C. albicans in a mouse model of infection (34, 36). Deleting GAT1 or GLN3 also reduces formation of chlamydospores, unusual globular structures that are formed during growth on some media (39).

The GATA repressors are not as well characterized in other fungi as they are in S. cerevisiae. DAL80 and GZF3 are ohnologs resulting from whole-genome duplication (WGD) in the ancestor of S. cerevisiae (40, 41). Fungi that diverged from S. cerevisiae before the WGD therefore have only one ortholog. The role of the single C. albicans ortholog (called GZF3) has not been studied. Y. lipolytica encodes four GATA transcription factors, including two (Gzf1 and Gzf2) that are similar to Gat1 and Gln3; Gzf3, which is related to S. cerevisiae Gzf3 and Dal80; and Gzf4, which is more closely related to iron response regulators (32). Gzf3 represses expression of nitrogen utilization genes in Y. lipolytica, similarly to S. cerevisiae, whereas Gzf4 is more closely related to iron response regulators (32). In many filamentous fungi, the GATA transcription factor AreA (Gat1 family) is an activator of gene expression whereas AreB (Gzf3) represses expression (42, 43).

Here, we used reverse genetics to assess the role of NCR regulators in the pathogenic yeast Candida parapsilosis, a relative of C. albicans in the CUG-Ser clade (44, 45). We found that GAT1 and GLN3 orthologs are activators of NCR and that Gzf3 acts as a general repressor of nitrogen catabolism genes. The roles of Put3 and Uga3 as regulators of proline and GABA metabolism, respectively, are also conserved with S. cerevisiae. However, the role of Dal81 is different. DAL81 is not required for allantoin or GABA metabolism in either C. albicans or C. parapsilosis. Transcriptomic analysis shows that in C. parapsilosis, Dal81 acts as a regulator of genes required for metabolism of arginine and especially as a repressor of arginine biosynthesis and transport. There has therefore been substantial rewiring of the Dal81 transcriptional network between S. cerevisiae and C. parapsilosis.

RESULTS

Nitrogen catabolite repression in C. parapsilosis.

Fungal species preferentially use nitrogen sources such as glutamate, glutamine, ammonium, and peptones, but they can also utilize nonpreferred sources, including allantoin, γ-aminobutyric acid (GABA), and other amino acids. We first tested if gene expression is subject to nitrogen catabolite repression in C. parapsilosis CLIB214 in a manner similar to that seen with other fungi by comparing the levels of gene expression of wild-type cells growing in media containing a preferred nitrogen source (yeast nitrogen base [YNB] with glucose and ammonium sulfate) and in media with a nonpreferred source (YNB with glucose and isoleucine).

Overall, 731 genes were upregulated and 292 genes were downregulated in cells grown on isoleucine compared to ammonium sulfate as the sole nitrogen source (see Data Set S1 in the supplemental material). Upregulated genes were enriched in transmembrane transporters and permeases, including transporters of ammonium (MEP1 and MEP2), urea (DUR3 and DUR4), allantoate (DAL4, DAL7, and DAL9), and oligopeptides (OPT2, OPT5, OPT6, and OPT8). Genes involved in amino acid transport and synthesis, especially those involved in arginine biosynthesis (e.g., ARG1, ARG3, ARG5,6 and CPA2), were also upregulated. Expression of the putative NCR activator GAT1 was increased by a log2 fold change (log2FC) value of ~6 and that of the putative NCR repressor GZF3 by a log2FC of ~2. These responses are consistent with an increase in levels of nitrogen scavenging pathways during growth on nonpreferred nitrogen sources and are similar to results of previous analyses of derepression of pathways subject to NCR in S. cerevisiae (4, 46).

RNA-seq analysis of nitrogen response of C. parapsilosis wild-type and dal81 deletion strains. Download DATA SET S1, XLS file, 1.3 MB.

Identification of NCR regulators in C. parapsilosis.

The roles of C. parapsilosis orthologs of known nitrogen regulators in S. cerevisiae were next tested by characterizing the phenotype of gene deletion strains growing on various nitrogen sources (Fig. 1). GAT1 and GZF3 deletions in C. parapsilosis were previously described by Holland et al. (47); deletions of GLN3, PUT3, GCN4, UGA3, and DAL81 were constructed using a similar methodology. Most of the deletion strains grew well on rich media with complex nitrogen sources (yeast extract-peptone-dextrose [YPD], glucose with yeast extract and peptone) and on YNB plus glucose when glutamate (a preferred nitrogen) was used as the sole nitrogen source. Loss of GLN3 resulted in poor growth on ammonium sulfate and under most of the nitrogen-limiting conditions tested, similarly to the orthologous deletion in C. albicans (36). Deletion of the GAT1 activator led to dramatically reduced growth on tryptophan, which is also observed in C. albicans (36). Deleting GAT1 resulted in a minor growth defect on glutamate, the preferred nitrogen source, which has not been reported in other fungal species. Overall, the growth phenotypes support the hypothesis that Gat1 and Gln3 are activators of genes required for utilization of nonpreferred nitrogen sources in C. parapsilosis, similarly to S. cerevisiae and C. albicans.

FIG 1 

Identification of regulators of nitrogen metabolism in Candida species. (A) C. parapsilosis strains were grown overnight in YPD, and increasing 1/5 dilutions were pinned on YPD and on minimal media (YNB with glucose and without amino acids or ammonium sulfate) supplemented with a 10 mM concentration of the indicated amino acid or allantoin or with 0.5% ammonium sulfate. Plates were photographed after 72 h, and growth of the deletion strains was compared to the growth of the control C. parapsilosis CPRI strain. (B) Disrupting DAL81 reduced growth of S. cerevisiae when allantoin or GABA was the sole nitrogen source but had no effect on growth of C. albicans and C. parapsilosis. Deleting UGA3 reduced growth of both C. parapsilosis and C. albicans on GABA. C. p, C. parapsilosis; C. a, C. albicans, S. c, S. cerevisiae. The relevant control strains for each species were used (C. parapsilosis CPRI, C. albicans STCA2 [for dal81], C. albicans SN152 [for uga3Δ {49}], and S. cerevisiae 23344c). The asterisks indicate that DAL81 was edited in C. albicans by introducing two stop codons. For the S. cerevisiae strains, uracil was added to the media at a final concentration of 0.08 g/liter. For C. albicans SN152 and the corresponding uga3Δ strain, arginine was added to reach a final concentration of 0.05 g/liter.

There is one ortholog of S. cerevisiae GZF3 and DAL80, called GZF3, in Candida species (40, 41). Deleting GZF3 in C. parapsilosis had no effect on growth on any of the nitrogen sources tested (Fig. 1). In S. cerevisiae, Gzf3 and Dal80 repress expression of metabolic genes such as those encoding transporters of ammonium and amino acids during growth on preferred sources (6). GAT1 is also one of the targets (48). We therefore tested the effect of deleting C. parapsilosis GZF3 on expression of some potential targets, including MEP2 (ammonium permease and sensor of nitrogen starvation), GAP2 (amino acid permease), and GAT1, which were upregulated when C. parapsilosis was grown on isoleucine as the major nitrogen source (Data Set S1). Table 1 shows that expression of all three genes was increased (8-fold to 80-fold) when GZF3 was deleted during growth on rich media with complex sources of nitrogen (YPD). Gzf3 therefore represses expression of genes required for metabolism of nonpreferred nitrogen sources.

TABLE 1 

Relative expression levels of NCR-sensitive genes in gzf3Δ strains grown in YPD

Gene	Fold change (2−ΔΔCT)a		P valueb
	C. parapsilosis CPRI	C. parapsilosis gzf3Δ/Δ
GAP2	1.0 (0.32–3.15)	8.85 (2.54–30.87)	0.0006
MEP2	1.0 (0.47–2.13)	79.5 (27.7–228.2)	0.0000001
GAT1	1.0 (0.41–2.46)	23.43 (7.0–78.43)	0.003

Threshold cycle (C) values were normalized to ACT1, and the expression level in C. parapsilosis CPRI was set to 1. Gene expression ranges (standard deviations) are shown in parentheses.

P values comparing expression levels in C. parapsilosis CPRI and gzf3Δ strains were calculated from three biological replicates by the use of a two-tailed Student t test.

The roles of some other transcription factor orthologs are also conserved in S. cerevisiae and C. parapsilosis. Deleting GCN4 in C. parapsilosis reduced growth on most media (Fig. 1). This is similar to results seen with both S. cerevisiae and C. albicans, where Gcn4 plays a key role in coordinating the response to amino acid starvation (18, 21). PUT3 was required for utilization of proline only and UGA3 for utilization of GABA only (Fig. 1). However, deleting DAL81 had no effect on growth under any set of conditions. To test whether this phenotype is unique to C. parapsilosis, we edited the DAL81 ortholog in C. albicans by introducing two stop codons by the use of a clustered regularly interspaced short palindromic repeat (CRISPR)-Cas9-based method. We found that the edited strain had no defect in growth on various nitrogen sources, even when allantoin or GABA was the sole nitrogen source (Fig. 1). In S. cerevisiae, Dal81 is required for metabolism of GABA (together with Uga3) and of allantoin (together with Dal82) (Fig. 1) (28, 29). The role of Uga3 in GABA metabolism is conserved in both C. albicans (49) and C. parapsilosis (Fig. 1B). There is no ortholog of DAL82 in Candida species.

Role of C. parapsilosis Dal81.

Our observation that deleting DAL81 did not affect growth of either C. albicans or C. parapsilosis on allantoin or GABA suggests that the protein has functions in Candida that are different from its functions in Saccharomyces species. To determine if DAL81 plays any role in regulating expression of genes required for GABA metabolism in C. parapsilosis, we compared the gene expression profiles of wild-type and dal81 deletion strains growing in minimal media supplemented with either ammonium sulfate or GABA as the sole nitrogen source. We used C. parapsilosis CPRI as a control strain (47). The gene deletions were constructed by replacing one allele with HIS1 from C. dubliniensis and one with LEU2 from C. maltosa. C. parapsilosis CPRI also contains one C. dubliniensis HIS1 (CdHIS1) allele and one C. maltosa LEU2 (CmLEU2) allele (47). In C. parapsilosis CPRI, expression of 198 genes was increased by a log2FC of >1 during growth on GABA compared to ammonium sulfate, and expression of 117 of these genes was also increased even when DAL81 is deleted (Fig. 2). Similarly, expression of 240 genes was decreased in the control strain during growth on GABA relative to ammonium sulfate, and 141 of these were also downregulated in the DAL81 deletion. Growth on GABA resulted in increased expression of genes involved in GABA metabolism, including the UGA transaminase (UGA1) and succinate-semialdehyde dehydrogenase (UGA2), even when DAL81 was deleted, by a log2FC of >5 (Data Set S1). We also used quantitative real-time PCR (RT-PCR) to confirm that DAL81 is not required for the GABA-specific induction of metabolic genes (Table 2). Expression of UGA1 and UGA2 was induced >100-fold during growth in GABA, in the presence and absence of DAL81.

FIG 2 

GABA induction of gene expression does not require DAL81. (A) Genes upregulated and downregulated in C. parapsilosis CPRI and dal81Δ strains grown in GABA as a sole nitrogen source. Gene expression is compared to growth on ammonium sulfate for each strain. (The full list is available in Data Set S1.) (B) Genes with the highest induction in expression under conditions of growth on GABA. The log2 fold change in C. parapsilosis CPRI and dal81Δ strains is shown. (The full list is available in Data Set S1.) WT, wild type.

TABLE 2 

GABA induction of gene expression in control and dal81Δ strains of C. parapsilosis

Gene	Fold change (2−ΔΔCT)a				P valueb
	C. parapsilosis CPRI		C. parapsilosis dal81Δ/Δ
	Ammonium	GABA	Ammonium	GABA
UGA1	1.0 (0.2–4.7)	142.5 (45.4–447.1)	1.0 (0.2–6.0)	118.6 (54.8–256.9)	0.57
UGA2	1.0 (0.2–6.1)	219.3 (31.7–1,516.6)	1.0 (0.2–6.8)	113.1 (28.7–445.3)	0.95

C values were normalized to ACT1 to generate ΔC values for each gene, and the expression level during growth in ammonium sulfate was set to 1. Gene expression ranges (standard deviations) are shown in parentheses. Cells were grown in YNB/glucose with 0.5% ammonium sulfate or 10 mM GABA.

P values comparing expression levels of control (CPRI) and dal81 deletion strains in GABA were calculated from three biological replicates by the use of a two-tailed Student t test.

In S. cerevisiae, GABA is imported by the Uga4 permease (50). There is no syntenic homolog of UGA4 in C. parapsilosis (51, 52). GABA may be imported by UGA6 (CPAR2_212440) and/or CPAR2_212360, a member of the same permease family, because expression of these genes was increased by a log2FC of >7.0 in media containing GABA (Fig. 2; see also Data Set S1). However, expression of both was also induced when isoleucine is the sole nitrogen source. Expression of DAL9, encoding a putative allantoate permease, was induced by GABA in the presence or absence of DAL81 (Fig. 2). It is therefore likely that in C. parapsilosis, and possibly in other Candida species, DAL81 does not regulate expression of GABA or allantoate metabolism genes, as indicated by the lack of a growth defect (Fig. 1).

Deleting DAL81 had little effect on overall gene expression levels when cells were grown in ammonium sulfate. Expression of only 40 genes was decreased in the dal81 deletion strain relative to C. parapsilosis CPRI, and expression of 15 genes was increased with a log2FC of >1 (Data Set S1). However, expression of arginine biosynthesis genes CPA2, ARG4, and ARG5,6 was increased and expression of CAR1 and CAR2, involved in arginine degradation, was decreased (Data Set S1). Expression of DUR1,2 (urea amidolyase) and of DUR3 (a putative urea transporter) was reduced (by a log2FC value greater than 1.8) in the dal81 deletion strain compared to a control strain grown in ammonium sulfate, suggesting that its role in regulation of urea metabolism may be partially conserved with S. cerevisiae (Data Set S1) (53). The expression of galactose metabolism genes (GAL1 and GAL7) was also reduced. However, expression of the DUR genes was induced by growth on GABA, even when DAL81 was deleted (Data Set S1).

To further characterize the role of DAL81, we compared the gene expression profile of a dal81 deletion strain to that of the wild type under conditions of cell growth in YPD, a rich medium containing complex nitrogen sources, including yeast extract and peptones. Expression of 493 genes was increased and of 337 genes was decreased in the dal81 deletion strain (Data Set S1). The upregulated genes were involved in processes associated with ribosome biogenesis and, in particular, with arginine metabolism (Data Set S1). Expression of several genes required for arginine biosynthesis and transport was increased, whereas expression of arginine catabolic genes was decreased (Fig. 3).

FIG 3 

DAL81 regulates expression of arginine metabolism in C. parapsilosis. The diagram shows a schematic representation of the arginine biosynthetic pathway in yeast. Expression of the genes highlighted in blue was increased in a dal81 deletion strain of C. parapsilosis growing in YPD. Expression of genes shown in red was decreased. The gene shown in black (ARG2) had no detectable change in expression. The log2FC data are indicated in parentheses. (Full data are available in Data Set S1.) CoA, coenzyme A.

To confirm that the observed increase in expression of ARG genes was caused by deleting dal81 and was not a result of a secondary mutation, we restored an intact copy of DAL81 at one of the deleted loci in C. parapsilosis (Fig. S1). Expression of ARG1 and ARG3 was found to have increased 6- to 60-fold when DAL81 was deleted and was restored to wild-type levels in the complemented strain (Table 3). We also used CRISPR-based editing to delete DAL81 in a clinical isolate, C. parapsilosis 90-137 (54). Expression of ARG3 was increased >200-fold in the dal81 deletion strain (Table 3). In addition, deleting DAL81 did not reduce growth of C. parapsilosis 90-137 on GABA or allantoin as the sole nitrogen source (Fig. S2). The effect of deletion of DAL81 on expression of arginine genes and the lack of a role in regulating allantoin and GABA metabolism are therefore not restricted to one isolate of C. parapsilosis. Deleting DAL81 did not reduce growth of C. parapsilosis or C. albicans on arginine or ornithine as the sole nitrogen source and did not affect susceptibility to canavanine, which is a toxic analog of arginine (Fig. S3). It is, however, likely that expression of arginine biosynthesis genes in Candida species is also regulated by other activators, similarly to S. cerevisiae (25, 26).

TABLE 3 

Reintroducing DAL81 restores expression of ARG genes in C. parapsilosis cells grown in YPD

Gene	Fold change (2−ΔΔCT)a
	CLIB214			90-137
	WTb	dal81Δ/Δ	dal81Δ::DAL81	WTb	dal81Δ/Δ
ARG1	1.0 (0.1–10.1)	6.3 (1.8–22.2)	0.6 (0.3–1.4)
ARG3	1.0 (0.1–12.1)	62.2 (15.3–252.1)	0.7 (0.3–1.4)	1.0 (0.9–1.1)	244.7 (179.6–333.5)

C values were normalized to ACT1, and the expression level in C. parapsilosis CLIB214 or C. parapsilosis 90-137 was set to 1. Gene expression ranges (standard deviations) are shown in parentheses.

WT, wild type.

Complementation of C. parapsilosis dal81 deletion. (A) DAL81 was deleted by replacing both alleles with the auxotrophic markers CmLEU2 and CdHIS1 (47). (B) To reintroduce the wild-type gene, the DAL81 ORF was cloned into plasmid pSFS2A (80). The selectable marker (SAT1 cassette) was flanked on one side by DAL81, and the other by part of the CdHIS1 ORF used to delete one of the DAL81 alleles. A linear fragment was transformed into the knockout strain. (C) Structure of one of the alleles after reintegration of DAL81 in the knockout strain. The construct was verified using primers (purple arrows) to amplify inside ORF (ORF presence check) and around the 5′ and the 3′ ends of the integration site. (E) The DAL81 ORF (802-bp) site was amplified from the wild-type strain and four putative reintegration isolates (RI 1, 2, 3, and 4), but not from the deletion strain (KO [knockout]). A 1,098-bp fragment was amplified from the wild-type strain and RI 1 but not from the others. A 1,103-bp fragment was amplified from the 3′ end of RI 1 only. The fragments from the 5′ and 3′ ends of RI 1 were sequenced to confirm the construction. Download FIG S1, EPS file, 0.9 MB.

Deleting DAL81 in C. parapsilosis 90-137 did not affect growth on allantoin or GABA as the sole nitrogen source. Both alleles of DAL81 were deleted in C. parapsilosis 90-137 using CRISPR. Strains were grown overnight in YPD, and increasing 1/5 dilutions were plated on minimal media (YNB with glucose without amino acids or ammonium sulfate) supplemented with 0.5% ammonium sulfate or with 10 mM allantoin or GABA where indicated. Download FIG S2, TIF file, 2.5 MB.

Disrupting DAL81 did not affect growth in C. albicans or C. parapsilosis during growth on arginine, ornithine, or canavanine as sole nitrogen source. C. albicans and C. parapsilosis strains were grown overnight in YPD, and increasing 1/5 dilutions were pinned on minimum media (YNB with glucose and without amino acids or ammonium sulfate), supplemented with a 10 mM concentration of arginine or ornithine or with 0.5% ammonium sulfate. Canavanine was added at a final concentration of 80 µg/ml where indicated. Plates were photographed after 72 h. Growth levels of the mutants were compared to the level seen with the relevant control for each species (C. albicans STCA2, C. parapsilosis CPRI, or C. parapsilosis 90-137). C.a, C. albicans; C.p, C. parapsilosis; *, edited strains; Δ, gene deletion strains. Download FIG S3, EPS file, 1.1 MB.

DISCUSSION

The capacity to utilize nonpreferred nitrogen sources is an important virulence factor as it allows growth and survival of pathogens in diverse host niches. For example, deleting the core regulators of nitrogen assimilation, GAT1 and GLN3, reduces virulence in C. albicans (34, 36). Here, we show that growth on poor nitrogen sources (e.g., isoleucine) triggers derepression of NCR-sensitive genes in C. parapsilosis, similarly to S. cerevisiae and other fungi (55, 56). Genes involved in transport and in amino acid metabolism are upregulated.

We found that the function of several core transcriptional regulators of the NCR is conserved in C. parapsilosis and other fungi. GAT1 and GLN3 are well studied in S. cerevisiae and C. albicans, where they encode activators of NCR-sensitive genes (4, 12–14, 34–36, 46). Deletion of either of these genes in C. albicans results in growth defects on many nitrogen-limiting media (36), which we also observed with the equivalent deletions in C. parapsilosis (Fig. 1). Expression of C. parapsilosis GAT1 was increased >60-fold during growth on isoleucine (see Data Set S1 in the supplemental material). Repression of GAT1 expression during growth on preferred nitrogen sources in other fungi has also been reported (20, 46).

In S. cerevisiae, two transcriptional repressors (GZF3 and DAL80) are required to maintain NCR as long as a preferred nitrogen source is available (1, 4, 6). There is only one ortholog (called GZF3) in Candida species, and that ortholog is poorly characterized. We recently reported that GZF3 regulates biofilm formation in C. parapsilosis, although the underlying mechanism is unknown (47). Here, we demonstrate that expression of NCR-sensitive genes (GAP2, MEP2, and GAT1) is increased when GZF3 is deleted in C. parapsilosis, showing that Gzf3 also mediates NCR in Candida species.

The function of several other specific nitrogen regulators is also conserved in C. parapsilosis. For example, GCN4 is required to utilize most amino acids as the sole nitrogen source. Deletion of PUT3 resulted in a growth defect when proline was the sole nitrogen source (Fig. 1). Put3 has recently been shown to regulate the ability of C. albicans to use proline as a nitrogen and carbon source, suggesting that its role as a nitrogen regulator is conserved across Candida and Saccharomyces species (57). However, the role of DAL81 is distinctly different (Fig. 1).

Dal81 proteins have a GAL4-like zinc finger DNA binding domain, which is found only in fungi (58). They also contain a “middle homology region” that probably has a regulatory function (59). DAL81 was first characterized in S. cerevisiae in the 1980s, when several studies showed that it was required to catabolize allantoin and urea (60–64). Talibi and Raymond (65) later discovered that Dal81, together with Uga3, binds to DNA sequences upstream of GABA-inducible genes. Uga3 and Dal81 are both required for normal GABA uptake and for expression of the GABA permease UGA4 (66). Dal81 was subsequently shown to act synergistically with other transcriptional activators driving expression of metabolic genes. For example, Dal81 and Dal82 interact at the promoters of allantoin-responsive genes (67) and, in response to external amino acids, Dal81 facilitates binding of the transcriptional activators Stp1p and Stp2p at the promoters of AGP1, BAP2, and BAP3 (31).

We found that GABA induction of gene expression in C. parapsilosis does not require DAL81, confirming that that its roles in S. cerevisiae and C. parapsilosis are very different. On the other hand, both C. parapsilosis and C. albicans require UGA3 to utilize GABA as a nitrogen source (Fig. 1). We do not know if Uga3 acts alone in Candida species, or together with an as-yet-unknown activator, to control expression of GABA metabolism.

We used transcriptomic analysis to further characterize the function of DAL81 in C. parapsilosis. During growth in rich media with complex nitrogen sources (YPD), deletion of DAL81 resulted in a major upregulation of the arginine biosynthesis pathway, in at least two different genetic backgrounds. Figure 3 shows that expression of almost every gene involved in arginine transport and biosynthesis was increased when DAL81 was absent, whereas expression of degradation enzymes was reduced. Furthermore, the expression levels of ARG1 and ARG3 were restored to wild-type levels when a single allele of DAL81 was reintroduced into the genome. DAL81 therefore regulates arginine metabolism in C. parapsilosis, both by repressing expression of arginine biosynthesis genes and by activating expression of arginine-degrading genes. The effect of deleting DAL81 was not quite as dramatic when cells were grown in minimal (YNB) media with ammonium sulfate as the sole nitrogen source, though the expression level of arginine genes was still increased. Overall, expression of arginine synthesis genes was higher in YNB than in YPD, probably because there is no arginine present in the former media, which may mask the effect of deleting DAL81 (Data Set S1).

In C. albicans, nitrogen starvation results in increased expression of DAL81, which, together with STP2, activates expression of the vacuolar transporter AVT11 (20). Ramachandra et al. (20) generated an in silico model of the regulatory networks that govern nitrogen metabolism in C. albicans which predicts involvement of Dal81 in the mobilization of nitrogen from vacuoles during nitrogen starvation. When nitrogen-starved cells were fed with arginine, expression of both DAL81 and AVT11 was decreased, suggesting that nitrogen storage and mobilization are not important under these conditions (20). The model suggests that DAL81 is an activator of GAT1 when cells are fed with arginine and that it activates STP1 and represses GLN3 when cells are fed with bovine serum albumin; both arginine and bovine serum albumin are poor nitrogen sources. Our data show that DAL81 represses, rather than activates, expression of arginine biosynthesis genes in C. parapsilosis. The roles of DAL81 in C. albicans and C. parapsilosis may be different, though this remains to be experimentally tested. In addition, we do not know if Dal81 directly regulates arginine gene expression or if it acts through another transcription factor. It is, however, clear that Dal81 does not regulate GABA or allantoin metabolism in either Candida species.

We show that there has been a dramatic change in the targets of the Dal81 transcription factor between S. cerevisiae and C. parapsilosis. Transcriptional rewiring is not uncommon in fungi, and several studies have identified divergent transcriptional circuits in S. cerevisiae and C. albicans (68–71). Significantly, many of these rewiring events affect central metabolic pathways such as carbohydrate metabolism (71, 72), ribosome biogenesis (70), and nucleotide biosynthesis (68). Lavoie et al. (73) suggested that fungi repurpose core metabolic regulators in order to adapt to different environmental niches but that their function often remains within a similar metabolic “field.” This may also be true in C. parapsilosis, where Dal81 does not regulate nitrogen assimilation from GABA and allantoin but does regulate the expression of many genes associated with arginine metabolism.

MATERIALS AND METHODS

Media and strains.

All strains were cultivated in YPD broth (Formedium; catalog no. CCM0210) or on solid YPD agar (Formedium; catalog no. CCM0110) at 30°C. For phenotype screening, minimal media (0.19% yeast nitrogen base without amino acids and ammonium sulfate [Formedium; catalog no. CYN0501], 2% glucose, with or without 2% agar) was used as the base medium. For S. cerevisiae strains, 0.08 g/liter uracil was added. For SN152-derived C. albicans strains, 0.05 g/liter arginine was added. The medium was then supplemented as indicated. For drop test plates, overnight cultures were collected by centrifugation at 13,000 rpm at room temperature for 30 to 60 s. Cells were washed by resuspension twice in 1 ml phosphate-buffered saline (PBS), and centrifugation was performed each time as described above. Washed cells were resuspended in 1 ml PBS, diluted to an A600 of 0.0625 in 1 ml PBS, and divided into aliquots and placed at 200 µl per well in the wells of a 96-well plate. Strains were then serially diluted 1:5 in 200 µl PBS to reach a final A600 of 0.0001. A 2-µl volume of each dilution was then transferred to solid media using a 48- or 96-pin replicator. Plates were incubated at 30°C and photographed at 48 and 72 h.

C. parapsilosis gene deletion strains were constructed as described by Holland et al. (47). S. cerevisiae strains 23344c and SBCY17 (74) were kindly provided by Mariana Bermúdez Moretti (Table S1). The C. albicans uga3Δ/Δ deletion strain, and the SN152 parent strain (49), were provided from the laboratory of A. D. Johnson. DAL81 was reintroduced into a C. parapsilosis dal81 deletion strain as shown in Fig. S1. The C. albicans dal81Δ mutant strain was constructed using the CRISPR method described by Vyas et al. in 2015 (75). C. albicans STCA2 (Table S1) contains CAS9 integrated at the ENO1 locus. This was transformed with pV1090-DAL81, which encodes a single guide RNA (DAL81_Guide_B) directed against DAL81 (Table S2) as well as nourseothricin resistance (NATr) and integrates at the RP10 locus. The plasmid was digested with KpnI/SacI and cotransformed with a repair template (CaDAL81-RTb; Table S2) designed to insert two premature stop codons after amino acid 199. Homozygous DAL81 mutant strains were confirmed by PCR amplification using primers ChkB-Fw and Check_Rv and by sequencing a fragment amplified with primers SeqChk_Fw and Check_Rv. DAL81 was deleted in C. parapsilosis 90-137 using a CRISPR-based method described by Lombardi et al. (54). A guide RNA targeting position (+354 bp) was introduced into plasmid pRIBO using primers CpDAL81_sgRNAa_T and CpDAL81_sgRNAa_B, generating plasmid pRIBO-DAL81a. A 5-μg volume of pRIBO-DAL81a was cotransformed with 8 to 10 μg of a linear repair template into C. parapsilosis 90-137. The repair template was generated by primer extension using 120-bp oligonucleotides containing 20-bp overlapping sequences (DAL81_RTdel_T and DAL81_RTdel_B) and consisted of 200 bp from 5′ and 3′ of the DAL81 open reading frame (ORF), flanking a 20-bp unique tag. Transformants were selected by colony PCR using primers flanking the ORF (DAL81_UPST and DAL81_DWST), and PCR products were verified by Sanger sequencing. Homology-directed repair resulted in deletion of the entire DAL81 ORF, replacing it with a 20-bp tag.

List of strains used. Download TABLE S1, DOCX file, 0.1 MB.

List of primers. Download TABLE S2, DOCX file, 0.1 MB.

RNA isolation.

Overnight cultures were washed twice in PBS and diluted to a starting A600 of ~0.3 in the desired medium. Cultures were incubated at 30°C and 250 rpm for 4 to 5 h until an A600 of approximately 1.0 to 1.5 was reached. Cells were harvested by vacuum filtration through 0.45-µm-pore-size filters and washed by subjecting filters to vortex mixing in 5 ml PBS. Finally, cells were collected again by vacuum filtration and washed from filters using 500 µl RNAlater (Qiagen; catalog no. 76104). Cells were snap-frozen in liquid nitrogen and stored at −80°C. For most RNA sequencing (RNA-seq) experiments, RNA extractions were performed using a RiboPure RNA purification kit (yeast) (Ambion; catalog no. AM1926). RNA quality was assessed using an Agilent 2100 Bioanalyzer instrument and an RNA 6000 Nano kit (Agilent; catalog no. 5067-1511). For cDNA synthesis and for analysis of cells growing with isoleucine as the sole nitrogen source, RNA was isolated using a Bioline isolate II RNA minikit (catalog no. BIO-52072). RNA concentrations were determined using a Thermo Scientific NanoDrop 2000 instrument. All purified RNA was stored at −80°C.

Quantitative real-time PCR.

PCRs were performed using two technical replicates and at least three biological replicates. Data from the technical replicates were averaged prior to statistical analysis. For cDNA synthesis, equal quantities of RNA (200 to 1,000 ng) from each sample were diluted in 4 µl nuclease-free distilled water (dH2O) and 1 µl 100 µg/ml oligo(dT)15 primer (Promega; catalog no. C1101) was added. Samples were incubated in a thermocycler at 70°C for 10 min and cooled to 4°C. Reverse transcription was performed using a Moloney murine leukemia virus (MMLV) reverse transcriptase kit (Promega; catalog no. M1701) with incubation at 37°C for 60 min and 95°C for 2 min. Each PCR mixture contained 1 µl cDNA, 7 µl dH2O, 2 µl 5 µM primer mix (Table S2), and 10 µl FastStart Universal SYBR green Master (Rox) (Roche; catalog no. 04 913 850 001). Reactions were performed in optical PCR tubes (Agilent; catalog no. 401427 and 401428) using an Agilent MX3005P QPCR system. Statistical analysis of biological replicates was carried out using the comparative threshold cycle (C) method described by Applied Biosystems.

RNA-seq.

RNA isolated from the YPD-grown C. parapsilosis CLIB214 and C. parapsilosis dal81Δ strains and from C. parapsilosis CLIB214 grown in YNB–2% glucose (with 10 mM isoleucine or 0.5% ammonium sulfate) was sequenced by BGI Global Genomics Services (100 bases; paired ends). RNA from C. parapsilosis CPRI and C. parapsilosis dal81Δ strains grown in YNB–2% glucose (with 10 mM GABA or 0.5% ammonium sulfate) was sequenced in-house. For in-house sequencing, RNA was quantified using a Qubit 2.0 Fluorometer (Life Technologies, Inc.). Library preparation was performed using an Illumina NeoPrep system and a TruSeq Stranded mRNA NeoPrep kit (Illumina; catalog no. NP-202-1001). The quality of the library was assessed using an Agilent 2200 TapeStation instrument with a high-sensitivity DNA 1000 kit (Agilent; catalog no. 5067-5584). Libraries were pooled and sequenced using an Illumina NextSeq sequencer with a NextSeq 500/550 High Output v2 kit (Illumina; catalog no. FC-404-2001) (75 cycles). All data were analyzed using established bioinformatic protocols (76). In brief, reads were mapped to the C. parapsilosis genome using TopHat version 2 (77), transcripts were counted using HTSeq (78), and differentially expressed genes were identified using DESeq2 (79). Genes with a log2FC value above 1 or below −1 and with an adjusted P value of <0.001 were retained.

Accession number(s).

All data were submitted to the Gene Expression Omnibus databases under accession number GSE109034.