Ure2 Is Involved in Nitrogen Catabolite Repression and Salt Tolerance via Ca2+ Homeostasis and Calcineurin Activation in the Yeast Hansenula polymorpha*

Disruption of HpURE2 resulted in a low expression of genes encoding nitrate-assimilatory proteins; sensitivity to Li+, Na+, and Cd2+; no induction of ENA1; low levels of the GATA-type transcription factor Gat1; and low intracellular Ca2+ levels. Gat1 levels were also very low in a Δcnb1 mutant lacking the regulatory subunit of calcineurin. The strain Δure2 was very sensitive to the calcineurin inhibitor FK506 and displayed several phenotypes reminiscent of Δcnb1. The reporter 4xCDRE-lacZ, containing calcineurin-dependent response elements in its promoter, revealed that calcineurin activation was reduced in HpΔure2. Expression of ScURE2 in Δure2 rescued nitrogen catabolite repression and Cd2+ tolerance but not those phenotypes depending on calcineurin activation, such as salt tolerance and nitrate assimilation gene derepression. HpΔure2 showed an increased expression of the gene PMR1 encoding the Golgi Ca2+-ATPase, whereas that of PMC1 encoding the vacuolar Ca2+-ATPase remained unaltered. PMR1 up-regulation was abolished by deletion of the GATA-type transcription factor GAT2 in a HpΔure2 genetic background, and normal Ca2+ levels were recovered. Moreover, overexpression of GAT2 or PMR1 yielded strains mimicking the phenotype of the HpΔure2. This suggests that the low Ca2+ levels in the HpΔure2 mutant are due to the high levels of Pmr1 that replenish the Golgi Ca2+ content, thus acting as a negative signal for Ca2+ entry into the cell. We conclude that HpUre2 is involved in salt tolerance and also in nitrate assimilation gene derepression via Ca2+ homeostasis regulation and calcineurin activation, which control the levels of Gat1.

In Saccharomyces cerevisiae, Ure2 plays a central role in nitrogen catabolite repression (NCR) 5 (i.e. the genes related to the utilization of poor nitrogen sources are repressed in the presence of preferred nitrogen sources) (1). The first insights into URE2 were obtained by Lacroute et al. (2,3), who isolated ure2 mutants incapable of carrying out NCR. They also isolated [URE3], a non-Mendelian, non-mitochondrial mutation with the same phenotype as ure2 (4,5).
[URE3] later proved to be an altered form of Ure2 inherited by prion mechanisms (6). NCR involves the localization of GATA-type transcription factors Gln3 and Gat1 outside the nucleus, in the presence of preferred nitrogen sources. This prevents the expression of genes related to the utilization of poor nitrogen sources. In this framework, Ure2 is involved in cytoplasm localization of Gln3 (1,(7)(8)(9). Several studies show the relationship between the TOR signaling pathways and NCR. Thus, in the presence of the Tor (target of rapamycin) kinase inhibitor rapamycin, both Gat1 and Gln3 are present in the nucleus, as occurs in poor nitrogen sources; in addition, Ure2 was found to be phosphorylated (7,10,11). The ⌬ure2 strain also shows an improvement in Na ϩ and Li ϩ tolerance, this being due to the induction of the Na ϩ extrusion ATPase gene, ENA1, which is positively modulated by Gat1 and Gln3 (12). Likewise, URE2 deletion suppresses the sensitivity of calcineurin mutants to Na ϩ , Li ϩ , and Mn 2ϩ , increasing their survival during treatment with mating pheromones; this depends on Gln3 and Ena1 (13). Molecular cloning of URE2 by Magasanik's group, besides confirming its role in NCR, revealed that it encodes a protein with high similarity to glutathione S-transferases (GST) (14). Despite this, such in vitro activity has not yet been detected. However, Cooper and coworkers (15) found that ⌬ure2 mutants are hypersensitive to cadmium and nickel ions and hydrogen peroxide. Nevertheless, using recombinant Ure2, glutathione peroxidase (GPx) activity with cumene hydroperoxide, hydrogen peroxide, or tert-butylhydroperoxide as substrates has been reported (16).
Hansenula polymorpha is able to use nitrate as sole nitrogen source. Nitrate enters cells via the high affinity nitrate transporter Ynt1 (17)(18)(19). It is reduced to ammonium by the consecutive action of nitrate and nitrite reductase. YNT1, YNR1, and YNI1 genes, respectively, encode the main high affinity nitrate transporter, nitrate reductase (NR) and nitrite reductase. These three genes are subjected to NCR in response to preferred nitrogen sources, such as ammonium or glutamine (19,20). Their expression is also dependent on nitrate induction mediated by two Zn(II) 2 Cys 6 transcriptional factors, Yna1 and Yna2 (20). However, the mechanisms underlying NCR in H. polymorpha are unknown. In Neurospora crassa, the protein NMR1 is involved in the negative modulation of nitrate assimilation genes and others involved in the utilization of non-preferred nitrogen sources. NMR1 interacts with the GATA factor NIT2 in the presence of preferred nitrogen sources to prevent NIT2dependent gene transcription (21,22). In Aspergillus nidulans, NmrA acts similarly to NMR1, on the GATA factor AreA (23). However, neither NMR1 nor NmrA is similar to Ure2. Moreover, gstA, a URE2 ortholog in A. nidulans, is not involved in NCR but contributes to heavy metal and xenobiotic compound tolerance (24).
Because the nitrate assimilatory pathway in the yeast H. polymorpha is known to be subject to NCR, we investigated the mechanism involved. One question was whether NCR is framed within Ure2 activity in H. polymorpha, as in S. cerevisiae, or via mechanisms closer to those reported for filamentous fungi. Once we found that Ure2 was involved in H. polymorpha NCR, we focused on its mechanisms of action because these have only been studied in depth in S. cerevisiae.
We found that Hp⌬ure2 showed sensitivity to Na ϩ and Li ϩ , in contrast to S. cerevisiae, and an unexpected drop in nitrate assimilation gene expression. We concluded that Ure2 is involved in nitrogen catabolite repression and salt tolerance via Ca 2ϩ homeostasis and calcineurin in H. polymorpha.

EXPERIMENTAL PROCEDURES
Yeast Strains and Growth Conditions-The H. polymorpha strains used in this work and their genotypes are listed in supplemental Table I. All strains are derivatives of NCYC495 leu2 ura3 strain. The wild type (WT) was obtained by transforming this strain with integrative vectors bearing HpURA3 and HpLEU2. Yeast cells were grown with shaking at 37°C in YPD medium (1% (w/v) yeast extract, 2% (w/v) peptone, 2% (w/v) glucose) or synthetic medium containing 0.17% (w/v) yeast nitrogen base, without amino acids and ammonium sulfate (Difco), 2% (w/v) glucose, and the nitrogen source indicated in each case. Synthetic medium containing 0.67% yeast nitrogen base without amino acids (Difco), 2% (w/v) glucose buffered to pH 5.5 with 50 mM MES-Tris (buffered synthetic medium) was used in some experiments. Whenever necessary, media were supplemented with 0.23 mM leucine and 0.19 mM uracil.
Nucleic Acid Isolation and Quantitative Real-time PCR-Yeast DNA and DNA were isolated as described elsewhere (25). Total RNA was extracted using the RNeasy Mini Kit TM , according to the manufacturer's instructions (Qiagen). RNA integrity was electrophoretically verified by ethidium bromide staining and A 260 /A 280 ratio. Genomic DNA was removed with 10 units of RNase-free DNase (Roche Applied Science) for each 10 g of RNA. DNase treatment efficiency was checked by the absence of PCR products using HpACT1 as a template. Total RNA (1 g) was reverse transcribed using the commercial TaqMan kit according to the manufacturer's instructions (Applied Biosystems). Quantitative RT-PCR (qRT-PCR) was carried out in the iCycler iQ real-time PCR detection system (Bio-Rad) using the Fast Start SYBR Green Master master mix (Roche Applied Science). Four serial dilutions of cDNA, 3 l of each, were amplified in triplicate for each amplicon in a volume of 20 l. The relative changes in gene expression from qRT-PCR experiments were analyzed as described (26). HpACT1 was used as a reference gene.
Gene Disruption and Yeast Vectors-URE2, CNB1, GAT1, GAT2, and PMC1 disruption is described in the supplemental material. All of the primer sets for PCR-mediated gene disruption, tagging, or qRT-PCR are described in supplemental Table  II. All vectors are described in supplemental Table III. nURE2, nGAT2, and nPMR1-Strains bearing several copies of URE2 (nURE2) GAT2 (nGAT2), or PMR1 (nPMR1) were obtained by multiple integration of vectors pGEM-URE2-URA3, pGEM-GAT2-URA3, and pGEM-PMR1-LEU2. These contain the genes URE2, GAT2, and PMR1 and were linearized at the URA3 marker gene with BglII or at the LEU2 marker gene with BstII to facilitate target integration.
lacZ Gene Fusions-The pHPI 359 vector (27) was used to fuse ENA1 and PMR1 gene promoters to lacZ, yielding pENA1-lacZ and pPMR1-lacZ. The regions from Ϫ1164 to ϩ39 relative to the ATG of ENA1 and Ϫ1009 to ϩ39 of the PMR1 were amplified by PCR using Pfu from genomic DNA using the primers ENA1-lacZ-F and ENA1-lacZ-R for ENA1 and using PMR1-lacZ-F and PMR1-lacZ-R for PMR1. The vector pAMS367 bearing 4xCDRE-lacZ (kindly provided by M. Cyert, Stanford University) was modified at the StuI site by introducing the H. polymorpha LEU2 marker.
Growth Tests-The sensitivity of different yeast strains to cations and other compounds was assayed by a drop test. Strains were grown in YPD liquid medium, and then cultures were diluted to obtain 10 6 to 10 2 cells in 5 l and then spotted on solid medium.
Epitope Tagging of Ure2, Gat1, and Gat2-Ure2, Gat1, and Gat2 were tagged at their C termini with six copies of the peptide epitope from the HA protein of human influenza virus using the vector pHA1. This vector derives from pANL31 (28), which contains the eGFP (green fluorescence protein) ORF and the zeocin resistance marker (ble). The eGFP sequence was replaced by a 256-bp fragment containing six copies of the HA epitope obtained by PCR using the pair of oligonucleotides tagF and tagR from the S. cerevisiae vector pYM3 (29). The ϩ1 to ϩ905 DNA region from URE2 was amplified by PCR using the primers Ure2exp-F and Ure2-HAtag-R, the ϩ401 to ϩ1523 DNA region from GAT1 was amplified using the primers GAT1GFP-F and GAT1GFP-R, and the ϩ46 to ϩ1278 DNA region from GAT2 was amplified using the primers GAT2-HA-F and GAT2-HA-R. All sequences were cloned in frame with the 6HA sequence to render the vectors pHA-URE2, pHA-GAT1, and pHA-GAT2. These were linearized at the BclI, NarI, and NruI sites, respectively, to facilitate their homologous integration and used to transform yeast. Transformants were selected by growth on YPD plates containing 100 g/ml Zeocin TM (Invitrogen). HA tagging was confirmed by PCR with primers designed to bind outside the construct (C-URE2-F, C-GAT1-F, and RT-DAL80-F) and at the HA epitope (tagR). Western blot analysis using anti-HA antibody (Roche Applied Science) provided further confirmation.
Measurement of Ca 2ϩ Intracellular Content-Ca 2ϩ content of the cells was determined as described previously (30). Briefly, samples of cells were filtered, washed twice with ice-cold 100 mM MgCl 2 , and extracted with 100 mM HCl. Ca 2ϩ was determined by atomic absorption spectrophotometry.
DNA Sequencing-DNA sequencing was performed using Amplytaq polymerase with a BigDye Terminator version 3.1 cycle sequencing kit on an automated DNA sequencer (ABI PRISM 3100 Genetic Analyzer).
Miscellaneous Methods-Yeast cells were electrotransformed as described previously (31). ␤-Galactosidase activity was determined as in Ref. 27. Determinations of lithium influx and efflux were previously described (32).

Molecular Cloning and Sequence Analysis of HpURE2-An
HpURE2 partial DNA sequence is present in the H. polymorpha genome data base (Genolèvures) (33). Based on this sequence, a 462-bp DNA fragment was obtained by PCR and used to screen an H. polymorpha EMBL3 genomic library (34). Several phages were isolated, and DNA sequencing of the phage 12 revealed the presence of the whole HpURE2 ORF, along with other ORFs. This contained 906 bp encoding a protein of 34.3 kDa. The sequence of HpURE2 has been deposited in GenBank TM under accession number AJ698949.
BLASTp analysis revealed the highest identity of HpUre2, about 70%, with Ure2 of different yeasts, such as Debaryomyces hansenii, Pichia stipitis, Candida maltosa, and Candida albicans. The identity with ScUre2 decreases to 63%. Identity was also found with different members of the GST superfamily. For more details, see the supplemental material.

HpUre2 Is Responsible for Nitrogen Catabolite Repression
and Derepression-To test the role of HpURE2 in NCR, a ⌬ure2 null mutant strain was obtained. NR activity, induced by nitrate and repressed by preferred nitrogen sources, was used as a readout of NCR. In nitrate plus preferred nitrogen sources, such as ammonium, NR was higher in ⌬ure2 than WT (Table 1), indicating that NCR was almost abolished in ⌬ure2. Indeed, qRT-PCR showed that in nitrate plus ammonium, nitrate assimilation gene expression was higher in ⌬ure2 than WT (Table 2). However, it was very striking that in nitrate, ⌬ure2 presented lower NR activity than WT and also lower nitrate assimilation gene expression, indicating that these genes are not fully derepressed in the absence of URE2.
To gain further insight into HpUre2 regulation, we studied its phosphorylation state in response to rapamycin, to which Hp⌬ure2 ( Fig. 1), like Sc⌬ure2, was very sensitive (35), and to nitrogen sources. As shown in Fig. 1, Ure2 was phosphorylated in response to preferred nitrogen sources, mainly in glutamine, and became dephosphorylated in synthetic medium without any nitrogen source or proline. Rapamycin also triggered Ure2 dephosphorylation.
These results allow us to conclude that URE2 is involved in NCR in H. polymorpha, although the lower expression of nitrate assimilation genes YNT1, YNR1, and YNI1 in ⌬ure2 also indicated a positive role of Ure2. Moreover, the HpUre2 phosphorylation state depends on the nitrogen source, most likely via the TOR signaling pathway.
Ure2 Is Involved in Na ϩ /Li ϩ Tolerance-In S. cerevisiae, Ure2 is involved in salt tolerance (12,13), which prompted us to

TABLE 2 Analysis of nitrate assimilation gene expression in WT and ⌬ure2 strains
Ammonium-grown cells of the WT and ⌬ure2 were washed and incubated in synthetic medium containing 5 mM ammonium, 5 mM nitrate, or 5 mM ammonium plus 2.5 mM nitrate. Relative expression was determined by qRT-PCR. The experiments, only one of which is shown, were repeated three times without significant differences. ND, not determined.

Nitrate assimilation gene expression
The expression is normalized to the WT in the same medium. b The expression is normalized to the expression in ammonium. HpACT1 was used as reference gene. NOVEMBER 26, 2010 • VOLUME 285 • NUMBER 48 study the role of HpUre2 in Na ϩ /Li ϩ tolerance. For this, ⌬ure2 was tested for growth in media containing Na ϩ or Li ϩ and was found to be more sensitive to these cations than WT (Fig. 2). This contrasts with S. cerevisiae, where ⌬ure2 shows greater Na ϩ and Li ϩ tolerance than WT (12). Cation influx indicated that ⌬ure2 accumulated Li ϩ faster than WT. Likewise, the efflux kinetics showed that ⌬ure2 was unable to extrude Li ϩ as WT did (Fig. 3). Because ENA1 encodes the main ATPase involved in Na ϩ /Li ϩ extrusion, we determined its expression levels in WT and ⌬ure2 by qRT-PCR. Na ϩ and Li ϩ induced the level of ENA1 in WT between 10-and 13-fold; in contrast, no induction was observed in ⌬ure2 (Fig. 4). These results reinforced our idea that efflux of these cations was impaired in ⌬ure2. The same results were found using a strain bearing the ENA1-lacZ construct (data not shown). Our findings allow us to conclude that Ure2 is positively involved in Na ϩ /Li ϩ tolerance, up-regulating ENA1 expression. Ure2 Participates in the Calcineurin-dependent Response-In S. cerevisiae, it has been shown that ENA1 expression is subject to a complex regulatory network, where Gln3, Gat1, and calcineurin are the positive signals (12,36,37). However, in Hp⌬ure2, the expected increase in ENA1 expression and concomitant cation tolerance was not observed. This suggested that GATA factors and/or calcineurin-dependent ENA1 expression could be negatively affected in ⌬ure2. Therefore, we tested calcineurin involvement using its inhibitor FK506 (38). We found that ⌬ure2 was very sensitive to this macrolide, even in the absence of Na ϩ or Li ϩ , which further increased ⌬ure2 sensitivity to FK506 (Fig. 5). This suggested a positive effect of Ure2 on calcineurin activation. In this framework, we observed a strong parallelism between ⌬cnb1, lacking the regulatory subunit of calcineurin, and ⌬ure2, in response to Mn 2ϩ , Na ϩ , and SDS (Fig. 6). ⌬ure2⌬cnb1 showed a phenotype very close to ⌬cnb1, although slightly more sensitive, indicating that Ure2 could even act beyond Cnb1 (Fig. 6). We also found that ENA1 was scarcely expressed in response to Na ϩ in ⌬cnb1, as in ⌬ure2 (data not shown). These results strongly suggested that calcineurin activation in ⌬ure2 was negatively affected.

URE2 Deletion Reduces Calcineurin-dependent Gene Expression and Total Cell Ca 2ϩ Content-
The effect of Ure2 on calcineurin was measured by 4xCDRE-lacZ reporter (39) in WT and ⌬ure2 strains. ⌬ure2 showed about a 2.5-fold lower induction of 4xCDRE-lacZ, in response to Na ϩ , indicating poor activation of calcineurin-dependent gene expression in ⌬ure2 (Fig. 7). Because calcineurin activation is Ca 2ϩ -dependent, we studied the role of Ca 2ϩ in ⌬ure2 phenotypes. The addition of 50 mM Ca 2ϩ restored ⌬ure2 tolerance to Li ϩ , whereas 6 mM EGTA increased ⌬ure2 sensitivity to it (Fig. 8). We also observed that extra Ca 2ϩ produced a higher ENA1-lacZ expression in WT than in ⌬ure2 (Fig. 9).  5-7). Cells incubated for 120 min in synthetic medium plus 5 mM ammonium or 1 mM proline are also shown (lanes 8 and 9). B, the Ure2 low mobility shift band is due to phosphorylation. Protein extracts from cells deprived of nitrogen or incubated in glutamine were treated with -protein phosphatase. 50 mM EDTA was added to inhibitprotein phosphatase. C, rapamycin causes Ure2-6HA dephosphorylation. Protein extracts from Ure2-6HA strain were incubated for 120 min in 5 mM glutamine with (ϩ) or without (Ϫ) 0.5 g/ml rapamycin or deprived of nitrogen. D, rapamycin growth sensitivity. Serial 10-fold dilutions of the WT and ⌬ure2 were spotted on solid medium containing YPD plus rapamycin at the indicated concentration. FIGURE 2. ⌬ure2 is sensitive to Li ؉ and Na ؉ . WT, ⌬ure2, ⌬ure2URE2, and nURE2 strains were grown in YPD. Serial 10-fold dilutions were spotted on pH 5.5 buffered synthetic medium plus LiCl and NaCl at the indicated concentrations. Cells were incubated at 37°C for 2 days. Furthermore, total cell Ca 2ϩ content was lower in ⌬ure2 with respect to WT (Fig. 10). Altogether, we conclude that Ure2 is involved in Ca 2ϩ homeostasis and consequently in calcineurin activation.      . Effect of Ca 2؉ on ENA1 gene expression in WT and ⌬ure2 strains. ENA1 expression was followed by assaying ␤-galactosidase activity in WT and ⌬ure2 bearing ENA1-lacZ. Cells were grown in synthetic medium plus 5 mM ammonium. ENA1 expression was determined in WT (F) and ⌬ure2 (E) incubated in the same medium plus 0.7 M NaCl, and in 0.7 M NaCl plus 50 mM CaCl 2 in WT () and ⌬ure2 (‚). Experiments, only one of which is shown, were carried out three times without significant differences. NOVEMBER 26, 2010 • VOLUME 285 • NUMBER 48

JOURNAL OF BIOLOGICAL CHEMISTRY 37555
To gain further insights into calcineurin activation in ⌬ure2, we studied the activation of vacuolar H ϩ /Ca 2ϩ exchanger Vcx1. In S. cerevisiae, calcineurin inhibits Vcx1dependent H ϩ /Ca 2ϩ exchange. Calcineurin decreases Ca 2ϩ tolerance of pmc1 mutants by inhibiting the function of Vcx1; this Ca 2ϩ tolerance is restored by inactivation of calcineurin (40). To test the activation of Vcx1 in ⌬ure2, we analyzed Ca 2ϩ sensitivity of ⌬pmc1, ⌬cnb1⌬pmc1, and ⌬ure2⌬pmc1. As shown in Fig. 11, ⌬pmc1 was very sensitive, whereas ⌬cnb1⌬pmc1 and ⌬ure2⌬pmc1 were more resistant. As expected, ⌬cnb1⌬pmc1 was much more tolerant to Ca 2ϩ than ⌬ure2⌬pmc1 because the latter conserves some calcineurin activity. These results also show the low activation of calcineurin in ⌬ure2.
Gat1 Levels Are Regulated by the Calcineurin Pathway, Being Lower in ⌬ure2-Because ENA1 and nitrate assimilation genes YNT1, YNR1, and YNI1 are down-regulated in ⌬ure2 and positively regulated by Gat1, we explored the levels of HpGat1 in ⌬cnb1 and ⌬ure2. HpGat1 is a positive GATA factor also involved in nitrate assimilation gene expression in H. polymorpha (supplemental Figs. S1 and S2). Accordingly with S. cerevisiae Ure2 mechanisms, HpGat1 would enter the nucleus in the Hp⌬ure2 strain, up-regulating ENA1 expression. The strain ⌬gat1 showed Li ϩ and Na ϩ sensitivity, although less than in ⌬ure2, and low levels of ENA1 expression (supplemental Figs. S3 and S4). We found that the Gat1 levels in ⌬ure2 were very low (Fig. 12), in contrast to those found in S. cerevisiae (41)(42)(43)(44). Therefore, HpURE2 deletion did not increase ENA1 activation via Gat1 but clearly lowered it. This also explains the lower derepression of the nitrate assimilation genes YNT1, YNR1, and YNI1 in ⌬ure2.
The close correlation between ⌬ure2 and ⌬cnb1 found throughout this work led us to ask whether Gat1 levels were calcineurin-dependent. We found that ⌬cnb1 showed even lower Gat1 levels than ⌬ure2 (Fig. 12), allowing us to report for the first time in yeast that levels of a GATA transcription factor were controlled by the calcineurin signaling pathway. In accordance with this, the GAT1 gene presents a putative calcineurin-dependent response element (CDRE) (45) in its 5Ј-non-coding region. Therefore, the lower levels of Gat1 present in ⌬ure2 are consistent with the low calcineurin activation in this strain compared with WT.
To clarify if calcineurin acts on ENA1 exclusively via Gat1, we measured the levels of ENA1 expression in ⌬gat1 with or without additional 50 mM Ca 2ϩ . An increase of ENA1 expression was observed in response to Ca 2ϩ (data not shown), suggesting that ENA1 induction is under the dual control of calcineurin-Crz1-Gat1 and calcineurin-Crz1. This also explains the higher sensitivity of ⌬ure2 than ⌬gat1 to Na ϩ and Li ϩ .
Expression of ScURE2 in Hp⌬ure2 Does Not Rescue Calcineurin Activation-Both ScUre2 and HpUre2 present high similarity to GSTs. ScUre2 shows GPx activity, but in vitro GST activity has not been reported. However, deletion of ScURE2 causes increased sensitivity to heavy metal ions, such as Cd 2ϩ and Ni 2ϩ (15). To show whether these potential enzymatic activities of HpUre2 are responsible for Ca 2ϩ homeostasis and calcineurin regulation, we tested whether ScURE2 expression was able to complement Hp⌬ure2 phenotypes. Hp⌬ure2ScURE2 transformants almost fully recovered Cd 2ϩ tolerance, indicating that GST and GPx activities associated with ScUre2 are active in H. polymorpha. NR activity in Hp⌬ure2ScURE2 was lower than Hp⌬ure2 in nitrate plus ammonium (supplemental Table IV). Consistent with this, Hp⌬ure2ScURE2 fully recovered tolerance to chlorate (Fig. 13), the chlorine analog of nitrate, which is reduced by NR to the toxic chlorite (46), indicating that ScUre2 was able to   . URE2 and CNB1 deletion decreases Gat1 levels. Gat1-6HA levels were analyzed by SDS-PAGE in WT, ⌬ure2, and ⌬cnb1. Cells were grown to early exponential phase in synthetic medium plus 5 mM ammonium, washed, and incubated for 120 min in the same medium plus 5 mM glutamine (Gln) (lanes 1 and 2), 5 mM ammonium (NH 4 ϩ ) (lanes 3 and 4), or 5 mM nitrate (NO 3 Ϫ ) (lanes 5 and 6). In A, 10 g of protein from WT and 40 g from ⌬cnb1 were analyzed; in B, 20 g of protein from WT and ⌬ure2 were analyzed. exert NCR in H. polymorpha. However, ScURE2 expression did not rescue either Li ϩ tolerance (Fig. 13) or NR activity levels (supplemental Table IV). These levels were lower in Hp⌬ure2ScURE2 than in WT, in both nitrate and nitrate plus ammonium (supplemental Table IV). In nitrate, NR activity is consistently lower in ScURE2Hp⌬ure2 than in Hp⌬ure2; this could be due to the low levels of Gat1 in this strain and also to the capacity of ScUre2 to bind Gat1 even in the presence of a non-preferred nitrogen source like nitrate (supplemental Table  IV). Therefore, these results indicate that ScUre2 was unable to activate calcineurin as HpUre2 does. The sensitivity of Hp⌬ure2ScURE2, like that of ⌬ure2 to FK506 is consistent with this (Fig. 14). PMR1 Expression Is Regulated by Ure2 through Gat2-To maintain cytosolic Ca 2ϩ homeostasis in S. cerevisiae, calcineurin regulates the expression of PMC1 and PMR1, encoding P-type Ca 2ϩ -ATPases involved in Ca 2ϩ transport into the vacuole and ER-Golgi, respectively (47,48). Analysis of PMR1 and PMC1 gene expression in H. polymorpha revealed that PMR1-lacZ levels were higher in ⌬ure2 than WT, whereas PMC1 expression remains unaltered in ⌬ure2 (Fig. 15). Moreover, adding extra Ca 2ϩ produced higher induction in WT than in ⌬ure2 (Fig. 15). This result suggests that the lower levels of intracellular Ca 2ϩ observed in ⌬ure2 are due to the constitutively high PMR1 expression because this would increase ER-Golgi Ca 2ϩ replenishment, and as a result, Ca 2ϩ entry into cells would decrease. In S. cerevisiae, the ⌬pmr1 mutant considerably increases Ca 2ϩ entry into cells via a capacitative Ca 2ϩ entry, similar to that found in mammals (49 -51). In light of the current S. cerevisiae Ure2 mode of action, in Hp⌬ure2 a GATA factor would enter the nucleus, activating PMR1. Analysis of PMR1 expression in strains lacking the positive GATA factors HpGAT1, HpGLN3, and HpGAT2 revealed low PMR1 expression in ⌬gat2, even in the presence of Ca 2ϩ (Fig. 15). HpGat2 presents a close sequence similarity with ScGat1 and HpGat1 (supplemental Figs. S1 and S2). Nevertheless, its deletion has no effect on Li ϩ and Na ϩ sensitivity, unlike Gat1 (supplemental Fig. S3). In accordance with the observed role of Gat2 in regulating PMR1, a strain bearing multicopy GAT2 (nGAT2) showed high PMR1 expression levels, equal to those seen in ⌬ure2. On the other hand, GAT2 deletion to obtain ⌬ure2⌬gat2 restored PMR1 expression to those levels observed in WT (Fig. 15). We also determined calcineurin-dependent gene expression by 4xCDRE-lacZ in ⌬gat2, ⌬ure2⌬gat2, and nGAT2; in the latter, this expression was the same as in ⌬ure2, whereas deletion of GAT2 led to higher levels of expression (Fig. 7). Furthermore, overexpression of PMR1 by increasing gene dosage reproduces ⌬ure2 phenotypes (Fig. 16). We concluded that Ure2 is involved in Ca 2ϩ homeostasis via Gat2, which is responsible for PMR1 regulation.
Gat2 Levels Are Regulated by Calcineurin and Gat1-Because Gat2 plays a key role downstream from Ure2, we determined Gat2 levels in ⌬cnb1 and ⌬ure2. ⌬cnb1 showed very low levels of Gat2, almost undetectable by Western blot, whereas in FIGURE 13. ScURE2 expression in Hp⌬ure2 rescues Cd 2؉ and chlorate tolerance but not Li ؉ tolerance. WT, ⌬ure2, and Hp⌬ure2ScURE2 strains were grown in YPD. Serial 10-fold dilutions were spotted on synthetic medium containing 5 mM ammonium and 1 mM nitrate plus CdCl 2 , LiCl, and KClO 3 at the indicated concentrations. Cells were incubated at 37°C for 2 days.  Cells were grown in synthetic medium plus 5 mM ammonium and then incubated in the same medium plus 5 mM CaCl 2 for 2 h. Experiments, only one of which is shown, were repeated three times without significant differences. ⌬ure2 they were higher than WT (Fig. 17). Because Gat1 levels were lower in ⌬cnb1 than WT, we studied whether Gat2 was under the control of Gat1. Indeed in ⌬gat1, Gat2 levels were about 50% of those in WT (Fig. 17). The high levels of Gat2 in ⌬ure2 are consistent with our results because we report that PMR1 expression, which is regulated by GAT2, is higher in this strain. These high GATA factor levels are apparently contradictory with the down-regulation of calcineurin in ⌬ure2. However, because Ure2 is absent, Gat1 can enter the nucleus freely to activate GAT2 transcription. The results here reported are summarized in the HpUre2 working model (Fig. 18).

DISCUSSION
ScUre2 plays a central role in the regulation of nitrogen metabolism (1,11,35,52) and to a lesser extent in salt tolerance (12,13,37). HpURE2 has been shown to be a highly pleiotropic gene because its deletion resulted in an assortment of different phenotypes. Some of the observable characteristics of ⌬ure2 are (i) rapamycin sensitivity; (ii) decreased tolerance to Na ϩ , Li ϩ , Mn 2ϩ , and SDS; (iii) pronounced FK506 sensitivity; and (iv) reduction of nitrate assimilation gene expression (Figs. 1, 2, 5, and 6 and Table 2). These traits reveal similarities but also important differences between ScUre2 and HpUre2. The main structural difference between them is the absence of the Q/Nrich N-terminal region involved in prion-like behavior (54,55). We have not yet investigated whether HpUre2 presents such behavior, either in H. polymorpha itself or using S. cerevisiae as the host. However, analysis of ⌬ure2 revealed a clear involvement of HpUre2 in NCR (Tables 1 and 2), which suggests that the above N-terminal region seems uninvolved in NCR. This is consistent with what was found in S. cerevisiae (55), although new insights into NCR suggest this prion domain contributes to Ure2 stability and functioning in this process. Actually, this domain is required for interaction with the GATA factor Gzf3 (56).
Besides HpUre2, we have also identified two positively acting GATA factors (HpGat1 and HpGat2 (see supplemental material)) and one with a negative role (HpGzf3), all showing significant identity with the S. cerevisiae GATA factors. 6 Altogether, these findings indicate that the overall nitrogen regulatory system in the nitrate-assimilating yeast H. polymorpha is closer to S. cerevisiae than to the filamentous fungi A. nidulans and N. crassa, which are also able to use nitrate as the sole nitrogen source (57,58). We also found that HpUre2 undergoes phosphorylation in response to preferred nitrogen sources and is dephosphorylated under nitrogen limitation conditions and by rapamycin (Fig. 1). This confirms our previous observation that in H. polymorpha, the TOR signaling pathway responds to nitrogen sources and regulates nitrogen assimilation gene expression (59). The phosphorylation state of Ure2 could play an important role in modulating its interaction with the GATA factors.
Another remarkable difference between HpUre2 and ScUre2 is their role in the response to Na ϩ and Li ϩ stress; whereas Hp⌬ure2 presents sensitivity to these cations (Fig. 2), Sc⌬ure2 is resistant. Greater Li ϩ accumulation and low ENA1 expression account for the Na ϩ /Li ϩ sensitivity of Hp⌬ure2 (Figs. 3 and 4). As occurs in S. cerevisiae, both calcineurin and Gat1 regulate ENA1 positively because strains lacking Gat1 (supplemental Fig. S4) and Cnb1 (data not shown) showed low levels of ENA1 expression. Two new findings open novel perspectives on this interesting protein. First, Ure2 clearly regulates Ca 2ϩ cell content via Gat2, which acts transcriptionally on PMR1 (Fig. 15). Second, this process is involved in activating the calcineurin pathway. Even more important, Gat1 levels were clearly regulated by calcineurin (Fig. 12). Indeed, because it is widely assumed that Ca 2ϩ activates calcineurin in response to different stimuli (60, 61), a poor activation of calcineurin was to be expected in ⌬ure2. Accordingly, a strong parallelism in the behavior of ⌬cnb1 and ⌬ure2 was seen: (i) the two strains were  . CNB1 and GAT1 deletion decreases Gat2 levels, whereas URE2 deletion increases them. Gat2-6HA levels in WT, ⌬ure2, ⌬cnb1, and ⌬gat1 were analyzed by SDS-PAGE. Cells grown to early exponential phase in synthetic medium plus 5 mM ammonium were washed and incubated for 120 min in synthetic medium plus 5 mM glutamine (Gln), 5 mM ammonium (NH 4 ϩ ), or 5 mM nitrate (NO 3 Ϫ ). 50 g of protein were analyzed.
FIGURE 18. Working model for role of Ure2 in Ca 2؉ homeostasis, nitrate assimilation, and ENA1 gene expression. Ure2 regulates Ca 2ϩ cell content via Gat2, which acts transcriptionally on PMR1. Pmr1 regulates Ca 2ϩ levels in ER-Golgi, which in turn act on total Ca 2ϩ content in the cell, modulating calcineurin activation. This induces ENA1 and nitrate assimilation gene up-regulation via Crz1 3 Gat1. ENA1 is also induced via Crz1 directly. Gat2 is transcriptionally activated by Gat1. Lines with arrowheads indicate positive events, whereas lines with bars are inhibitory. Although not shown here, the model assumes that Ure2 also retains Gat1 in the cytosol in the presence of preferred nitrogen sources.
sensitive to Mn 2ϩ , Na ϩ , and SDS; (ii) ⌬ure2⌬cnb1 showed a phenotype very close to ⌬cnb1 (Fig. 6); (iii) ENA1 expression levels were identically low in both strains (data not shown); (iv) Gat1 levels were very low in both (Fig. 12); (v) ⌬ure2 was remarkably sensitive to FK506 (Fig. 5); and (vi) the lower calcineurin activation in ⌬ure2 was shown using 4xCDRE-lacZ (Fig. 7). The slight increase in ⌬cnb1⌬ure2 sensitivity with respect to ⌬cnb1 could be due to the role of Ure2 in detoxification processes (15). These findings allow us to understand the scarce induction of ENA1 in ⌬ure2 in response to Na ϩ /Li ϩ . We are also aware that regulation of Gat1 by calcineurin has a special significance for nitrate assimilation gene derepression in H. polymorpha. We emphasize that our results allow us to conclude that NCR is almost abolished in ⌬ure2. In this regard, HpUre2 seems to operate as in S. cerevisiae, retaining GATA factors outside the nucleus when the medium contains preferred nitrogen sources (1,(7)(8)(9). However, in ⌬ure2, nitrate assimilation gene derepression in nitrate is negatively affected. Now we know that this is due to the low levels of Gat1 present in ⌬ure2.
Expression of ScURE2 in Hp⌬ure2 also revealed that HpUre2 displays a more complex regulatory network than ScUre2. Thus, ScUre2 rescued Cd 2ϩ tolerance (Fig. 13) and NCR in Hp⌬ure2 ( Fig. 13 and supplemental Table IV). This rescue is associated with the GST and GPx activities of ScUre2, respectively, and its capacity to retain HpGat1 outside the nucleus in the presence of preferred nitrogen sources, such as ammonium. In contrast, Li ϩ tolerance and NR activity were not complemented, indicating the incapacity of ScURE2 to activate the calcineurin in Hp⌬ure2 (Fig. 13 and supplemental Table IV). These results are consistent with our model of the action mechanisms of HpUre2 (i.e. its capacity to interact directly with Gat1 but also indirectly acting on its levels via Ca 2ϩ homeostasis and calcineurin). Moreover, ScURE2 expressed in H. polymorpha seems to interact only with HpGat1, whereas HpUre2 interacts with HpGat1 and HpGat2, which is involved in PMR1 up-regulation to maintain Ca 2ϩ homeostasis (Fig. 15).
Calcineurin has been widely reported to be associated with GATA factors in mammalian cells, playing an important role in muscle regeneration and hypertrophy in association with NFATc1 and GATA-2 (62,63). Blockage of calcineurin downregulates GATA-6-DNA binding in differentiated vascular smooth muscle cells (64). The hypertrophic effects of calcineurin in cardiomyocytes have been linked with its interaction with GATA-4 transcription factor (65).
Regarding lower levels of Ca 2ϩ in ⌬ure2 than in WT, we observed that a lack of Ure2 led to Ca 2ϩ -ATPase PMR1 gene up-regulation (Fig. 15), presumably as a consequence of a GATA factor being freed from Ure2 and entering the nucleus. Once the GATA factor is in the nucleus, it can act directly on PMR1 or on a second GATA factor affecting PMR1. We therefore propose that PMR1 is regulated by Ure2 through Gat2 in H. polymorpha, because (i) GAT2 deletion led to PMR1 downregulation (Fig. 15), (ii) its overexpression led to PMR1 up-regulation (Fig. 15), and (iii) ⌬ure2 presents high levels of Gat2 (Fig. 17). Because GAT2 is down-regulated in ⌬gat1 and ⌬cnb1 (Fig. 17), activation of GAT2 expression by Gat1 cannot be ruled out. In S. cerevisiae, the ER-Golgi Ca 2ϩ store has been shown to induce a signal that modulates extracellular Ca 2ϩ entry (49 -51), depending in turn on Ca 2ϩ -ATPase Pmr1 levels. We consider that the up-regulation of PMR1 via Gat2 in ⌬ure2 leads to a decrease in Ca 2ϩ entry, resulting in a reduced cytosolic Ca 2ϩ store. Deletion of GAT2 from the ⌬ure2 genetic background abolished PMR1 up-regulation (Fig. 15), and in accordance with our hypothesis, Ca 2ϩ content was restored to WT levels (Fig. 10). Furthermore, the activation of Vcx1 in ⌬ure2 could also act concertedly or sequentially along with PMR1 to reduce the cytoplasmic Ca 2ϩ available for calcineurin activation.
The Pmr1 Ca 2ϩ /Mn 2ϩ -ATPase negatively regulates the rapamycin-sensitive TOR complex (TORC1) in S. cerevisiae, Mn 2ϩ in the Golgi being involved in TORC1 signaling inhibition (66). However, our results are consistent with the role of Pmr1 in Ca 2ϩ transport. In agreement with this, Ca 2ϩ addition produced a lower ENA1-lacZ induction in ⌬ure2 than in WT (Fig. 9), as would be expected for a strain where Ca 2ϩ homeostasis is jeopardized.
We have elucidated the mechanisms underlying Ca 2ϩ homeostasis, Na ϩ /Li ϩ tolerance, and nitrate assimilation gene derepression involving Ure2 in H. polymorpha, which provide new insights into the role of this protein. Whether or not the same mechanisms exist in other yeasts is an intriguing question. So far, how Ure2 acts in nitrogen regulation and facilitates Na ϩ / Li ϩ tolerance has only been characterized in S. cerevisiae. Indeed, the function of ScUre2 in NCR is fully complemented by the Ure2p of different Candida and Saccharomyces yeast species (53). In contrast, we report here that ScURE2 expressed in Hp⌬ure2 complements NCR but not those phenotypes where calcineurin is involved, suggesting that the mechanisms so far elucidated for ScUre2 are not universal.
In summary, we uncover the central role of Ure2 in Ca 2ϩ homeostasis and its implication in calcineurin pathway activation. The GATA factor Gat1 is also shown to be regulated by calcineurin.