Genetic analysis of suppressor mutants of a pho 84 disruptant in the search for genes involved in intracellular inorganic phosphate sensing in Saccharomyces cerevisiae

To achieve inorganic phosphate (Pi) homeostasis, cells must be able to sense intracellular and extracellular Pi concentrations. In the Pi signaling (PHO) pathway in Saccharomyces cerevisiae, high Pi represses genes involved in Pi uptake (e.g., PHO84) and Pi utilization (PHO5); conversely, the cyclin-dependent kinase inhibitor Pho81 inhibits the activity of the Pho80-Pho85 cyclin-cyclin dependent kinase complex in low-Pi conditions, leading to induction of these genes. However, how yeast senses Pi availability remains unresolved. To identify factors involved in Pi sensing upstream of the Pho81-Pho80-Pho85 complex, we generated and screened suppressor mutants of a Δpho84 strain that shows constitutive PHO5 expression. By a series of genetic tests, including dominance–recessiveness, complementation and tetrad analyses, three sef (suppressor of pho84 [pho eighty-four]) mutants (sef8, sef9 and sef10) were shown to contain a novel single mutation. The sef mutants suppressed the phenotype of constitutive PHO5 expression at the transcriptional level, but did not show restored Pi uptake capacity. An epistasis– hypostasis test revealed that the sef mutations were hypostatic to pho80 mutation, indicating that their gene products function upstream of the Pho81-Pho80-Pho85 complex in the PHO pathway. The sef mutations identified are associated with gene(s) that may be involved in the homeostasis of an intracellular Pi level-sensing mechanism in S. cerevisiae.


INTRODUCTION
Cells must cope with environmental changes including alterations in nutrients, temperature and osmotic pressure.A system that senses environmental changes and provides an adequate response is important for cell survival.The phosphate signaling pathway (PHO pathway) is involved in the sensing and response system that achieves phosphate homeostasis in Saccharomyces cerevisiae, and is one of the well-studied models of a signal transduction pathway.As a constituent of nucleic acids and high-energy compounds such as ATP, phosphate is indispensable for numerous cellular functions.Therefore, phosphate homeostasis is essential for cells.
When cells are grown in a low inorganic phosphate (Pi) environment, the expression of secreted acid phosphatases (Pho5, Pho11 and Pho12) and high-affinity phosphate transporters (Pho84 and Pho89) is highly induced so that phosphate is liberated from extracellular compounds and imported into the cell (Ljungdahl and Daignan-Fornier, 2012).Conversely, when cells are surrounded by a high concentration of Pi, transcription of these genes is strongly repressed.In high-Pi conditions, the cyclin-cyclin dependent kinase (CDK) complex Pho80-Pho85 phosphorylates the transcription factor Pho4, which triggers its export from the nucleus and, as a result, transcriptional activation of Pho4 target genes does not occur (Fig. 1).When the Pi concentration is low, by contrast, the CDK inhibitor Pho81 inhibits the kinase activity of Pho85, allowing the non-phosphorylated form of Pho4 to accumulate inside the nucleus and activate the transcription of Pho4 target genes together with the transcription factor Pho2 (Conrad et al., 2014).
Although components of the PHO pathway have been characterized, how yeast cells sense environmental Pi conditions is a fundamental issue that remains unresolved.Yeast cells sense both extracellular and intracellular Pi concentration.The low-affinity phosphate transporters Pho87 and Pho90 have a hydrophilic motif at the amino terminus that is similar to the carboxy-terminal motif found in the glucose sensors Snf3 and Rgt2, suggesting that they are involved in extracellular sensing (Pinson et al., 2004;Ljungdahl and Daignan-Fornier, 2012).When cells are grown in low Pi, the concentration of α-myo-d-inositol heptakisphosphate (IP 7 ) is increased in a VIP1-dependent manner (Lee et al., 2007).IP 7 physically interacts with Pho81, which enhances the inhibitory effect of Pho81 toward the Pho80-Pho85 complex and thereby facilitates the induction of Pho4/Pho2 target genes (Lee et al., 2008).A Pho84-disrupted strain (Δpho84) shows constitutive expression of the repressive acid phosphatase (APase) encoded by the PHO5 gene; this is because intracellular Pi is depleted by the loss of Pi uptake, even when the extracellular Pi concentration is high (Auesukaree et al., 2004).However, this constitutive expression phenotype can be suppressed by the overexpression of low-affinity Pi transporters, suggesting that Pho84 itself does not have a phosphate-sensing function and that there is an as-yet-unknown intracellular Pi-sensing mechanism (Wykoff and O'Shea, 2001).Nev-ertheless, how cells sense intracellular Pi concentration and how this signal is connected to the regulation of IP 7 levels remain unsolved.
In this study, to identify factors involved in sensing intracellular levels of Pi, we generated and isolated mutants that suppressed the constitutive APase expression phenotype of Δpho84 mutants and performed genetic analyses.Among the suppressor mutants identified, three were shown to contain an unknown single mutation.In addition, all three mutations were found to have an effect upstream of the Pho80-Pho85 complex.Therefore, these three mutations are likely to be mutations in a novel signal transduction factor in the PHO pathway.

MATERIALS AND METHODS
Strains and media Saccharomyces cerevisiae strain NBW7 was used as the parental strain and detailed strain information is shown in Table 1.Yeast cells were grown in YPDA medium consisting of 5% YPD broth (Sigma Aldrich) and 0.04% adenine (Wako), or in SC medium consisting of 0.67% yeast nitrogen base without amino acids (Difco) and 2% glucose supplemented with appropriate amino acids when required.The concentration of Pi was adjusted to 11 mM and 0.22 mM for high-and low-Pi conditions, respectively, unless stated otherwise.Low-Pi YPDA medium was prepared by treating YPDA medium with MgSO 4 and aqueous ammonia (Rubin, 1974).For sporulation, 0.5% sodium acetate was added.For solid media, 2% agar was added to the medium.Random mutagenesis by ethyl methanesulfonate treatment Yeast strain SH8333 (Δpho84) was cultivated overnight in 10 ml of YPDA medium.The cells were collected by centrifugation, washed with 3 ml of 0.1 M phosphate buffer (pH 7.0), and then resuspended in 6 ml of 0.1 M phosphate buffer.Aliquots of cell suspension (2 ml) were mixed with 7.5 ml of 0.1 M phosphate buffer, 0.2 ml of 40% glucose solution, and 0.3 ml of ethyl methanesulfonate (EMS) with gentle shaking at 30 °C for 45 min.Next, cells were collected by centrifugation, resuspended in 5 ml of 20% sodium thiosulfate, and incubated at 30 °C for 10 min to inactivate EMS.After several washes, the cells were diluted appropriately and spread onto high-Pi YPDA plates.
Genetic analysis A dominance-recessiveness test was carried out on 67 sef Δpho84 mutants as follows.All sef Δpho84 mutants (MATa) were crossed with strain TSY1 (MATα Δpho84), and the resulting diploids were checked for the production of APase on high-Pi YPDA plates.
For tetrad analysis, 12 sef Δpho84 mutants with novel sef mutations were crossed with TSY1 (Δpho84).The resulting diploids were grown on YPDA plates, transferred to sporulation medium, and incubated at 23 °C for 3 to 5 days.Asci were treated with Zymolyase-100T for 5 min and dissected by a micromanipulator (Singer Instruments MSM system).Cells from the spores were subjected to an APase staining assay on high-Pi YPDA plates.
Classification into complementation groups was performed as follows.Six sef Δpho84 mutants (MATa) with a predicted single sef mutation were crossed with TSY1 (MATα), the resulting diploids were sporulated, and segregants inheriting the MATα mating type and original sef mutation were selected.The six sef Δpho84 mutants were then crossed in all combinations to determine their complementation group.
Construction of single sef mutants was performed as follows.sef8 Δpho84, sef9 Δpho84 and sef10 Δpho84 mutants were crossed with NBW7 (MATα) and the resulting diploids were sporulated and subjected to tetrad analysis.Because 2 + :2-, 1 + :3-and 0 + :4-segregation for APase production in high-Pi medium was observed in the tetrads, we judged that the single sef mutants showed an APase -phenotype under high-Pi conditions.
Detection and measurement of APase activity A colony-staining assay for detecting APase activity on a solid medium was performed as described previously (Toh-E and Oshima, 1974).APase activity was measured as described previously (Toh-E et al., 1973) with some modifications.An aliquot of cell culture was collected by centrifugation (3,000 rpm, 5 min), washed once with sterile water, and then resuspended in an appropriate amount of sterile water.Forty microliters of the cell suspension was then mixed with 10 μl of 1 M acetate buffer (pH 4.0), 40 μl of p-nitrophenylphosphate solution (3.2 mg/ml) and 110 μl of sterile water, and incubated at 35 °C for 10 min.The reaction was stopped by the addition of 200 μl of 10% (vol/vol) trichloroacetic acid, and 400 μl of saturated sodium carbonate was added.After mixing and centrifugation, absorbance of the supernatant at 420 nm was measured.One unit of APase activity was defined as the amount of enzyme that liberated 1 μmol of p-nitrophenol in 1 min, which was calculated using 1.6 × 10 4 as the molar extinction coefficient of p-nitrophenol.Enzyme activity was expressed after normalization to the optical density at 600 nm (OD 600 ) value of the cell suspension.
Northern blot hybridization sef8 Δpho84, sef9 Δpho84 and sef10 Δpho84 cells grown in 50 ml of either  high-or low-Pi YPDA medium were collected when the OD 600 value reached mid-log phase.Total RNA was extracted as described previously (Auesukaree et al., 2005), and 15 μg of each RNA sample was subjected to gel electrophoresis and subsequent northern blot analysis (Brown, 2001).DNA probes for detecting the PHO5 gene and ACT1 gene as a control were prepared as described previously (Auesukaree et al., 2005).Probe labeling was performed using a PCR DIG Probe Synthesis Kit (Roche).
Measurement of Pi uptake activity sef8 Δpho84, sef9 Δpho84 and sef10 Δpho84 cells grown in 10 ml of highor low-Pi YPDA medium at 30 °C were collected when the OD 600 value reached 1.0.Phosphate-free Burkholder synthetic minimal medium (no-Pi medium) was prepared by substitution of 1,500 mg of KCl for 1,500 mg of KH 2 PO 4 per liter in the medium (Ueda and Oshima, 1975).The cells were washed in no-Pi medium, inoculated into 10 ml of no-Pi medium containing 100 μl of cycloheximide (500 μg/ml) at an OD 600 value of 0.1, and shaken at 30 °C for 30 min.The Pi uptake activity of the cell suspension was measured as described previously (Ueda and Oshima, 1975;Auesukaree et al., 2003).

RESULTS
Isolation of suppressor mutants of a pho84 disruptant by EMS mutagenesis A disruptant of the PHO84 gene encoding a high-affinity phosphate transporter is almost unable to take up extracellular Pi and shows constitutive APase expression (Tamai et al., 1985).To identify genes involved in the signaling process upstream of the Pho81-Pho80-Pho85 complex, we attempted to obtain suppressor mutants of the Δpho84 disruptant (Supplementary Fig. S1).We treated Δpho84 cells by EMS random mutagenesis and screened for mutants in which the constitutive APase expression phenotype of Δpho84 was suppressed.After EMS treatment, the cells were spread and grown on high-Pi plates, and then a colony-staining assay for APase activity was carried out on almost 10 5 colonies.As a result, we obtained 67 mutants in which constitutive APase expression had been abolished.The 67 mutants, named sef (suppressor of pho84 [pho eightyfour]), were subjected to a colony-staining assay for APase activity on low-Pi medium.Thirty-three mutants showed no expression of APase even in the low-Pi condition, whereas the remaining 34 mutants showed some APase expression.

Genetic analysis of sef mutants
We conducted a series of genetic analyses on the sef mutants as summarized in Supplementary Fig. S1.First, we performed a dominance-recessiveness test in which the 67 sef Δpho84 mutants were crossed with the Δpho84 mutant.The resulting hybrid diploid strains were subjected to a col-ony-staining assay for APase activity under the high-Pi condition.One strain showed weak pink staining, whereas the other 66 strains showed a clear red color (data not shown), indicating that among the 67 suppressor strains, only one had a dominant mutation, while 66 had a recessive mutation.At present, six gene mutations are known to suppress the constitutive APase expression phenotype of the Δpho84 disruptant (pho81, pho4, pho2, pho5, phm3 and phm4) (Auesukaree et al., 2004).Next, therefore, we subjected the sef mutations to a complementation test with these gene mutations.We crossed the 66 recessive sef Δpho84 mutants individually with Δpho81 Δpho84, Δpho4 Δpho84, Δpho2 Δpho84, Δpho5 Δpho84, Δphm3 Δpho84 and Δphm4 Δpho84 mutants.The resulting hybrid diploids were subjected to a colony-staining assay for APase activity under the high-Pi condition.The results revealed that 16, 16, 13 and 9 sef mutants were classified into pho81, pho4, pho2 and pho5 complementation groups, respectively, because the resultant diploids displayed no APase activity on high-Pi medium.None of the sef mutants was classified into the phm3 or phm4 complementation group.The remaining 12 sef mutants were not classified into any of the six known complementation groups.Therefore, we concluded that these 12 sef mutants contained mutations in a novel gene.All 12 sef Δpho84 mutants produced APase under low Pi.
To determine how many different genetic mutations were present in the 12 sef mutants, we performed a tetrad analysis of the hybrid diploids used in the dominance-recessiveness test and carried out the APase staining assay under high Pi for each segregant.Six sef mutants showed a clear 2 + :2-segregation pattern, while the remaining six sef mutants did not (data not shown).This result indicated that six of the sef mutants had a single mutation associated with suppression of the Δpho84 phenotype.
To classify the six sef mutants potentially harboring a single mutation, we performed a complementation test.We prepared six sef Δpho84 strains with two mating types, crossed them to make diploids, and examined their APase activity.The six sef mutants were divided into three complementation groups, which we named sef8, sef9 and sef10 (Table 2).APase activity of sef mutants We measured the APase activity of the sef8 Δpho84, sef9 Δpho84 and sef10 Δpho84 strains under high-and low-Pi conditions for a quantitative assessment of APase activity (Fig. 2A).The three mutants showed low APase activity similar to the wild type on high-Pi medium, and displayed high APase activity on low-Pi medium, consistent with the results of the colony-staining assay.Next, we constructed single sef mutants (sef8-1, sef9-2 and sef10-1) by introduction of a functional PHO84 gene and measured APase activity (Fig. 2B).sef8-1 and sef10-1 mutants showed a significant decrease in APase activity on low-Pi medium as compared with wild type, whereas the sef9-2 mutant showed almost the same APase activity as wild type.
sef mutations suppress constitutive APase expression in Δpho84 at the transcriptional level To investigate whether the suppression of APase activity under high Pi in sef mutants was exerted at the transcriptional level, we performed northern blot analysis to detect PHO5 mRNA (Fig. 3).PHO5 mRNA was not detected under high Pi in any of the sef Δpho84 mutants, similar to wild type, whereas it was clearly detected under low Pi.Thus, this result indicated that the sef mutations transcription of PHO5 in the Δpho84 disruptant under the high-Pi condition.
Transcriptional repression of PHO5 in sef8 Δpho84, sef9 Δpho84 and sef10 Δpho84 mutants is not due to recovery of phosphate uptake Because Δpho84 cells are defective in taking up extracellular Pi, their intracellular Pi level is low even under conditions of high extracellular Pi (Auesukaree et al., 2004).To clarify whether the transcriptional repression of PHO5 observed in the sef Δpho84 mutants was due to a recovery of Pi uptake, we measured the Pi uptake activity of the sef Δpho84 mutants under high and low Pi (Table 3).The maximum velocity (V max ) of wild type in high-Pi medium was 1.39 ± 0.21 (nmol min − 1 OD 600 − 1 ), whereas that of Δpho84 was 0.56 ± 0.22.In the sef Δpho84 mutants, the V max value ranged from 0.59 to 0.73, which was almost the same as that of the Δpho84 disruptant.In addition, because the Pi concentration in this experiment was much higher than the K m value, the actual uptake velocity in all strains was likely to be almost the same as the V max in both high-and low-Pi conditions.This result suggested that the intracellular Pi level in the sef Δpho84 mutants was not recovered to that of the wild type; therefore, suppression of the constitutive expression phenotype of Δpho84 in the sef mutants was not likely to be due to a recovery of extracellular Pi uptake.

Sef8, Sef9 and Sef10 function upstream of Pho80-Pho85
To determine where the gene products of the sef mutations function in the PHO pathway, we performed an epistasis-hypostasis test between the sef Δpho84 mutations and a pho80 mutation.We disrupted the PHO80 gene in sef8 Δpho84, sef9 Δpho84 and sef10 Δpho84 strains, and measured the APase activity of the resultant strains (Fig. 4).Under the high-Pi condition, the three mutants (sef8 Δpho84 Δpho80, sef9 Δpho84 Δpho80 and sef10 Δpho84 Δpho80) showed strong APase activity, which was even higher than that of Δpho80.This result clearly demonstrated that the pho80 mutation is epistatic to the sef mutations, indicating that the SEF gene products function upstream of the Pho81-Pho80-Pho85 complex in the PHO pathway.
sef9 mutant shows low-temperature sensitivity To characterize the three sef mutants, we investigated their growth phenotype under a low-temperature condition, because it has been reported that Pi uptake is a limiting factor for growth at low temperature (Vicent et al., 2015).The sef Δpho84 strains were streaked on YPDA plates and incubated at 13 °C.Among the sef Δpho84 mutants, only sef9 Δpho84 showed weak growth, which was similar to that of the control Δsit4 strain (Fig. 5).This low-temperature-sensitive phenotype co-segregated with high APase activity under the high-Pi condition (data not shown).We concluded that the sef9 mutation is responsible for a low-temperature-sensitive phenotype.

High Pi
Low Pi

PHO5 ACT1
Fig. 3 Fig. 3. Northern blot analysis of the PHO5 gene transcript in sef Δpho84 mutants.Total RNA was extracted from sef8 Δpho84, sef9 Δpho84, and sef10 Δpho84 cells grown on YPDA medium containing high or low Pi.Fifteen micrograms of RNA per sample was subjected to gel electrophoresis and subsequent northern blot analysis.The ACT1 gene was used as a loading control.

DISCUSSION
How yeast cells sense intracellular Pi and transmit the signal to the downstream PHO pathway has been a longstanding mystery.In particular, little is known about the genes that act upstream of the Pho81-Pho80-Pho85 complex and downstream of Pho84.In this study, we identified three novel SEF genes, named SEF8, SEF9 and SEF10, which are involved in the Pi signaling pathway and probably act upstream of the Pho81-Pho80-Pho85 complex.
Recently, Choi et al. identified many genes involved in signaling upstream of the Pho81-Pho80-Pho85 complex through a systematic screen of disruption mutations, and investigated a subset of these genes in more detail (Choi et al., 2017).From the results of an epistasis-hypostasis test of their mutations with both pho80DΔ (causing attenuated PHO80 mRNA expression) and pho81Δ mutations, they reported that their mutations affected the PHO pathway upstream of Pho85-Pho80.We also performed an epistasis-hypostasis test of sef8, sef9 and sef10 with the Δpho80 deletion mutation and demonstrated that sef Δpho80 Δpho84 triple mutants display much higher PHO5 expression than sef Δpho84 double mutants (Fig. 4), suggesting that the SEF genes also act upstream of Pho85-Pho80.
Although we did not expect that leaky mutations of the known pho5, pho4, pho2, pho81, phm3 and phm4 mutants would be caught by our strategy of mutant hunting, a study like ours has not previously been conducted.Nevertheless, our complementation test excluded the possibility that the sef8, sef9 and sef10 mutations were present  in known genes.Therefore, even if the gene products of sef8, sef9 and sef10 are involved in transcription of PHO5 as chromatin components, we are sure that they are novel genes that have not been uncovered by previous screening.Deletion of genes encoding adenosine kinase (ADO1) and adenylate kinase (ADK1) leads to constitutive expression of PHO5 (Huang and O'Shea, 2005).Choi et al. recently discovered that deletion of AAH1, which is also involved in adenine nucleotide metabolism, leads to derepression of the PHO pathway under high Pi in a Vip1-dependent manner (Choi et al., 2017).Thus, these three genes repress activation of the PHO pathway under high Pi and are involved in negative regulation of the PHO pathway.By contrast, the single sef8 and sef10 mutants exhibited decreased expression of PHO5 under low Pi (Fig. 2B), suggesting that their gene products act positively on the PHO pathway.
Notably, all sef Δpho84 mutants showed repression of the PHO pathway, and Pi uptake capacity was not restored in any of them under the high-Pi condition (Table 3).Therefore, it appears that extracellular Pi level is unlikely to affect the level of intracellular Pi.However, we cannot exclude the possibility that yeast cells have two systems to sense both extracellular and intracellular Pi concentration, and that an extracellular Pi sensor(s) transmits the Pi signal into the cell even in the pho84 mutant.Although it is not known what gene(s) are responsible for sensing extracellular Pi level, we think that the low-affinity Pi transporters Pho87 and Pho90 are candidates.
On the other hand, we have shown in a previous paper (Auesukaree et al., 2004) that PHO5 expression is strongly correlated with the levels of both intracellular orthophosphate and intracellular polyphosphate, and that the signaling defect in the pho84 strain is likely to result from insufficient intracellular phosphate caused by a defect in phosphate uptake.If SEF genes are involved in control or sensing of the intracellular Pi level, the sef Δpho84 strain may show repression of the PHO pathway in the high-Pi condition but derepression in the low-Pi condition, as shown in Figs. 2 and 3. Intracellular Pi level may be altered by metabolic change, such as energy metabolism or polyphosphate synthesis in the vacuole, possibly caused directly or indirectly by sef mutation.However, since we did not examine intracellular Pi level in this study, we cannot exclude at least the following two possibilities.First, sef mutation may increase intracellular Pi level through metabolic alteration, as revealed in Δpho84 phm3 and Δpho84 phm4 mutants (Auesukaree et al., 2004).Second, Sef protein may sense intracellular Pi level and subsequently transmit the signal downstream to activate the PHO pathway.Although we prefer the second possibility, i.e., the gene products of SEF8, SEF9 and SEF10 are candidates for sensing the intracellular Pi level, it should be noted that we cannot exclude the first possibility.To clarify this issue, more detailed analysis including the measurement of intracellular Pi level in pho84 and sef pho84 strains will be needed.
The sef9 mutant showed sensitivity to cold stress.Many heat shock proteins such as HSP12, HSP26, HSP42 and HSP104 are induced under cold stress conditions (Aguilera et al., 2007).These HSP proteins need ATP for their function.Indeed, Vicent et al. reported that overexpression of PHO84 restored the cold sensitivity of HSP mutants (Vicent et al., 2015).One possible explanation for this observation is that ATP level might be important for cold tolerance or adaptation.Therefore, the intracellular ATP level in the sef9 mutant may be decreased relative to the other sef mutants and the pho84 disruptant.Choi et al. recently reported that intracellular ATP level is linked to intracellular Pi availability (Choi et al., 2017).Thus, the SEF9 gene may be involved in adenine nucleotide metabolism.
In conclusion, we have identified three novel SEF genes (SEF8, SEF9 and SEF10) that most probably function upstream of the Pho81-Pho80-Pho85 complex.Our discovery of these SEF genes will provide a resource for further studies to understand the molecular mechanisms by which the PHO pathway is regulated.Toward this goal, cloning of the SEF8, SEF9 and SEF10 genes is now underway.

Fig. 1 .
Fig. 1. Outline of the PHO pathway under high-and low-Pi conditions.

Fig. 2 .
Fig. 2. Acid phosphatase (APase) activity in sef Δpho84 mutants and sef mutants.sef Δpho84 mutants grown on high-Pi (left panel) and low-Pi (right panel) plates were subjected to APase staining.Red-colored and white-colored colonies show APase activity and no APase activity, respectively.Bar graphs show APase activity in liquid culture.(A) APase activity in sef Δpho84 mutants.(B) APase activity in sef mutants.The values are the means and standard deviations of results from three independent experiments.

Fig. 4 .
Fig. 4. Epistasis-hypostasis test between sef Δpho84 and pho80 mutations.APase staining and APase activity measurement were performed in sef Δpho84 and sef Δpho84 Δpho80 mutants under the high-Pi condition.The values are the means and standard deviations of results from three independent experiments.

Fig. 5 .
Fig. 5. Low-temperature sensitivity of sef Δpho84 and sef mutants.sef Δpho84 and sef mutants were streaked onto YPDA plates and incubated at 13 °C.Δsit4 was used as a control strain.

Table 1 .
Saccharomyces cerevisiae strains used in this study

Table 2 .
Result of complementation test

Table 3 .
Pi uptake capacity in sef Δpho84 mutants