Introduction

Inorganic polyphosphate (polyP), a linear polymer containing a few to several hundred orthophosphate residues linked by energy-rich phosphoanhydride bonds, is an essential and multifunctional component of microbial cells (Rao et al. 2009; Albi and Serrano 2016; Kumar et al. 2016; Clotet 2017). PolyP is involved in architecture and functioning of yeast cell wall. Cell wall polyP plays a key role in the regulation of polysaccharide biosynthesis (Kalebina et al. 2008) and in the tolerance to heavy metal ions (Ryazanova et al. 2016). PolyP is an essential component of specific supramolecular complexes in the cell wall of Candida maltosa which are formed at grown on hexadecane (Zvonarev et al. 2017). Cell wall–associated polyP in Cryptococcus neoformans is required for capsule assembly (Ramos et al. 2017). The metabolism of yeast cell wall polyP remains poorly understood. Dolichyl-diphosphate:polyphosphate phosphotransferase was suggested to be responsible for synthesis of this polyP (Shabalin and Kulaev 1989). The С. neoformans knockout mutant in VTC4 gene had no detectable polyP in cell wall (Ramos et al. 2017). This indicates the key role of vacuolar Vtc4 polyphosphate synthase in cell wall polyP synthesis. The Рpx1 polyphosphatase degrading polyP was purified from the cell envelope of S. cerevisiae (Andreeva and Okorokov 1993). However, the enzyme has a neutral pH optimum (Andreeva and Okorokov 1993), while S. cerevisiae cells acidified the culture liquid during the growth (Kane 2016). There are no data on the potential involvement of acid phosphatases in polyP metabolism in yeasts.

Acid phosphatases (ACPases) in S. cerevisiae are involved in the system of phosphate homeostasis (Secco et al. 2012; Yadav et al. 2016; Eskes et al. 2018). There are several highly homologous ACPases in S. cerevisiae (Oshima 1997). The two adjacent genes YBR092C and YBR093C encode the constitutive Pho3 and repressible Pho5 enzymes, respectively; another two repressible ACPases are Pho11/YAR071W and Pho12/YHR215W (https://www.yeastgenome.org).

Due to numerous technical advantages, the PHO system was used for studies of the mechanisms of transcription regulation in yeast (Kornberg 2005; Korber and Barbaric 2014). The РНО5 gene is interesting with regard to the structural specificity of its promoter (Kornberg 2005; Korber and Barbaric 2014), and the features of its expression during cell cycle (Pondugula et al. 2009) and under phosphate (Yadav et al. 2016) or potassium starvation (Anemaet and van Heusden 2014; Yadav et al. 2016).

The limited information on the substrate specificity of yeast ACPases raises the question about their potential involvement in the metabolism of polyP. It is known that Pho3 hydrolyzes thiamin phosphates in the periplasmic space, increasing thiamin uptake (Nosaka 1990), and Pho5 has nucleoside-triphosphatase and nucleoside-triphosphate pyrophosphatase activities (Kennedy et al. 2005). PHO5 is expressed during phosphate starvation (Oshima 1997). The level of high-molecular alkali-soluble polyP probably localized in the cell wall also decreased in these conditions (Vagabov et al. 2000).

The goal of this work was to clarify the potential role of acid phosphatase Pho5 in polyphosphate metabolism using a S. cerevisiae strain constitutively overexpressing PHO5.

Materials and methods

Yeast strains and plasmid construction

The parent strain was YPH857 (MATa leu2d1, lys2-801 ade2-101 his3-D 200 trp1-D 63 ura3-52 cyh2) (Spencer et al. 1993). To construct the PHO5 overexpressing strain, we used the previously developed pMB1 expression vector (Eldarov et al. 2013). The vector is derivative of pRS316 centromeric vector with URA3 marker (Sikorski and Hieter 1989) and contains strong constitutive TDH3 promoter and PGK terminator. The PHO5 coding sequence was obtained as a 1.4-kbp PCR fragment with Phuzion DNA polymerase (NEB) and two primers PHO5_1 (5′ ggtacATGTTTAAATCTGTTGTTTATTCA) and PHO5_2 (5′ ggtctcgagATT TAG TTA TAC AAA ACT ATT), containing flanking PciI and XhoI sites, respectively. After cleavage with PciI and XhoI, the fragment was cloned into NcoI/XhoI digested pMB1 vector. Plasmid clones with correct PHO5 CDS inserts were picked after plasmid sequencing and used to transform the recipient YPH857 strain by standard lithium acetate procedure (Gietz and Schiest 2007). The kits and enzymes for DNA manipulations (isolation of plasmid DNA, purification of DNA fragments, DNA hydrolysis by restriction enzymes, ligation of DNA fragments, PCR amplification, and site-directed mutagenesis) were from SibEnzyme (Russia), Fermentas MBI (Lithuania), and Stratagene (USA). Pho5 overproduction in obtained Ura+ transformants was verified with colorimetric assay (see below).

Culture conditions

The strains were maintained on an agarized YNB medium (Difсo, USA) supplemented with amino acids and nucleotides as described earlier (Eldarov et al. 2013). For biomass and enzyme production, the cultures were grown at 29 °C in flasks with 200 mL of the YPD medium (glucose, 2%; peptone, 2%; yeast extract, 1%) at 150 rpm for 48 h. The medium contained 3.7 mM phosphate (Pi). The biomass was harvested by centrifugation at 3000g and twice washed with cold distilled water. In some cases, the cells were cultivated in YPD medium with low Pi (0.04 mM) obtained as described earlier (Rubin 1973).

Pi and polyP assay

The polyP fractions were extracted from biomass samples as described earlier (Vagabov et al. 2000). The amounts of Pi and polyP were assayed accordingly (Eldarov et al. 2013).

Pho5 localization

Spheroplasts were obtained from the cells of YPH857/pMB1_PHO5 strain. The cells were washed with 0.8 M mannitol in 5% citrate buffer, pH 6.5, and incubated in the same buffer with lyophilized gastric juice of Helix pomatia (150 mg per 1 g of biomass) and dithiothreitol (10 mg per 1 g of biomass) for 2 h at 30 °C. Spheroplasts were washed with the same buffer and lysed in 20 mM Mes (2-(N-morpholino)ethanesulfonic acid) –NaOH, pH 4.5, with 0.5 mM PMSF (phenylmethylsulfonyl fluoride) (Eldarov et al. 2013). The cell debris was removed by centrifugation at 12,000g for 60 min, and the supernatant was used for ACPase activity assay.

Three variants of ACPase extraction from the whole cells were compared. (a) The cells were suspended in MES-NaOH, pH 4.5, with 0.5 mM PMSF, and sonicated for 30 s 5 times; then the samples were centrifugated at 12,000g for 3 min. (b) The cells were suspended in MES-NaOH, pH 4.5, with 0.5 mM PMSF and 1 M NaCl, and incubated for 30 min at 0 °C under stirring. The cells were removed by centrifugation at 12,000g for 3 min. (c) The cells were suspended in MES-NaOH, pH 4.5, with 0.5 mM PMSF and 1 M NaCl, exposed to ultrasound for 30 s 5 times, and then incubated for 30 min at 0 °C under stirring. The cells were removed by centrifugation at 12,000g for 3 min.

Purification of Pho5

The culture liquid after cultivation of YPH857/pMB1_PHO5 strain for 48 h in YPD medium was used for Pho5 purification. (NH4)2SO4 was added to the culture liquid up to 70% saturation. The pellet was removed by centrifugation at 12,000g for 60 min. The supernatant was concentrated using Millipore membrane YM-10 and washed threefold with 20 mM MES-NaOH, pH 4.5, under the same ultrafiltration conditions. The preparation was placed on the column (0.5х7 cm) with DEAE-Toyopearl 650M (TOSON, Japan) equilibrated with 20 mM MES-NaOH, pH 4.5. The column was washed with 100 mL of the same buffer. The acid phosphatase was eluted at a flow rate of 24 ml/h with increasing linear gradient of KCl (0–0.8 M) in the same buffer. The gradient volume was 200 mL. The fractions with ACPase activity were pooled and used for analysis. The purification procedure was performed at 0 °C. The purified Pho5 preparation was analyzed in standard SDS-PAGE.

Enzyme activity assay

The ACPase activity was determined at 30 °C with p-nitrophenylphosphate colorimetric assay (Bergmeyer et al. 1974). The incubation medium contained 20 mM MES-NaOH, pH 4.5, 2.5 mM MgSO4, and 2.7 mM p-nitrophenylphosphate. The reaction was stopped by 50 mM NaOH, and the absorbance at 400 nm was measured. The activities with other substrates were measured in the same buffer by Pi release. All substrates were incubated in the same conditions without enzyme, and the amounts of Pi in substrate preparations were assayed and used as control for activity quantification. The Pi was assayed as described earlier (Eldarov et al. 2013). For assay of exopolyphosphatase activity, the commercial preparations of polyP of average 15 phosphate residues (polyP15) and polyP of average 208 phosphate residues (polyP208) were purified by chromatography of Sephadex G-10 to avoid Pi and pyrophosphate admixtures (Andreeva and Okorokov 1993). PolyP208 was from Monsanto (USA), and polyP15 and other substrates were from Sigma-Aldrich (USA). The concentration of GTP, NADP, and ribose-5-phosphate was 1.25 mM, and the concentration of other substrates was 2.5 mM. The concentration of polyPs was calculated in terms of phosphate residues. The concentration range of 0.165–2.7 mM was used for Michaelis constant evaluation. The amount of the enzyme forming 1 μmol of Pi (or nitrophenol in the case of p-nitrophenylphosphate) per 1 min was taken as a unit of enzyme activity (U). In the case of pyrophosphatase, the amount of enzyme cleaving 1 μmol of pyrophosphate per 1 min was taken as a unit of enzyme activity (U). Protein concentration was assayed with BSA as a standard by the Bradford method according the Manual from Thermo Fischer Scientific.

All experiments were repeated three times, and the mean values of three experiments with standard deviations calculated by standard Excel program are presented in the tables and figures.

Results and discussion

ACPase levels and effects of Pi content and growth stage

The parent YPH857 strain and the YPH857/pMB1_PHO5 strain showed similar growth in YPD medium with 3.7 mM and 0.04 mM of Pi (Fig. 1a).

Fig. 1
figure 1

The growth of parent strain YPH857 and Pho5 overproducing strain YPH857/pMB1_PHO5 in the media with different Pi concentrations (a). The effects of Pi and growth stage on total acid phosphatase activities in whole cells (black columns) and culture liquid (gray columns) of parent strain YPH857 (b) and Pho5 overproducig strain YPH857/pMB1_PHO5 (c). The time of cultivation and Pi concentration in the medium are shown on X-axis. The activity was assayed with p-nitrophenylphosphate. The wet biomass yield was the same for both strains: 10 g/L for 28 h of cultivation in the media with 3.7 mM Pi and 0.04 mM of Pi, and 20 and 17 g/L at 48 h of cultivation in the media with 3.7 mM Pi and 0.04 mM of Pi, respectively

Full size image

The Рho5 is a cell wall enzyme inducible under Pi limitation (Toh-e et al. 1973). We compared the ACPase activities of whole cells and the culture liquids of the parent strain and the YPH857/pMB1_PHO5 strain grown to the logarithmic stage and stationary stage in Pi-sufficient and Pi-limited media. In the parent strain, the activity was in all cases associated mainly with the cells, while the culture liquid contained much less activity. In agreement with previous studies (Shnyreva et al. 1996) in the stationary growth stage in the Pi-limited medium, there was an approximately 2-fold increase in activity compared to the cultivation in the Pi-sufficient medium (Fig. 1b). The total ACPase activity of the whole cells and culture liquid of the YPH857/pMB1_PHO5 strain exceeded that of the parent strain approximately 10- and 20-fold, respectively (Fig. 1c). PHO5 expression in YPH857/pMB1_PHO5 strain is under the control of the strong TDH3 promoter. Therefore, the dependence of total ACPase activity on Pi concentration in the medium differs from that in parent strain. In the stationary growth stage in the Pi-limited medium, the total ACPase activity was lower than on the Pi-sufficient medium (Fig. 1c). The decrease in total ACPase in the overexpressing strain at low Pi concentrations is probably caused by a slight suppression of growth.

The cellular localization of Pho5

We have assessed the localization of Pho5 in the cells of YPH857/pMB1_PHO5 strain grown to the stationary growth phase in the medium with 3.7 mM Pi. Figure 2 shows the similar amounts of this activity in association with the cell surface and in the culture liquid. In cell-free extract, the activity was very low. The cells were treated in three different ways as described in “Materials and methods,” and the best extraction technique (yielding ~ 30% activity) was the sonication in the presence of 1 М NaCl. Therefore, a significant part of Pho5 in overproducing strain is tightly associated with the cell wall, as well as it is known for the wild-type enzyme (Toh-e et al. 1973; Shnyreva et al. 1996).

Fig. 2
figure 2

The distribution of Pho5 in the overproducing strain of S. cerevisiae: 1, whole cells; 2, culture liquid; 3, cell-free extract; 4, extraction from cells by sonication; 5, extraction from cells by NaCl; 6, extraction from cells by sonication and NaCl (for details, see “Materials and methods” section) The activity was assayed with p-nitrophenylphosphate

Full size image

PolyP content

The strain YPH857/pMB1_PHO5 strain had constitutive high level of Pho5 in the cell wall and was suitable for studying the effect of PHO5 overexpression on polyP metabolism. The levels of all polyP fractions were similar in the cells of the parent strain and the transformant strain grown in YPD with 3.7 mM of Pi (Table 1). No difference was found in the level of alkali-soluble polyP extracted at рН 11–12 (Table 1). This fraction is considered as cell wall associated polyP (Kulaev et al. 2004).

Table 1 Pi and polyphosphate (polyP) content in the cells of S. cerevisiae, parent strain YPH857 and Pho5 overproducing strain YPH857/pMB1_PHO5 (μmol P/g wet biomass)
Full size table

Purification of Pho5

The high level of ACPase activity in the YPH857/pMB1_PHO5 strain simplified Pho5 purification. Pho5 was purified from the culture liquid, where specific activity was 1.06 U/mg protein. The purified preparation had a specific activity of 45.8 U/mg protein similar to the known preparation (Shnyreva et al. 1996). The procedure of purification was simple (see “Materials and methods” section), and enzyme yield was 95 U/L of culture liquid.

The molar mass of Pho5 calculated from gene sequence is 52.832 kD. However, Pho5 preparation showed a broad protein band with molar mass higher than 94 kD (Fig. 3). Similar result was described earlier (Shnyreva et al. 1996). This is due to high glycosylation of Pho5 (Kozulić et al. 1984; Shnyreva et al. 1996). The proteins with high glycosylation levels give in SDS-PAGE broad and diffuse bands (Skelton et al. 1998; Magnelli et al. 2011).

Fig. 3
figure 3

SDS-PAGE of Pho5 preparation (2). The protein markers (1) were phosphorylase b (94 kD), albumin (67 kD), ovalbumin (43 kD), carbonic anhydrase (30 kD), trypsin inhibitor (20.1 kD), and α-lactalbumin (14.4 kD)

Full size image

Pho5 properties

The Pho5 activity is usually assayed with artificial substrates, p-nitrophenylphosphate or α-naphthylphosphate. Purified Pho5 had a pH optimum at ~ 5.0 with p-nitrophenylphosphate (Fig. 4) similar to previous studies (Shnyreva et al. 1996). Pho5 hydrolyzes well nucleotide phosphates (Kennedy et al. 2005) and phosphorylated sugars, glycerophosphate, and pyrophosphate (Table 2). The presence of 2.5 mM MgSO4 had no effect on Pho5 activity with all substrates used in this study. The Michaelis constants determined for several substrates (Table 2, Fig. 5) were in the similar concentration range.

Fig. 4
figure 4

The pH dependence of activity of purified Pho5 with p-nitrophenylphosphate

Full size image
Table 2 Substrate specificity of purified Pho5 of S. cerevisiae
Full size table
Fig. 5
figure 5

The Michaelis–Menten plots for purified Pho5. All experiments were repeated triply and standard Excel program was used for calculations

Full size image

The activity with polyPs was negligible (Table 2). The activity with p-nitrophenylphosphate was only poorly inhibited by polyP with a chain length of about 208 phosphate residues (2.5 mM) and heparin (10 μg/mL) by 12 and 23%, respectively.

Pho5 well hydrolyzes all organic compounds with phosphoester bonds used in this work. Among the inorganic phosphates with such bond, the enzyme hydrolyzes well only pyrophosphate. This activity is not related to the admixture of pyrophosphatases, since it was not stimulated by magnesium ions. Unlike Pho5, the known yeast pyrophosphatases requires divalent metal cation for their activities (Halonen et al. 2002). The negligible activity of purified Pho5 with polyP and the absence of noticeable changes in the polyP level in Pho5 overproducing strain suggest that this acid phosphatase is not involved in polyP breakdown. Note that Pho8, an alkaline phosphatase localized in vacuoles, is capable of polyP hydrolysis and participates in polyP homeostasis (Kizawa et al. 2017).

The substrate specificity of Pho5 and its extracellular localization suggest its function: the hydrolysis of organic compounds with phosphoester bounds (i.e., nucleoside phosphates, phosphorylated sugars) for obtaining of Pi under phosphate limitation conditions. Pho5 had no function in controlling polyP in the cell wall.