Phosphorylation of the GTS1 gene product of the yeast Saccharomyces cerevisiae and its effect on heat tolerance and flocculation

Abstract

The GTS1 gene from the yeast Saccharomyces cerevisiae showed pleiotropic effects on yeast phenotypes, including an increase of heat tolerance in stationary-phase cells and an induction of flocculation. Here, we found that the GTS1 product, Gts1p, was partially phosphorylated at some serine residue(s) in cells grown on glucose. Studies using mutants of protein kinase A (PKA) and CDC25, the Ras-GTP exchange activator, showed that PKA positively regulated the phosphorylation level of Gts1p. Overexpression of Gts1p in a mutant with attenuated PKA activity did not show any increase of heat tolerance and partially decreased flocculation inducibility, suggesting that phosphorylation of Gts1p is required for induction of these phenomena.

Budding yeast, GTS1, Phosphorylation, Heat tolerance, Flocculation, Protein kinase A

1 Introduction

We isolated the GTS1 gene of the yeast Saccharomyces cerevisiae as a gene which influences the timing of budding and cell size [1] and overexpression of the gene was shown to increase the heat tolerance of stationary-phase cells [2]. Independently, Bossier et al. [3] isolated GTS1 from genomic library-transformed yeast which regrow after lethal heat shock, and found that overexpression of the gene resulted in constitutive flocculation of yeast. They also reported that overexpression of GTS1 in strain S18-1D, which has low protein kinase A (PKA) activity, failed to cause constitutive flocculation, suggesting that phosphorylation of Gts1p is required for the flocculation effect although they did not investigate whether Gts1p is a phosphoprotein [3]. On the other hand, a strong correlation has been observed between the activity of PKA and the extent of heat tolerance in yeast (for review see [4]). Cells with low PKA activity are remarkably thermotolerant, whereas cells with high constitutive PKA are heat-sensitive [5–7]. These results raised a possibility that both the increase of the heat tolerance and flocculation inducibility are regulated by PKA via phosphorylation of Gts1p. In the present study, we investigated the phosphorylation of Gts1p and found that PKA was involved, if not directly, in the phosphorylation of Gts1p. The phosphorylation of Gts1p was shown to be required for the induction of heat tolerance and flocculation.

2 Materials and methods

2.1 Yeast strains and media

Strains of S. cerevisiae used in this study are listed in Table 1. Compositions of YPAD (rich) and synthetic (SD) media have been described previously [1].

1

Genotypes of yeast strains used

Strain	Parental strain and genotypea	Source or reference
W303-1A	MATa ade2-1 his3-11 trp1-1 leu2-3 ura3-1 can1-100	K. Suzukib
TMpGTS1/W303	W303-1A, YEpGTS1	This study
TMΔgts1/W303	W303-1A, gts1::URA3	This study
SP-1	MATa ade8 his3 trp1 leu2 ura3	[5]
TMpGTS1/SP	SP-1, YEpGTS1	This study
S18-1D	SP-1, tpk1w1 tpk2::HIS3 tpk3::TRP1	[5]
TMpGTS1/S18	S18-1D, YEpGTS1	This study
RS13-58-A1	SP-1, tpk1w1 tpk2::HIS3 tpk3::TRP1 bcy1::leu2	[5]
S13-3A	SP-1, tpk2::HIS3 tpk3::TRP1 bcy1::leu2	[5]
STX491-13A	MATαcdc25-1 tyr1 his7 ade2 ura1 lys2	YGSCc
JC35-1b	MATa cka1::HIS3 cka2::TRP1 CEN/ARS URA3 GAL1-CKA1	[15]
RPG41-1a	MATa cka1::HIS3 cka2::TRP1 CEN4/ARSH4 URA3 CKA2	[15]
RPG62-6b	MATa cka1::HIS3 cka2::TRP1 CEN/ARS URA3 GAL1-Dmα GAL10-Dmβ	[15]

^aRelevant genotype is written in italic letters and plasmid genotype is in normal letters.

^bA gift from Dr. K. Suzuki (University of Hiroshima).

^cObtained from Yeast Genetic Stock Center, University of California at Berkeley.

2.2 Conditions of cell labeling, cell extraction, immunoprecipitation and immunoblots

When cell density measured by optical density at 550 nm reached 1.0, 0.1 mCi (3×10⁶ Bq) [³²P]orthophosphate ml⁻¹ (Japan Isotope Energy Research Institute, Tokai, Japan) or 0.1 mCi (3×10⁶ Bq) [³⁵S]methionine ml⁻¹ (Trans-label, ICN Radiochemicals, Irvine, CA, USA) was added. After growth for 2–3 h at 30°C, cells were harvested by centrifugation, washed once with ice-cold water and stored at −80°C until use. Immunoprecipitation of Gts1p was performed as described previously [8] using anti-Gts1p IgG. The immunoprecipitates were separated by polyacrylamide gel electrophoresis in the presence of SDS (SDS–PAGE). Western blotting of Gts1p was performed as described previously [1].

2.3 Analysis of phosphoamino acids and phosphatase treatment of labeled Gts1p

For analysis of phosphoamino acids, immunoprecipitated ³²P-labeled Gts1p was purified by extracting the protein from the gel after SDS–PAGE, and analyzed as described by Boyle et al. [9]. For treatment with phosphatase, immunoprecipitated ³⁵S-labeled Gts1p was incubated at 37°C for 45 min with or without 60 units of alkaline phosphatase in APS buffer (50 mM Tris–HCl, pH 8.0, 1 mM MgCl₂) containing protease inhibitors. Treated Gts1p was then eluted from Immunoprecipitin (Gibco BRL, MD, USA) and detected by autoradiography after SDS–PAGE.

2.4 Plasmids and yeast transformation

To obtain a GTS1-overexpressing transformant, named TMpGTS1, the EcoRI–SpeI fragment containing the sequence from −354 to +1507 of GTS1[1] was inserted into the multicopy plasmid YEp24. The resulting plasmid, named YEpGTS1, was used for transformation [10] using the strains W303-1A, SP-1 and S18-1D obtaining transformants TMpGTS1/W303, TMpGTS1/SP and TMpGTS1/S18, respectively (Table 1). A GTS1-disrupted mutant, named TMΔgts1/W303, was produced by replacing the BamHI–KpnI region of the GTS1 locus of strain W303-1A with URA3[1] (Table 1).

2.5 Determination of heat tolerance, trehalose content and intensity of flocculation

Cells were cultured in the synthetic medium at 30°C and cells in the stationary phase were heated at 52.5°C. At the indicated times, 2.5-μl aliquots of cells containing about 1×10⁵ cells were removed, spotted on YPAD agar plates and cultured at 30°C for 2 days. Cellular trehalose levels were determined as described previously [11].

The intensity of flocculation was examined by light microscopic observation of cells in the stationary phase after shaking in a test tube for 30 min at 30°C.

3 Results and discussion

3.1 Phosphorylation of the GTS1 protein

To examine whether Gts1p was phosphorylated in vivo, cell lysates labeled with [³²P]orthophosphate were immunoprecipitated with anti-Gts1p antibodies and analyzed by SDS–PAGE (Fig. 1a). The immunoprecipitates from the wild-type and GTS1-overexpressing (TMpGTS1/W303) cells exhibited a weak and a strong signal, respectively, corresponding to the position of Gts1p with a molecular mass of about 45 kDa. The lysate from the GTS1-disrupted cell (TMΔgts1/W303) did not show any significant signal at the position of 45 kDa (Fig. 1a). When cell lysates labeled with [³⁵S]methionine were analyzed by immunoblotting, two bands were detected on the autoradiogram at the position corresponding to Gts1p, and the slow-migrating band disappeared after treatment with protein phosphatase (Fig. 1b) suggesting that the slow-migrating band represented a phosphorylated form of Gts1p. Furthermore, analysis of the phosphoamino acids of ³²P-labeled Gts1p by thin-layer chromatography (Fig. 1c) revealed that a strong signal was detected at the position corresponding to phosphoserine, suggesting that Gts1p was phosphorylated at some serine residue(s).

1

Phosphorylation of Gts1p in vivo (a, b) and analysis of phosphoamino acids (c). a: Autoradiogram of immunoprecipitates (IP) with anti-Gts1 antiserum (α-Gts1p) and with preimmune serum from GTS1-overexpressing (TMpGTS1/W303), wild-type, and GTS1-disrupted (TMΔgts1/W303) cells labeled with [³²P]orthophosphate. b: Autoradiogram of immunoprecipitates with anti-Gts1p antiserum from GTS1-overexpressing (TMpGTS1/W303) and wild-type cells labeled with [³⁵S]methionine with and without protein phosphatase treatment. c: Analysis of [³²P]phosphoamino acids of Gts1p by autoradiography. ³²P-labeled proteins immunoprecipitated with anti-Gts1 antibody were digested with acid and analyzed by two-dimensional thin-layer chromatography. Positions of phosphorylated derivatives of serine (P-Ser), threonine (P-Thr) and tyrosine (P-Tyr) were determined by detecting co-migrated cold phosphoamino acids with ninhydrin. ori, the position where sample was applied.

3.2 Kinase(s) involved in the phosphorylation of Gts1p

Searching the data base Prosite for putative phosphorylation sites in Gts1p revealed that there are five potential sites for protein kinase C, three for casein kinase II and one for PKA (data not shown). The phosphorylation state of Gts1p was not affected in cells treated with staurosporine, an inhibitor of protein kinase C, or in three strains with mutant casein kinase II, JC35-1b, RPG41-1a and RPG62-6b [12] (Table 1), suggesting that neither protein kinase C nor casein kinase II is involved in the phosphorylation of Gts1p (data not shown). In strain S18-1D (tpk1^w1 tpk2 tpk3 BCY1), which has an attenuated TPK1 allele of one of the three catalytic subunit genes of PKA [5], Gts1p was almost completely dephosphorylated, and present at an increased level (Fig. 2a, lane 2). In the mutants with inactivated regulatory subunit of PKA (bcy1), RS13-58-A1 and S13-3A, which contain catalytic subunits tpk1^w1 and TPK1, respectively, Gts1p was moderately phosphorylated, and one additional slow-migrating band was present (Fig. 2a, lanes 3 and 4). As the additional band was found only in bcy1 cells, there may be a phosphorylation site in Gts1p which is attacked in the absence of the regulatory subunit of PKA. Furthermore, in strain STX491-13A carrying cdc25, a temperature-sensitive mutant of the Ras-GTP exchange activator [13,14], dephosphorylated Gts1p was accumulated at the restrictive temperature (Fig. 2b). These results suggest that PKA was involved, if not directly, in phosphorylation of Gts1p.

2

Effect of the Ras-PKA pathway on the phosphorylation of Gts1p. a: Phosphorylation states of Gts1p in the wild-type strain SP-1, and PKA mutants S18-1D (tpk1^w1 tpk2 tpk3 BCY1), S13-3A (TPK1 tpk2 tpk3 bcy1) and RS13-58-A1 (tpk1^w1 tpk2 tpk3 bcy1). b: Time course of phosphorylation state of Gts1p in cdc25 cells after a temperature shift to the restrictive temperature. Western blots of Gts1p and actin as a control were performed as described in Section 2.

Although the result suggested that Gts1p was phosphorylated at some serine residue(s) by PKA, the only putative phosphorylation site for PKA that fits the consensus pattern ([RK](2)-x-[ST]) contains the threonine residue at 258. Furthermore, Gts1p with a Thr-258 to Ala substitution was normally phosphorylated as assessed by SDS–PAGE, and Gts1p in which amino acid residues 229–292 were deleted was also phosphorylated (data not shown). However, as over 20% of experimentally verified phosphorylation sites for PKA do not match the consensus pattern [15], it cannot be ruled out that PKA phosphorylates Gts1p. Alternatively, Gts1p may be directly phosphorylated by other kinase(s) under the positive regulation of PKA.

3.3 Effect of phosphorylation of Gts1p on heat tolerance and flocculation

We previously reported that the overexpression of Gts1p increased the heat tolerance of stationary-phase cells [2]. To test whether phosphorylation of Gts1p is required for the increase of heat tolerance, the heat sensitivity of S18-1D cells, which have low PKA activity, was compared to that of the wild-type strain SP1 with and without overexpression of Gts1p (Fig. 3). While the heat tolerance of wild-type cells was increased by the overexpression of Gts1p, that of S18-1D cells which have high thermotolerance [4–7] was actually decreased by overexpression, suggesting that phosphorylation of Gts1p was required for the increase in the heat tolerance of yeast. It should be added that overexpression of Gts1p did not change the level of intracellular trehalose in both S18-1D and the wild-type cells (data not shown). As heat shock proteins and trehalose have been reported to play an important role in the acquisition of heat tolerance of yeast [4], it is possible that Gts1p affects, if not directly, expression of some heat shock or related proteins. On the other hand, overexpression of Gts1p caused constitutive flocculation in S18-1D to a lesser extent than that in wild-type cells, and light microscopic photographs revealed that the flocs in the mutant were smaller than those in the wild-type cells (Fig. 4). The effect of Gts1p on flocculation here was smaller than that in Bossier et al. [3] who reported that Gts1p overexpression abolished constitutive flocculation. This discrepancy may be due in part to the different media used for cell suspension and in part to the strains used.

3

Effect of Gts1p phosphorylation on heat tolerance of yeast cells in the stationary phase. a: Heat tolerance (left panel) of the wild-type (SP-1) and GTS1-overexpressing transformant (TMpGTS1/SP). b: Heat tolerance (left panel) of the PKA mutant S18-1D (tpk1^w1 tpk2 tpk3 BCY1) and GTS1-overexpressing transformant TMpGTS1/S18. To verify that approximately equal numbers of cells were spotted, samples at time zero were spotted after sequential dilution (×10) (right panels).

4

Effect of Gts1p phosphorylation on flocculation of yeast cells observed by light microscopy. Photographs of (a) the wild-type (SP-1), (b) GTS1-overexpressing transformant (TMpGTS1/SP), (c) the PKA mutant S18-1D (tpk1^w1 tpk2 tpk3 BCY1) and (d) GTS1-overexpressing transformant TMpGTS1/S18.

Flocculation is defined as the Ca²⁺-dependent, non-sexual aggregation of yeast cells into flocs and reportedly 33 genes are involved in flocculation including four dominant structural genes (FLO1, FLO5, FLO9 and FLO10) [16,17]. Bossier et al. [3] reported that GTS1-induced flocculation is hardly sensitive to mannose and occurred in cells lacking the FLO1 gene suggesting that the flocculation mediated by FLO1 and GTS1 is unlinked. However, inconsistent with their results, we found that the flocculation was significantly reduced by 500 mM mannose but not by glucose, and that disruption of FLO1 abolished flocculation suggesting that the GTS1-induced flocculation is mediated by FLO1 (Iha et al., unpublished results). We are carrying out further experiments to resolve this discrepancy.

Although yeast cells showing strong flocculation are usually heat-resistant, the two phenotypes are probably unlinked because cells dispersed in the presence of 25 mM EDTA remained heat-resistant and because cells overexpressing GTS1 mutants lacking flocculation inducibility were still heat-resistant (unpublished results).