Isolation and Characterization of High-Osmolarity-Sensitive Mutants of Fission Yeast

Abstract

For the fission yeast Schizosaccharomyces pombe, adaptation to high-osmolarity medium is mediated by a mitogen-activated protein (MAP) kinase cascade, involving the Wis1 MAP kinase kinase and the Sty1 MAP kinase. The MAP kinase pathway transduces an osmotic signal and accordingly regulates the expression of the downstream target gene (gpd1⁺) that encodes NADH-dependent glycerol-3-phosphate dehydrogenase, in order to adaptively accumulate glycerol inside the cells as an osmoprotectant. We previously characterized a set of high-osmolarity-sensitive S. pombe mutants, including wis1, sty1, and gpd1. In this study, we attempted to further isolate novel osmolarity-sensitive mutants. For some of the mutants isolated, profiles of glycerol production in response to the osmolarity of the growth medium were indistinguishable from that of the wild-type cells, suggesting that they are novel types. They were classified into three distinct types genetically and, thus, were designated hos1, hos2, and hos3 (high osmolarity sensitive) mutants. One of them, the hos1 mutant, was characterized in detail. The hos1 mutant was demonstrated to have a mutational lesion in the known ryh1⁺ gene, which encodes a small GTP-binding protein. Disruption of the ryh1⁺ gene results not only in osmosensitivity but also in temperature sensitivity for growth. It was also found that the Δryh1 mutant is severely sterile. These results are discussed with special reference to the osmoadaptation of S. pombe.

Exposure of cells to high-osmolarity conditions in their environment led to dehydration and a decrease in cell viability. Accordingly, the ability of cells to adapt to external osmotic stress is a fundamental biological process that protects the organism against fluctuation in the water activity and solute content of their environment. In fact, many types of both prokaryotic and eukaryotic cells have developed mechanisms to adapt to severe osmotic stresses in their environment (often called osmoregulation) (4, 7). Recently, much attention has been focused on osmoregulation, with special emphasis on the molecular mechanism underlying signal transduction in response to osmotic stimuli.

The fission yeast Schizosaccharomyces pombe is an organism of choice to gain insight at the molecular level into signal transduction in response to an external osmotic stimulus (or stress) (1, 8, 9, 16, 22, 23, 25, 31). In general, the accumulation of osmoprotective, compatible solutes inside cells up to the concentrations necessary to counteract the elevation of external osmolarity is a well-documented aspect of osmoregulation (7). In S. pombe, glycerol appears to be the main compatible solute; this solute is synthesized from the glycolytic intermediate dihydroxyacetonephosphate in two steps that are catalyzed by an NADH-dependent glycerol-3-phosphate dehydrogenase and a phosphatase (4). We have recently cloned the gpd1⁺ gene, encoding osmoinducible glycerol-3-phosphate dehydrogenase, which was demonstrated to be crucially responsible for osmoregulation in S. pombe (19).

Recent extensive studies of S. pombe have begun to shed light on the stress-activated signal transduction mechanism by which the gpd1⁺ gene is activated in response to high external osmolarity (1, 22, 25, 31). A mitogen-activated protein (MAP) kinase cascade was found to be involved in this osmosensing signal transduction. The uncovered central elements of this MAP kinase cascade are the Sty1 MAP kinase (also known as Spc1 and Phh1) (12, 16, 23), the Wis1 MAP kinase kinase (MEK) (24), and the Wak1 MAP kinase kinase kinase (MEKK) (22). Interestingly, a His-Asp phosphotransfer signaling mechanism (or a two-component regulatory system) also appears to be implicated in the upstream region of this signaling pathway. In fact, a pair of histidine kinases (Mak1 and Mak2) and a response regulator (Mcs4) were suggested to be involved in the osmosensing pathway (22). It should be noted that a similar osmoregulation scenario has been well documented for the budding yeast Saccharomyces cerevisiae, in which the Sln1p-Ypd1p-Ssk1p phosphotransfer signaling pathway and the Hog1 MAPK cascade are crucially involved (6, 15, 20, 30, 32). In S. pombe, the downstream region of the MAP kinase cascade is less clear at present. However, recent studies have uncovered a basic leucine zipper (bZIP) type of transcription factor, Atf1 (also known as Gad7), which is a direct target of the Sty1 kinase (11, 25, 27, 31). In short, the osmoinducible transcription of the gpd1⁺ gene is greatly reduced in the wis1, sty1, and atf1 mutants, and consequently these mutants, as well as gpd1 mutants, exhibit an osmosensitive phenotype.

As mentioned above, we have been extensively studying osmoregulation in S. pombe (1, 18, 19, 33, 34). In this study, to gain new insight into the molecular mechanisms underlying osmoregulation in S. pombe, we attempted to isolate novel types of mutants, each of which show a osmosensitive phenotype. Here we isolated a set of S. pombe mutants that were found to be novel in that the mutational events are not apparently linked to the well-characterized MAP kinase-Atf1-Gpd1 pathway. Furthermore, one of the novel osmosensitive S. pombe mutants was characterized in detail.

MATERIALS AND METHODS

Strains, plasmids, and media.

The S. pombe strains used in this study are listed in Table 1. These strains were grown either in YPD medium containing 10 μg of adenine per ml or in SD medium, composed of 0.67% yeast nitrogen base without amino acids (Difco), supplemented with 2% glucose and other necessary growth requirements in standard amounts. Edinburgh minimal medium (EMM) and MEA medium, composed of 3% malt extract (Difco) and 2% agar, were also used (17). The plasmid used is also listed in Table 1.

TABLE 1.

List of S. pombe strains and plasmid relevant to this study

Strain or plasmid	Genotype or relevant remarks	Source
Strains
JY333	h− leu1 ade6-M216	M. Yamamoto
DG1	h− leu1 ade6-M216 ura4-D18 gpd1::ura4+	Our laboratory stock
DW746	h+ leu1 ade6-M210 ura4-D18 wis1::ura4+	Our laboratory stock
JY741	h− leu1 ade6-M216 ura4-D18	M. Yamamoto
JY879	h90leu1 ade6-M210 ura4-D18	M. Yamamoto
JY808	h90leu1 ade6-M210	Our laboratory stock
HAI001	h− leu1 ade6-M216 ura4-D18 ryh1::ura4+	This study
HAI002	h90leu1 ade6-M210 ura4-D18 ryh1::ura4+	This study
M6	h− leu1 ade6-M210 hos1	This study
M10	h+ leu1 ade6-M216 hos2	This study
M26	h− leu1 ade6-M210 hos3	This study
Plasmid
pLBDblet	Marker (LEU2)	Our laboratory stock

Isolation of hos (high-osmolarity-sensitive) mutants.

To isolate hos mutants, we used the mutagenesis procedure described previously by Ohmiya et al. (19). Briefly, S. pombe JY333 was mutagenized with N-methyl-N′-nitrosoguanidine (final concentration, 300 μg/ml) for 30 min at 30°C in 50 mM Tris-maleic acid (TM) buffer (pH 6.0) to give a 30 to 50% rate of survival. The mutagenized cells were washed three times with TM buffer and then once with yeast extract-peptone-dextrose (YPD) medium and allowed to recover from mutagenesis by growth in YPD medium for 4 h at 30°C. These were then plated onto YPD agar plates to generate master plates. After 3 days, colonies were replica plated from the master plates onto YPD agar containing 2 M sorbitol, and colonies unable to grow after 4 days under high-osmolarity conditions were identified and purified.

Assay for mating frequency.

Homothallic haploid cells (h⁹⁰) were grown in EMM to 4 × 10⁶ cells/ml at 30°C. Cells were washed with and reinoculated in nitrogen-free minimal medium (EMM−N) at a density of 8 × 10⁶ cells/ml and further incubated at 30°C. Aliquots were taken at 24 and 48 h after the reinoculation, and the mating frequency was calculated by the method described by Kunitomo et al. (13).

Glycerol assay.

Glycerol was analyzed enzymatically with a commercial glycerol assay kit (F-kit; Boehringer-Mannheim), as described previously (2). Exponentially growing cells in YPD medium were collected and resuspended in fresh YPD medium or in the same medium containing 0.9 M KCl and then incubated for 2 h at 30°C. The amount of glycerol was determined with the glycerol assay kit and calculated as an absolute amount of glycerol (micromole) per a certain number of cells giving 1 A₆₀₀ U of optical density.

Northern hybridization analysis.

Northern hybridization analysis was carried out as described previously (1). Exponentially growing cells in YPD medium were collected and resuspended in fresh YPD medium containing 0.9 M KCl. A total RNA fraction was prepared from the cells at each time. After denaturation with formamide-formaldehyde, RNA (5 μg) was analyzed on a 1.4% agarose gel containing formaldehyde, followed by alkali blotting onto Hybond-N⁺ (Amersham International). Hybridization was carried out with a ³²P-labelled probe which specifically encompassed the gpd1⁺ coding sequence at 65°C for 2 h in Rapid-hyb buffer, as recommended by the supplier (Amersham International).

Gene disruption.

For ryh1 gene disruption, the 368-bp SpeI-SpeI region in the ryh1⁺ gene on plasmid pNo1 was replaced with the S. pombe ura4⁺ gene to construct pHAI200 (see Fig. 4 and 5). The wild-type strain (JY741) was transformed with a 5.5-kb NheI-NheI fragment of pHAI200, and then stable Ura4⁺ transformants were selected. The chromosomal DNA was digested with StuI and then subjected to hybridization analysis with the 3.5-kb StuI-StuI fragment of pNo1 (ryh1⁺ probe) as a probe. The fragment carrying the ura4⁺ gene was also used as probe (ura4⁺ probe). It was revealed that the 3.3- and 1.5-kb fragments were hybridized with both ryh1⁺ and ura4⁺ gene probes for HAI001 as expected (see Fig. 5B).

FIG. 4.

Schematic representation of S. pombe genomic DNA fragments encompassing the ryh1⁺ gene and its ability to complement the mutant M6. (A) The DNA fragment cloned in each plasmid is shown. The regions in which the ura4⁺ cassette was inserted are also shown. (B) Strain M6 was transformed by each indicated plasmid, whereas JY333 was used as the wild type. These cells were streaked on an SD agar plate and an SD agar plate supplemented with 2 M glucose and then incubated at 30°C. Another SD agar plate streaked with the same set of strains was incubated at 37°C. After 3 days of incubation, the plates were photographed.

FIG. 5.

Construction of a deletion mutant of ryh1⁺ and its osmosensitivity for growth. (A) To construct the Δryh1 mutant, the ryh1⁺ coding region was replaced by the ura4⁺ marker. (B) This construct was confirmed by Southern hybridization analysis. The chromosomal DNA of Δryh1 (lanes 1 and 3) and wild-type (lanes 2 and 4) strains were digested with StuI and then subjected to Southern hybridization analysis with a ryh1⁺ probe (lanes 1 and 2) or a ura4⁺ probe (lanes 3 and 4). The osmosensitivity and temperature sensitivity phenotypic characteristics of the Δryh1 mutant were tested (C). After 4 days of incubation, the plates were photographed.

RESULTS

Isolation of a number of high-osmolarity-sensitive mutants.

An attempt was made to isolate osmoregulation-defective mutants of S. pombe by screening mutagenized cells for failure to grow on a high-osmolarity medium supplemented with 2 M sorbitol (i.e., YPD agar plates plus 2 M sorbitol). Of 40,000 colonies thus screened, 30 candidates were selected as putative high-osmolarity-sensitive mutants. They were then analyzed extensively by means of standard yeast genetics, including complementation analyses. Based on the results, they were classified into seven complementation groups. Among them, four groups were shown to be gpd1, atf1, sty1, and wis1 mutants, as anticipated (see the introduction). Others were clearly different from these ascribed alleles. Since we intended in this study to isolate novel high-osmolarity-sensitive mutants, we decided to further characterize these apparently novel ones, which were designated as hos (high-osmolarity-sensitive) mutants (namely, hos1, hos2, and hos3).

For further analyses, three strains, M6 for hos1, M10 for hos2, and M26 for hos3, were selected as representatives of strains with mutations in these alleles. It should be noted that these mutants have been purified genetically by repeated back-cross with the wild-type genetic background (JY333; see Table 1). That the osmosensitive phenotype of these mutants was recessive was also confirmed (data not shown). First, their osmosensitivity phenotypes were verified, as shown in Fig. 1. These three mutants as well as two well-characterized osmosensitive mutants (i.e., Δgpd1 and Δwis1) were streaked on YPD plates containing either 2 M glucose, 2 M sorbitol, or 0.9 M KCl. All of these mutants showed a growth defect in these high-osmolarity media (Fig. 1B to D). The mutant M6 exhibited a relatively low growth rate even on the standard YPD agar plate (Fig. 1A). The sensitivity of the mutant M26 to 0.9 M KCl was less evident (Fig. 1D). It was also found that the mutant M6 clearly exhibited a temperature sensitivity for growth at 37°C (Fig. 1E) (it is important to also note that a temperature-sensitive phenotype has been reported for the Δwis1 mutant, as shown in Fig. 1E).

FIG. 1.

Osmosensitivity of S. pombe strains for growth. The indicated strains of S. pombe were streaked onto YPD agar plates (A and E) and YPD agar plate supplemented with 2 M sorbitol (B), 2 M glucose (C), or 0.9 M KCl (D) and incubated at 30°C (A to D) or 37°C (E). After 4 days of incubation, the plates were photographed. Strains used as controls were JY333 (wild type [wild]) DW746 (Δwis1) and DG1 (Δgpd1).

hos mutants are novel types.

In the previous study, we have presented evidence that the expression of gpd1⁺ mRNA and the accumulation of intracellular glycerol are important for cells to grow on high-osmolarity medium (1, 19). To clarify whether the osmosensitive phenotypes of the isolated mutants were due to the defect in production of the osmoprotectant glycerol, we first examined the expression of gpd1⁺ mRNA in these mutants. As shown in Fig. 2, hos mutant cells growing exponentially in YPD medium were transferred into fresh medium supplemented with 0.9 M KCl, and total RNA was isolated from the cells harvested after the times indicated. Each RNA fraction was subjected to Northern hybridization analysis with an appropriate gpd1⁺ probe. Upon the shift to the high-osmolarity medium, for each hos mutant, the amount of gpd1⁺ mRNA increased substantially within 0.5 h as in the case of the wild type. In marked contrast, no or very small amounts of gpd1⁺ mRNA were detected in the Δgpd1 and Δwis1 mutants, respectively, regardless of the medium osmolarity. The latter observations are highly consistent with previous results (1, 18). Essentially, the same results were obtained even when 2 M sorbitol was used as an alternative osmotic solute (data not shown). It was thus found that the osmoinducible expression of gpd1⁺ mRNA appears not to be impaired in these hos mutants.

FIG. 2.

Northern hybridization analysis. Total RNA was isolated from the indicated strains after osmotic upshift for 0, 0.5, 1, and 1.5 h in YPD medium containing 0.9 M KCl and then subjected to Northern hybridization analysis by using a probe for the gpd1⁺ gene. In the lower panel, the ethidium bromide-stained agarose gel is shown as a control for the amounts of RNA loaded. Wild, wild type.

We then needed to measure directly the level of intracellular glycerol, since the result described above does not necessarily mean that glycerol is normally accumulated in the mutant cells in response to high-osmolarity stress. To examine this, the mutant cells were grown in YPD medium and then transferred into the same medium supplemented with 0.9 M KCl or unsupplemented. After incubation for 2 h, the intracellular accumulation of glycerol was measured for these cells (Fig. 3). In the case of the wild type, a marked accumulation of intracellular glycerol was observed upon the upshift to the high-osmolarity medium, as has been well documented (1, 19). In the Δgpd1 and Δwis1 mutant cells, the osmoresponsive intracellular accumulation of glycerol was greatly reduced, as described previously (1, 19). In the mutant cells isolated in this study, however, the intracellular accumulation of glycerol was found to occur as normally as it did in the wild type.

FIG. 3.

Glycerol production in response to osmotic upshift. Intracellular glycerol produced by the indicated strains was measured for cells grown for 2 h in either YPD (KCl−) or YPD supplemented with 0.9 M KCl (KCl+). Wild, wild type.

From these results, we confirmed that the hos1, hos2, and hos3 mutants are novel and that their osmosensitive phenotypes are not simply explained by the defect in production of the osmoprotectant glycerol. Therefore, extensive characterization of these hos mutants should shed light on the molecular mechanisms underlying the osmoregulation in this particular eukaryotic microorganism. This view encouraged us to further characterize these hos mutants and, with this end in mind, we selected the hos1 mutant strain (M6) for such detailed analyses.

Isolation of a gene that complements hos1.

The hos1 mutant is unable to grow on plates containing 2 M glucose and exhibits a temperature sensitivity for growth at 37°C (Fig. 1). In the hope of finding S. pombe genes that are relevant to the mutation, we screened a genomic DNA library to look for such clones on a multicopy plasmid that can suppress both of the phenotypic characteristics (i.e., osmosensitivity and temperature sensitivity). A number of plasmid clones were isolated as candidates, each of which carried a certain DNA insert with different lengths relative to each other. However, as judged from the results of restriction analyses and hybridization analyses, it was found that they all contain a common genomic DNA region. The simplified result is shown in Fig. 4A, in which the DNA inserts in two isolates (plasmids pNo1 and pNo20) are schematically shown. As shown in Fig. 4B, the 1.9-kb insert in pNo20 has the ability to complement both of the mutational lesions of M6. The nucleotide sequence of this insert was determined, and it was revealed that this region encompasses a known open reading frame, which was previously designated as the ryh1⁺ gene that encodes a small GTP-binding protein (10). To verify that the ryh1⁺ gene is indeed responsible for the observed complementation ability, the SpeI-SpeI region was replaced by the ura4⁺ marker on pNo1 to yield pHAI200. This plasmid had lost the complementation ability, as shown in Fig. 4B. From these results, we concluded that the ryh1⁺ gene is responsible for the complementation ability observed for the hos1 mutant.

We then wanted to determine whether hos1 is a mutant allele of the ryh1⁺ gene. To clarify this, pHAI201, in which the ura4⁺ marker was inserted at the HpaI site upstream of the ryh1⁺ gene, was also constructed, as shown in Fig. 4A. We confirmed that this particular clone still has the ability to complement hos1 (Fig. 4B). A Ura⁻ derivative of M6 was transformed by the NheI-NheI fragment encompassing the ura4⁺ marker as well as the ryh1⁺ gene, and the stable Ura4⁺ transformants were selected. It was revealed that all of them grew on YPD plates containing 2 M glucose and at 37°C. It should be noted that for several such Ura4⁺ Osm^r transformants, we confirmed by Southern hybridization that the NheI-NheI fragment was inserted into the right place, not elsewhere on the chromosome, via homologous recombination (data not shown). These results supported the idea that the hos1 mutation is in the ryh1⁺ gene.

Construction of Δryh1.

Since the ryh1⁺ gene is known to be dispensable for growth (10), a one-step gene disruption method utilizing a haploid strain (JY741) and the ura4⁺ marker was adopted to construct a Δryh1 mutant allele, in order to characterize the gene with special reference to osmoregulation, as shown in Fig. 5A. The resulting mutant (named HAI001) was confirmed by Southern hybridization to contain the disrupted gene, ryh1::ura4⁺, as expected (Fig. 5B). The phenotypic characteristics of the Δryh1 mutant was confirmed by showing that it exhibits osmosensitivity and temperature sensitivity for growth (Fig. 5C). Furthermore, these phenotypic characteristics were reverted to those of the wild type by introducing the ryh1⁺ gene on pNo20 (Fig. 5C). We then concluded that the ryh1⁺ gene is somehow implicated in the osmotic adaptation of S. pombe.

Disruption of the ryh1⁺ gene results in sterility.

To gain further insight into the function of the ryh1⁺ gene in terms of osmotic adaptation, several other phenotypic characteristics of the Δryh1 mutant were explored. First, vacuole biogenesis and cell wall integrity in the Δryh1 mutant were examined (see Discussion). We investigated the intracellular distribution of vacuoles in exponentially growing cells by visualization with a reagent (named FM4-64). We could not detect any noticeable difference with regard to the number and size of vacuoles between the Δryh1 mutant and wild-type cells (data not shown). To examine cell wall integrity, the sensitivities of Δryh1 mutant and wild-type cells to glucanase were compared, but no evident difference was detected in this respect (data not shown).

During the course of such examinations, however, we noticed that the Δryh1 mutant haploid strain may be severely sterile. To confirm this intriguing finding, we constructed a homothallic h⁹⁰ strain (HAI002) carrying the ryh1::ura4⁺ allele. Upon nitrogen starvation, the h⁹⁰ wild-type strain (JY808) was able to conjugate and form spores in up to 45% of the cells, while no spores were detected in the h⁹⁰ Δryh1 mutant cells, as quantitatively shown in Fig. 6. This defect in mating was suppressed by introducing the ryh1⁺ gene on a plasmid into the mutant strain, as also shown in Fig. 6. It was thus suggested that the ryh1⁺ gene plays a role, either directly or indirectly, in the mating processes of S. pombe.

FIG. 6.

Mating frequency of a Δryh1 and a wild-type strain. Homothallic strain HAI002 (h⁹⁰ Δryh1) was transformed with pLBDblet (closed circles) or pNo20 (open circles). JY808 transformed with pLBDblet (open squares) was used as the wild type. For these cells, mating frequency was assayed.

DISCUSSION

In S. pombe, glycerol appears to be the main compatible solute which is accumulated inside the cells in response to high medium osmolarity in order to maintain an osmotic homeostasis. Thus, the failure to accumulate glycerol should result in an osmosensitive phenotype for growth. In fact, a number of such osmosensitive mutants have already been isolated and characterized (e.g., gpd1, wis1, sty1, and atf1) (see the introduction). However, one can suppose that certain other mutational lesions may also result in an osmosensitive phenotype for growth. Keeping this assumption in mind, in this study we attempted to isolate novel types of high-osmolarity-sensitive (hos) mutants to gain insight into the molecular mechanisms underlying the complicated osmotic adaptation in S. pombe. Indeed, we succeeded in isolating such novel ones, the hos1, hos2, and hos3 mutants. Extensive characterization of these hos mutants should shed light on the relevant issues mentioned above. Studies concerned with this are currently under way in our laboratory. In the meantime, in this study we characterized the hos1 mutant at the molecular level by demonstrating that hos1 is a mutant allele of the gene known as ryh1⁺.

The ryh1⁺ gene encoding a GTP-binding protein (or G protein) of 201 amino acids and belonging to the ras superfamily was originally isolated by Hengst et al. by using the protein-coding region of the cloned S. cerevisiae YPT1 gene as a hybridization probe (10). As is well known, members of the Ras superfamily of proteins can be further classified into five subfamilies, namely, Ras, Rho, Rab (Ypt), Ran, and Arf. From the entire genome sequence of S. cerevisiae, 11 genes, whose protein products bear sequence features justifying their membership in the Rab (Ypt) subfamily, have been recently identified by Lazar et al. (14). On the basis of this current knowledge, it would be of interest to examine the relationship between the S. pombe Ryh1 sequence and those of the S. cerevisiae Rab (Ypt) family of sequences by constructing a reliable phylogenetic tree. As shown in Fig. 7, the results confirmed that S. pombe Ryh1 appears to belong to the Rab (Ypt) family and is closely related to the recently identified S. cerevisiae Ypt6 sequence (it is important to note that, from the comparison, Ryh1 appears to be distantly related to Ypt1). Such an inspection also revealed that among sequences in the current databases, the most homologous protein to Ryh1 is human Rab6 (73% identical in amino acids). In any case, based on the current knowledge of the Ras superfamily, the Rab (Ypt) subfamily of proteins is generally believed to play certain roles in the directed transport of vesicles between different intracellular compartments of the secretory pathway. Each Rab (Ypt) subfamily of proteins in a given species plays a role at each distinct step of the presumed multisteps of vesicular transport (3, 14).

FIG. 7.

Neighbor-joining tree indicating phylogenetic relationships between Ryh1 and various Ypt or Rab GTPases. Amino acid sequences of Ypt or Rab GTPases from S. cerevisiae (Sc), human (Hs), Arabidopsis (At), and S. pombe (Sp) were aligned, and a phylogenetic tree was constructed by the neighbor-joining method (21) by using the program CLUSTAL X (29).

Based on the fact that the ryh1⁺ gene encodes a member of the Rab (Ypt) family of proteins, how might one explain the osmosensitivity for growth, as observed in this study, of the hos1 (or Δryh1) mutant? As emphasized above, the possibility that Ryh1 is somehow implicated in the production of the osmoprotectant glycerol has been dismissed (Fig. 2 and 3). For S. pombe, it was recently suggested that some Ypt homologs (i.e., Ypt4 and Ypt7) seem to be involved in a process of vacuole fusion and fission (3, 5). Vacuole fusion and fission appear to be homeostatic mechanisms that restore the concentration of the cytosol, and vacuole fusion is a rapid and specific process of membrane fusion in response to external stimuli, including the medium osmolarity (5). Thus, we suspected that the hos1 (Δryh1) mutant might have a defect in the process of vacuole biogenesis. However, this appears not to be the case, as mentioned above (data not shown). We also suspected that Ryh1 might somehow be involved in the maintenance of cell wall integrity. But, the results of our glucanase treatment experimentation could not support this idea, as also mentioned (data not shown). Therefore, another plausible explanation(s) should be considered for the function of Ryh1 in relation to the osmosensitivity.

As mentioned above, Ryh1 seems to be closely related to S. cerevisiae Ypt6 and human Rab6, as far as their amino acid sequences are concerned. Both Ypt6 and Rab6 were suggested to play a role in the Golgi event of vesicular transport (14). Ryh1 might play a similar role, and addressing this issue by using the hos1 (Δryh1) mutant is of interest. In fact, in the original study on the Δryh1 mutant by Hengst et al. (10), the authors pointed out the possibility that an under glycosylation of invertase may occur in their Δryh1 S. pombe mutant. This is consistent with the current view that protein glycosylation processes are closely linked to vesicular transport processes. In this respect, we recently found that our Δryh1 mutant showed a phenotype of hypersensitivity to hygromycin B (25 μg/ml) and vanadate (4 mM) (28a). A similar phenotype has been reported for the S. pombe gms1 mutant that is defective in protein glycosylation (26, 28). It is thus tempting to speculate that the hos1 (Δryh1) mutant may have a defect in a process of protein glycosylation in such a way that the mutation affects the cell surface structure (or integrity). Such a mutational lesion might in turn result in the phenotype of high osmolarity sensitivity. Another intriguing finding, that the Δryh1 mutant is severely sterile, may also be explained by assuming that a process of conjugation (or cell-cell contact) is impaired in the mutant (Fig. 6), which likely has an altered cell surface structure.

In short, in this study we intended to isolate novel types of high-osmolarity-sensitive S. pombe mutants and succeeded in doing so. Characterization of these mutants will provide us with clues toward understanding the molecular mechanisms underlying the osmotic adaptation in S. pombe. But also, they provide us with new insight into the functions of relevant genes that are involved in other important cellular processes, as demonstrated for the ryh1⁺ gene encoding a GTP-binding protein. Other mutants, hos2 and hos3, are also of interest for further examination.