A second osmosensing signal transduction pathway in yeast. Hypotonic shock activates the PKC1 protein kinase-regulated cell integrity pathway.

Yeast cells respond to hypertonic shock by activation of a (MAP) mitogen-activated protein kinase cascade called the (HOG) high osmolarity glycerol response pathway. How yeast respond to hypotonic shock is unknown. Results of this investigation show that a second MAP kinase cascade in yeast called the protein kinase C1 (PKC1) pathway is activated by hypotonic shock. Tyrosine phosphorylation of the PKC1 pathway MAP kinase increased rapidly in cells following a shift of the external medium to lower osmolarity. The intensity of the response was proportional to the magnitude of the decrease in extracellular osmolarity. This response to hypotonic shock required upstream protein kinases of the PKC1 pathway. Increasing external osmolarity inhibited tyrosine phosphorylation of the PKC1 pathway MAP kinase, a response that was blocked by BCK1-20, a constitutively active mutant in an upstream protein kinase. These results indicate that yeast contain two osmosensing signal transduction pathways, the HOG pathway and the PKC1 pathway, that respond to hypertonic and hypotonic shock, respectively.

Yeast cells respond to hypertonic shock by activation of a (MAP) mitogen-activated protein kinase cascade called the (HOG) high osmolarity glycerol response pathway. How yeast respond to hypotonic shock is unknown. Results of this investigation show that a second MAP kinase cascade in yeast called the protein kinase C1 (PKC1) pathway is activated by hypotonic shock. Tyrosine phosphorylation of the PKC1 pathway MAP kinase increased rapidly in cells following a shift of the external medium to lower osmolarity. The intensity of the response was proportional to the magnitude of the decrease in extracellular osmolarity. This response to hypotonic shock required upstream protein kinases of the PKC1 pathway. Increasing external osmolarity inhibited tyrosine phosphorylation of the PKC1 pathway MAP kinase, a response that was blocked by BCK1-20, a constitutively active mutant in an upstream protein kinase. These results indicate that yeast contain two osmosensing signal transduction pathways, the HOG pathway and the PKC1 pathway, that respond to hypertonic and hypotonic shock, respectively.
Osmotic stress induces specific cellular responses that include changes in the activity of solute transporters (1,2) and enzymes involved in solute accumulation (3,4), the expression of genes encoding enzymes required for solute synthesis (5)(6)(7)(8), stress resistance (5,9), and cell wall structure (10). Despite their importance for cell growth and survival, the signaling mechanisms responsible for mediating osmotic stress-specific responses are not nearly as well understood as those which mediate responses to ligands such as growth factors or hormones. Our understanding of how eukaryotic cells sense and respond to changes in osmolarity has been helped recently by studies of this problem in the budding yeast Saccharomyces cerevisiae.
In yeast, a protein kinase cascade called the HOG 1 pathway (11) plays a central role in mediating cellular responses to an increase in external osmolarity. This pathway is defined by the HOG1 (11) and PBS2 (11)(12)(13) genes encoding members of the MAPK (mitogen-activated protein kinase) and MAPKK (MAP kinase kinase) family, respectively (14,15). Addition of NaCl or sorbitol to increase the osmolarity of the medium induces yeast to accumulate glycerol (6) and thereby restore the osmotic gradient across the cell membrane. This response, which involves increased expression of the glycerol-3-phosphate dehydrogenase gene GPD1 (5-7), is blocked in a hog1⌬ mutant (7). Other responses to an increase in osmolarity such as reorientation of cell growth and division (16) and induction of gene expression (17) are also defective in hog1⌬ and pbs2⌬ mutants. HOG pathway activation involves increased phosphorylation of a Hog1p tyrosine residue conserved among all MAP kinases which is required for growth at high osmolarity (17). Mammalian cells contain structural and functional homologs of Hog1p, suggesting that the HOG pathway is conserved among eukaryotes (18 -21).
In its natural environment, yeast cells are exposed to not only increases but also decreases in osmolarity. Although the HOG pathway has a clear role in mediating cell responses to increases in osmolarity, little is known about how yeast sense and respond to decreases in osmolarity. Yeast cells contain four known MAP kinase cascades (22)(23)(24). One of these, referred to here as the PKC1 pathway, is mediated by a protein kinase C-like protein encoded by the gene PKC1 (25). Other protein kinases on the PKC1 pathway have been identified using different genetic approaches and placed into a linear pathway that proceeds downward from Pkc1p to MAPKKK (called Bck1p or Slk1p) (26 -28) to two MAPKK (Mkk1p and Mkk2p) (29) to MAPK (called Mpk1p or Slt2p) (30,31). A comparison of deletion mutants in the HOG pathway to those in the PKC1 pathway reveal opposite phenotypes. For example, a hog1⌬ MAPK mutant grows in low but not high osmolarity medium while a mpk1⌬ MAPK mutant grows in high but not low osmolarity medium (30), a phenotype exacerbated by growth at elevated temperature, i.e. 37°C. Mutants in other genes of the PKC1 pathway show a phenotype similar to that of mpk1⌬ (26,29,(31)(32)(33). Although there are other possible explanations, this observation is consistent with a model in which the PKC1 pathway, like the HOG pathway, mediates an osmotic signal and induces cellular responses required for growth at the new (lower) osmolarity. At the time this work was initiated there was no known activating signal for this kinase cascade. In this report we test the hypothesis that the PKC1 kinase cascade is an osmosensing signal transduction pathway which responds to hypotonic shock as an activating signal.

EXPERIMENTAL PROCEDURES
Materials-The yeast strain used for most experiments was YPH102 (MATa ura3 leu2 his3 ade2 lys2) (34), into which different plasmids were introduced by LiAc-based transformation (35) with selection on uracil-deficient medium for URA3 carried on the plasmid. Other strains used in experiments are described in the legends of the figures in which they were used. Plasmids and PKC1 pathway mutant strains were obtained from Michael Snyder (Yale), Kunihiro Matsumoto (Nagoya University), and Carl Mann (Centre d'Etudes de Saclay, Gif-sur-Yvette). The MPK1-hemagglutinin (HA) plasmid (36) rescued the mpk1⌬ mutant phenotype of reduced growth in low osmolarity medium indicating that addition of the HA epitope did not interfere with the normal function of the Mpk1p. A similar plasmid carrying a mutation which codes for a substitution of phenylalanine for the conserved tyrosine in the MPK1 gene (pMPK1-HA Y192F) (36) did not rescue the mpk1⌬ phenotype and was deleterious to a wild-type strain transformed with the plasmid (data not shown).
Growth Conditions-Cultures of plasmid-bearing yeast were grown overnight in uracil-deficient medium and then grown to log phase on the day of the experiment in YEPD (2% Bacto-peptone, 2% glucose, 1% yeast extract), 20% YEPD (0.4% Bacto-peptone, 0.4% glucose, 0.2% yeast extract), or 20% YEPD containing additional solute (NaCl, sorbitol, or glucose) to raise the osmolarity. In some experiments, the osmolarity was increased by the addition of a concentrated solution of solute or decreased by addition of either water or conditioned medium with a lower osmolarity. For the latter type of experiment, two log phase cultures were grown to the same cell density, one grown in 20% YEPD, the other in 20% YEPD plus 1 M sorbitol. Cells were removed from the former culture by centrifugation and the resulting supernatant added to the latter culture to lower osmolarity without changing the levels of nutrients or other components in the growth medium. As a control for the effects of dilution, conditioned medium from a culture grown in 20% YEPD ϩ 1 M sorbitol was added to a 20% YEPD ϩ 1 M sorbitol culture.
Preparation of Cell Extracts and Immunoblot Analysis-Activation of the PKC1 pathway was detected using a previously described procedure for immunoblot analysis of MAPK (Mpk1p) tyrosine phosphorylation (37). Briefly, after different experimental manipulations to change the external osmolarity, yeast cell cultures were quickly chilled, and cells were collected by rapid centrifugation. Ice-cold buffer containing protease and phosphatase inhibitors (50 mM Tris HCl, pH 7.5, 1% sodium deoxycholate, 1% Triton X-100, 0.1% sodium dodecyl sulfate, 50 mM NaF, 5 mM sodium pyrophosphate, 0.1 mM sodium vanadate, 0.05% phenylmethylsulfonyl fluoride, 0.05 g/l aprotinin, 0.01 g/l leupeptin, and 0.01 g/l pepstatin) was added to the cell pellet and the cells lysed using glass bead agitation in a mini-beadbeater. The resulting homogenate was centrifuged for 15 min in a microcentrifuge, and proteins from the high speed supernatant were resolved by SDS-polyacrylamide gel electrophoresis, loading the same amount of protein (usually 20 g) for each sample. Protein concentrations were measured by the Bradford (38) method with bovine serum albumin as a standard. After transfer of proteins to nitrocellulose (0.2 m, Schleicher & Schuell) using a semidry blotting apparatus (Bio-Rad), tyrosine phosphorylation of different yeast proteins was detected by incubation of the membrane blot with an anti-phosphotyrosine monoclonal antibody (Upstate Biotechnology, Inc.), followed by an alkaline phosphatase-conjugated goat anti-mouse IgG antibody (Promega), and visualization of immune complexes with the chromogenic alkaline phosphatase substrate 5-bromo-4-chloro-3-indoyl phosphate/nitro blue tetrazolium. To detect the Mpk1p MAPK containing the HA epitope (39), immunoblots were probed with the 12CA5 anti-HA monoclonal antibody (Babco or Boehringer Mannheim) and then a horseradish peroxidase-conjugated sheep anti-mouse IgG antibody (Amersham Corp.). Immune complexes were detected using an enhanced chemiluminescence procedure for detecting peroxidase activity (Amersham Corp.).

Hypotonic Shock Induces the Tyrosine Phosphorylation of
Mpk1p-Specific extracellular signals activate MAP kinases by inducing their phosphorylation on a single conserved threonine and a nearby tyrosine (40,41). For Hog1p and the PKC1 pathway MAPK Mpk1p, like other MAP kinases, phosphorylation of this conserved tyrosine residue is required for pathway function because substitution of the non-phosphorylatable residue phenylalanine for the conserved tyrosine blocks pathway-specific responses (17,31). Therefore, to determine whether hypotonic shock activates the PKC1 pathway, we exposed the yeast strain YPH102 and YPH102 mpk1::HIS3 (mpk1⌬) to decreases in external osmolarity and then assayed cell proteins for tyrosine phosphorylation by immunoblot analysis with an antibody to phosphotyrosine. One minute after the osmolarity of the medium was lowered by reducing the concentration of sorbitol from 1 M to 0.2 M, an increase in tyrosine phosphorylation of a single band with an apparent size of 68 kDa was observed in the wild type but not in the mpk1⌬ strain ( Fig. 1 (left)). A gene fusion coding for Mpk1p tagged at the COOH terminus with a HA epitope (38) was introduced on a high copy 2 plasmid into an mpk1⌬ strain. Decreasing the osmolarity stimulated the tyrosine phosphorylation of a band that migrated more slowly than that in the wild type strain (YPH102). In the mpk1⌬ strain transformed with a plasmid identical to the MPK1-HA- Left, anti-phosphotyrosine immunoblot analysis of phosphorylation in response to hypotonic shock. Cells were exposed for 1 min to either no change in osmolarity (Iso) or a decrease in osmolarity (Hypo) before rapid cooling and preparation of cell extracts (see "Experimental Procedures"). The four strains tested were YPH102 with the control 2 plasmid pRS426 (34) (lanes 1 and 2), YPH102 mpk::HIS3 (mpk1⌬) with the control plasmid pRS426 (lanes 3 and 4), pMPK1-HA ( lanes 5 and 6), and pMPK1-HA Y192F (lanes 7 and 8). Right, anti-HA immunoblot analysis of Mpk1p-HA in cell extracts. Shown are the immunoblots of samples identical to those in lanes 5-8. containing plasmid with the exception of a point mutation in MPK1 which substitutes a phenylalanine for the conserved tyrosine, no band is seen in the anti-phosphotyrosine immunoblot. An immunoblot of the same cell extracts with an anti-HA epitope antibody ( Fig. 1 (right)) revealed an immunoreactive protein found in the MPK1-HA and mutant MPK1-HA containing strains, which had the same mobility as the tyrosine phosphorylated band. We interpret these observations to mean that the PKC1 pathway MAPK Mpk1p is tyrosine phosphorylated in response to hypotonic shock.
Time Course of Mpk1p Phosphorylation-MAP kinase pathways induce rapid (Ͻ1 min) changes in MAPK tyrosine phosphorylation in response to specific signals. As shown in Fig. 2, lowering external osmolarity induced an increase in Mpk1p tyrosine phosphorylation that occurred within 15 s of the stimulus and persisted for 10 -15 min. After 30 min when the tyrosine phosphorylation of Mpk1p had dropped to nearly prestimulus level, a further reduction in the osmolarity of the medium caused by addition of water induced a second rapid increase in Mpk1p tyrosine phosphorylation. Because cells containing a single chromosomal copy of MPK1 produce a small amount of Mpk1p which makes detection of phosphotyrosine difficult, this and following experiments were carried out with cells containing multiple copies of MPK1-HA.
Osmotic Dependence of MAP Kinase Phosphorylation-To extend and confirm the observation that Mpk1p and Hog1p respond in opposite fashion to osmotic changes, the in vivo tyrosine phosphorylation of both kinases was measured after changing the external osmolarity to a range of higher and lower levels. Specifically, a culture was grown to log phase in medium containing 1 M sorbitol, and then the osmolarity decreased or increased by addition of water with varying concentrations of sorbitol. Cells were collected 1 min after the osmotic change and MAP kinase tyrosine phosphorylation measured as before using an immunoblot procedure. Compared to the control cells (Fig. 3, marked by an asterisk (*)) where external osmolarity was unchanged, decreasing osmolarity induced an increase in Mpk1p phosphorylation that was proportional to the magnitude of the osmotic shock. Increasing osmolarity induced an increase in tyrosine phosphorylation of a band that we has the same mobility relative to molecular weight standards as that which was previously identified as Hog1p and, as expected from previous results, was absent in hog1⌬ cells (not shown). We noted that it was easier to detect both basal and high osmolarity-induced increases in tyrosine phosphorylation in Hog1p in cells containing a high copy Mpk1p (or Mpk1p-HA) plasmid, although anti-Hog1p immunoblot analysis (17) showed that the amount of Hog1p was unchanged under these different conditions (data not shown). This Mpk1p overexpression-induced increase in the amount of tyrosine phosphorylated Hog1p is blocked in mutants lacking protein kinases upstream of Mpk1p on the PKC pathway (see below). The physiological significance and explanation of this phenomenon is unknown.
Solute Independence of the Hypotonic Response-In experiments decribed above, external osmolarity was changed by altering the concentration of sorbitol. To determine if the phosphorylation of Mpk1p was due to osmotic changes or a sorbitol specific response, we tested whether changes in the concentration of other solutes, namely glucose or NaCl, would also activate Mpk1p phosphorylation. As shown in Fig. 4, cells were grown to log phase in 20% YEPD plus 1 M glucose (or 0.5 M NaCl) and then shifted to the medium with the same (Iso), lower (Hypo), or higher (Hyper) concentration of glucose (or NaCl). In both cases, Mpk1p and Hog1p were tyrosine-phosphorylated in response to hypotonic (Hypo) and hypertonic (Hyper) shock, respectively. Therefore, the Mpk1p phosphorylation responses are independent of the varied solute.
Osmotic Regulation of Mpk1p Phosphorylation Involves Upstream Kinases in the PKC Pathway-To determine whether Mpk1p phosphorylation by hypotonic shock is mediated through the PKC1 pathway, we measured this response in different mutant strains. As shown in Fig. 5 (top), hypotonic shock-induced tyrosine phosphorylation of Mpk1p-HA was not detectable in strains containing deletions in the genes that encode protein kinases upstream of Mpk1p on the PKC1 path- way. These include a protein kinase C mutant (pkc1⌬) (25), a MAPKKK mutant (bck1⌬) (27), and a MAPKK mutant (mkk1⌬ mkk2⌬) (29). The failure to detect Mpk1p-HA tyrosine phosphorylation could be explained by PKC1 pathway-dependent expression of Mpk1p-HA. However, immunoblot analysis with the anti-HA antibody (Fig. 5, bottom) showed that the amount of Mpk1p-HA was independent of the upstream kinases in the PKC1 pathway.
Tyrosine phosphorylation of Mpk1p was inhibited in cells exposed to an increase in osmolarity of the medium while Hog1p tyrosine phosphorylation increased (Fig. 6). To test whether these responses involved the PKC1 pathway, this experiment was repeated using a BCK1-20 mutant which is reported to encode a constitutively active form of the Bck1p protein kinase (26). This strain no longer shows the high osmolarity-induced decrease in Mpk1p phosphorylation. Note that the high osmolarity-induced increase in Hog1p phosphorylation was relatively unaffected by this mutation.

Osmosensing MAP Kinase Pathways in Yeast-Our results
show that there are two osmosensing signal transduction pathways in yeast, each containing structurally similar protein kinases (22,24). The symmetry between the pathways is striking. The HOG pathway genes HOG1 and PBS2 are required for cell growth at high osmolarity and high osmolarity rapidly induces a transient, PBS2-dependent hyperphosphorylation of the Hog1p MAP kinase (11). The PKC1 pathway genes PKC1, BCK1 (SLK1), MKK1/MKK2, and MPK1 (SLT2) are required for cell growth at low osmolarity (26, 29 -33), and low osmolarity rapidly induces a transient, MKK1/MKK2-dependent hyperphosphorylation of the Mpk1p MAP kinase (this study). Mpk1p kinase activity is rapidly elevated in cells exposed to a hypotonic shock (36). The HOG pathway has a fairly well defined role in the cellular response to an increase in osmolarity. Based on the phenotypes of hog1⌬ and pbs2⌬ mutants, the HOG pathway is required for high osmolarity-stimulated transcription of specific genes (7,17) leading to increased synthesis of the principal osmolyte glycerol (11) and general stress resistance (17). The PKC1 pathway is required for constructing a cell wall. Cells without PKC1 die by cell lysis (32,33). Deletion mutations in BCK1 (SLK1), MKK1/MKK2, or MPK1 (SLT2) have similar phenotypes: cell lysis that is accentuated by growth at higher temperatures. This temperature-sensitive cell lysis phenotype is suppressed by growth on high osmolarity medium and is correlated with a decrease in glucan content of the cell wall (10,42). The mechanism responsible for the weakened cell walls in PKC1 pathway mutants is not known with any certainty but may involve defects in polarized vesicle secretion/cell growth (43) or changes in glucan content (10). Therefore, one possible role of the PKC1 pathway is to regulate cell wall properties in response to changes in external osmolarity. This type of physiological response has been observed in fungi. The constant growth rate of the fungus Achyla bisexualis in medium of different osmolarity is correlated with changes in the mechanical properties of their cell wall with a stronger wall at low osmolarity than at high osmolarity (44).
The complex phenotype of PKC1 pathway mutants suggests that this pathway responds to physiological signals beside changes in external osmolarity. Besides the sensitivity to low osmolarity, such mutants are altered in cell morphogenesis (27,45). Mutants lacking the PKC1 pathway MAP kinase kinase kinase BCK1 (SLK1) are sensitive to starvation with defects indicative of a failure to exit the vegetative growth cycle (27). Compared to wild-type (BCK1 ϩ (SLK1 ϩ )) cells, bck1⌬ (slk1⌬) mutants do not accumulate glycogen, fail to undergo meiosis, are heat shock-sensitive, continue to form buds in stationary phase cultures, and lose viability in nutrientpoor medium (45). These phenotypes are independent of the osmolarity of the medium (45). These defects in growth control suggest that the PKC1 pathway has a role in nutrient sensing. The PKC1 pathway is required for growth at elevated temperature and the Mpk1p kinase activity is strongly activated by exposure of cells to higher temperature (36). How functions of the PKC1 pathway such as osmosensing, temperature-sensing, and nutrient sensing are coordinated with each other remains to be determined.
An important aspect of the two yeast osmosensing MAP kinase pathways is that similar pathways appear to exist in cells from other eukaryotes including mammals. In the case of the HOG pathway, Hog1p shows a high degree of similarity in amino acid sequence to a subgroup of MAP kinases, several members of which show increased tyrosine phosphorylation in cells exposed to an increase in osmolarity. Two mammalian members of this subgroup, p38 (18) and JNK1 (19), have been expressed in a yeast hog1⌬ strain and shown to complement the high osmolarity-sensitive growth phenotype of this mutant. The PKC1 pathway Mpk1p is closely related in amino acid sequence to a second subgroup of MAP kinases that includes the mammalian ERK2 (p44 mapk ) and ERK1 (p42 mapk ) (15). Strikingly, studies in a human intestinal cell line show that both of these MAP kinases show increased tyrosine phosphorylation after exposure of cells to a decrease in external osmolarity (46). The low osmolarity-sensitive growth phenotype of an mpk1⌬ mutant and a bck1⌬ (slk1⌬) mutant are complemented by expression in yeast of a Xenopus ERK2 MAP kinase and a mammalian MAP kinase kinase kinase (MEKK), respectively (31,47). Although studies of osmosensing pathways in mammals and yeast are just beginning, these similarities encourage the idea that other functions of such pathways (48,49) will also be conserved.
Cross-regulation between Osmosensing Pathways-The presence in a single cell of two different signaling pathways (HOG and PKC1) that are regulated in opposite directions by changes in external osmolarity raises several questions. For example, do these pathways regulate each other? We have used a genetic approach to address this question and found no evidence of synergy or suppression in the growth phenotype of double mutants nor was the growth phenotype of mutants in one pathway affected by overexpression of a protein kinase in the other pathway (data not shown). These data suggest that the two osmosensing pathways act independently of the other in supporting growth at different osmolarity. In wild-type cells with an intact PKC1 pathway (Fig. 5, left two lanes), decreasing external osmolarity caused a decrease in Hog1p tyrosine phosphorylation. PKC1 pathway mutants started out with a lower basal level of tyrosine phosphorylation of Hog1p. For each of the mutants examined, this amount of Hog1p phosphorylation did not change after a decrease in the osmolarity of the medium. These data suggest that the PKC1 pathway affects signaling through the HOG pathway when cells are exposed to a decrease in external osmolarity. The physiological significance of this apparent cross-talk between osmosensing pathways will require more information about the currently unknown targets of the PKC1 pathway.