Negative Functional Interaction Between Cell Integrity MAPK Pathway and Rho1 GTPase in Fission Yeast

Abstract

Rho1 GTPase is the main activator of cell wall glucan biosynthesis and regulates actin cytoskeleton in fungi, including Schizosaccharomyces pombe. We have obtained a fission yeast thermosensitive mutant strain carrying the rho1-596 allele, which displays reduced Rho1 GTPase activity. This strain has severe cell wall defects and a thermosensitive growth, which is partially suppressed by osmotic stabilization. In a global screening for rho1-596 multicopy suppresors the pmp1⁺ gene was identified. Pmp1 is a dual specificity phosphatase that negatively regulates the Pmk1 mitogen-activated protein kinase (MAPK) cell integrity pathway. Accordingly, elimination of Pmk1 MAPK partially rescued rho1-596 thermosensitivity, corroborating the unexpected antagonistic functional relationship of these genes. We found that rho1-596 cells displayed increased basal activation of the cell integrity MAPK pathway and therefore were hypersensitive to MgCl₂ and FK506. Moreover, the absence of calcineurin was lethal for rho1-596. We found a higher level of calcineurin activity in rho1-596 than in wild-type cells, and overexpression of constitutively active calcineurin partially rescued rho1-596 thermosensitivity. All together our results suggest that loss of Rho1 function causes an increase in the cell integrity MAPK activity, which is detrimental to the cells and turns calcineurin activity essential.

RHO GTPases are conserved proteins that regulate the actin cytoskeleton organization in all eukaryots (Heasman and Ridley 2008). Additionally, in yeast and other fungi, these GTPases provide the coordinated regulation of cell wall biosynthesis and actin organization, which is required to maintain cell integrity and polarized growth (Levin 2005; Park and Bi 2007; Perez and Cansado 2010). The Schizosaccharomyces pombe genome contains six genes coding Rho GTPases. Among them, rho1⁺ is essential (Arellano et al. 1996). Rho1 function is mediated by its interaction with at least three different targets: the β(1,3)-glucan synthase (Arellano et al. 1996), which is responsible for the synthesis of the major cell wall component, and the kinases Pck1 and Pck2 (Arellano et al. 1999; Sayers et al. 2000). Through both kinases, Rho1 also regulates the cell wall synthesis. Rho2 also interacts with Pck2 and, therefore, both GTPases regulate the α-D-glucan synthesis through Pck2 (Katayama et al. 1999; Calonge et al. 2000). Lack of Rho1 is lethal, and this phenotype is not suppressed by osmotic stabilization (Arellano et al. 1997), suggesting that defective biosynthesis of the cell wall is not the unique cause of death. On the contrary, Rho2-less cells are viable, although they become slightly rounded and more sensitive to treatment with glucanases (Hirata et al. 1998). Rho2 and Pck2 participate in the activation of the cell integrity mitogen-activated protein kinase (MAPK) signaling pathway (Ma et al. 2006). This signaling cascade responds to different extracellular stress stimuli such as hyper- or hypotonic conditions, oxidative stress, cell wall damaging compounds, and glucose deprivation (Madrid et al. 2006, 2013; Barba et al. 2008), and is involved in the maintenance of cell integrity, cytokinesis, ion homeostasis, and vacuole fusion. The components of the MAPK cascade module are Mkh1 (MAPKKK), Pek/Shk1 (MAPKK), and Pmk1/Spm1 (MAPK) (Toda et al. 1996; Zaitsevskaya-Carter and Cooper 1997; Sugiura et al. 1998; Loewith et al. 2000). Single deletion of genes coding any of the above-mentioned components causes multiseptation, hypersensitivity to hypo- or hypertonic stress and to β(1,3)-glucanases, and defective vacuole fusion (Toda et al. 1996; Zaitsevskaya-Carter and Cooper 1997; Bone et al. 1998; Sugiura et al. 1999; Loewith et al. 2000). Pmk1 is structurally similar to Slt2/Mpk1 from Saccharomyces cerevisiae (Toda et al. 1996; Zaitsevskaya-Carter and Cooper 1997) and to the mammalian extracellular signal-regulated kinases (ERKs) (Roux and Blenis 2004). Several targets of Pmk1 MAPK have been described, including Atf1, the transcription factor that signals in the stress-activated MAPK pathway (SAPK), which includes Sty1/Spc1 (Takada et al. 2007); Nrd1, an RNA recognition motif (RRM)-type RNA-binding protein (Satoh et al. 2009); and the cell surface protein Ecm33 (Takada et al. 2010). It has been proposed that Nrd1 may serve as a novel mechanism for the regulation of myosin mRNA and cytokinesis by the Pmk1 pathway (Satoh et al. 2009).

Fission yeast dual-specificity phosphatase Pmp1 is the main negative regulator of Pmk1 (Sugiura et al. 1998; Madrid et al. 2007). Tyrosine phosphatases Pyp1 and Pyp2, and serine/threonin phosphatase Ptc1 are also able to associate in vivo and dephosphorylate activated Pmk1 (Madrid et al. 2007). Interestingly, Pyp1 and Pyp2 phosphatases also negatively regulate the stress-activated Sty1/Spc1 MAPK (Millar et al. 1995), and their expression is positively regulated by this MAPK and the transcription factor Atf1, creating a negative feedback loop (Degols et al. 1996; Madrid et al. 2007).

Calcineurin is a highly conserved calcium-dependent serine/threonine protein phosphatase that mediates the Ca²⁺-dependent signaling to a wide variety of cellular responses. In mammals, calcineurin regulates a variety of physiological processes, including T-cell activation, cardiac muscle development, skeletal muscle-fiber-type switching, apoptosis, long-term potentiation in learning and memory, neuronal plasticity, and oxidative stress (Steinbach et al. 2007). In S. cerevisiae calcineurin cooperates with the MAPK cell integrity pathway in response to cell wall damage. Upon cell stress, the calcineurin-activated transcription factor Crz1 immediately induces the expression of FKS2, coding for a glucan synthase catalytic subunit (Stathopoulos and Cyert 1997; Zhao et al. 1998), whereas maintenance of high levels of FKS2 expression under chronic cell wall stress is controlled by the MAPK cell integrity pathway (Zhao et al. 1998; Jung and Levin 1999). By contrast, in fission yeast calcineurin activates at least two distinct signaling pathways, the transcription factor Prz1-dependent branch and a Prz1-independent pathway that functions antagonistically with the Pmk1 MAPK pathway, regulating chloride ion homeostasis and the Ca²⁺ influx via the Cch1–Yam8 channel complex (Ma et al. 2011b). Calcineurin plays a functional role in the control of chloride ion homeostasis, cell polarity, mating, cytokinesis, spindle pole body positioning, and bipolar growth (Sugiura et al. 1998, 2002; Zhao et al. 1998; Jung and Levin 1999; Kume et al. 2011).

We have obtained a S. pombe thermosensitive mutant strain carrying a hypomorphic rho1-596 allele that causes cell death at high temperatures. Exhaustive characterization of this mutant has unveiled the existence of a functional relationship between Rho1 and calcineurin, which is antagonized by the cell integrity MAPK pathway.

Materials and Methods

Strains, growth conditions, and genetic methods

Standard S. pombe media and genetic manipulations were used (Moreno et al. 1991). All the strains used were isogenic to wild-type (wt) strains 972 h^– and 975 h⁺, and they are described in Supporting Information, Table S1. The strains were constructed by tetrad dissection or random spore germination method. Cells were usually grown in either rich medium yeast extract with supplements (YES) or minimal medium (EMM) with appropriate supplements (Moreno et al. 1991). Escherichia coli DH5α was used as host for propagation of plasmids. Cells were grown in LB medium supplemented with 100 µg/ml ampicillin when needed. Solid media contained 2% agar.

Recombinant DNA methods

All DNA manipulations were carried out by established methods. Enzymes were used according to the recommendations of the suppliers. S. pombe was transformed by the lithium acetate method (Ito et al. 1983). Error-prone PCR for obtaining rho1 open reading frame (ORF) mutants was performed as described previously for Cdc42 (Martin et al. 2007). rho1-596 strain identification and isolation was carried out as described (Martin et al. 2007). Screening for rho1-596 multicopy suppressors was performed using a S. pombe genomic library (pURSP1, American Type Culture Collection) to transform the mutant strain. Transformant clones were selected at 35°; the plasmids were recovered, and the DNA insert was sequenced.

Plasmids pALrho1⁺ (García et al. 2006), and nmt promoter-containing vectors pREP3X, pREP41X, and pREP81X (Forsburg 1993) were used for the overexpression of different genes (rho1⁺, pmp1⁺, ppb1⁺, etc.). ppb1ΔC was created by cloning ppb1⁺ in the pBluescript-KS and amplifying the fragment from amino acid 1–445 that was subsequently cloned into pREP3X.

In vivo analysis of Rho1 activity

The expression vector pGEX–Rhotekin binding domain (RBD) (Manser et al. 1998) was used to transform E. coli and produce GST fused to the mammalian RBD for RhoA. The fusion protein was produced according to the manufacturer’s instructions and immobilized on glutathione sepharose (GS) beads. The amount of GTP-bound Rho1 was determined using a pull-down assay as described (Coll et al. 2003). Extracts from strains carrying integrated HA-rho1⁺ or HA-rho1-596 were obtained as described (Calonge et al. 2003), using 500 µl of lysis buffer (50 mM Tris pH 7.5, 20 mM NaCl, 0.5% NP-40, 10% glycerol, 0.1 mM DTT, 1 mM NaF, 2 mM Cl₂Mg, containing 100 µM phenyl methanesulfonyl fluoride, leupeptin, pepstatin, and aprotinin). Cell extracts (3 mg total protein) were incubated with 10 μg of GST–RBD protein coupled to GS beads for 2 hr, washed four times, and resolved in 12% SDS–PAGE gels. Proteins were transferred to PVDF membranes, incubated with mouse monoclonal anti-HA antibody (12CA5), and revealed with antimouse HRP-conjugated secondary antibodies and the the ECL detection kit. Total HA-Rho1 levels in whole-cell extracts (45 μg of total protein) were monitored by Western blot.

Microscopy techniques

For calcofluor white staining of S. pombe cells, exponentially growing cells were harvested, washed, and resuspended in a solution of a calcofluor solution (0.1 mg/ml) for 5 min at room temperature. After washing with water, cells were observed in a microscope with the corresponding UV filter. The vital stain methylene blue (MB) was used to visualize live and dead cells using bright-field microscopy.

Purification and detection of activated Pmk1

Cell homogenates were prepared under native conditions employing chilled acid-washed glass beads and lysis buffer (10% glycerol, 50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% Nonidet NP-40, plus specific protease and phosphatase inhibitors from Sigma Chemical). The lysates were cleared by centrifugation at 13,000 rpm for 15 min. Pmk1-HA6His was purified by using Ni²⁺-NTA agarose beads (Qiagen). The purified proteins were resolved in 10% SDS–PAGE gels, transferred to nitrocellulose filters, and incubated with either monoclonal mouse anti-HA or polyclonal rabbit antiphospho-p44/42 antibodies (Cell Signaling Technologies). The immunoreactive bands were revealed with antimouse or antirabbit HRP-conjugated secondary antibodies (Sigma Chemical) and the ECL detection kit.

Analysis of cell wall polysaccharides

Labeling and fractionation of cell wall polysaccharides were performed as described (Arellano et al. 1997). S. pombe cells were grown to log phase. Then 10 ml of cells (1 × 10⁷/ml) were labeled by addition of [U-14C]glucose (0.5 µCi/ml) and incubated for an additional 6 hr. Cells were harvested. Total glucose incorporation was monitored in an aliquot by measuring the radioactivity in 10% cold trichloroacetic acid-insoluble material. Mechanical breakage of cells was done in a Fast-Prep apparatus as described (Calonge et al. 2003). Cell walls were pelleted at 1000 × g for 5 min and washed three times with 5% NaCl and twice with 1 mM EDTA. Aliquots (100 µl) of the total cell wall were incubated with 100 units of Zymolyase 100T (Seikagaku Kogio) or Quantazyme (Quantum Biotechnologies) for 36 hr at 30°. Aliquots without enzymes were included as controls. Samples were centrifuged, and the supernatant and pellet were counted separately. The supernatant from the Zymolyase 100T reaction was considered β-glucan plus galactomannan, and the pellet was considered α-glucan. The supernatant from the Quantazyme reaction was considered β-glucan, and the pellet was considered α-glucan plus galactomannan.

Other methods

In vitro glucan synthase assay was performed as described (Arellano et al. 1997).

Detection of calcineurin activity in vivo was done as described (Deng et al. 2006). We crossed rho1-596 strain with a strain carrying an integrated copy of pKK2 plasmid into the ura4 locus on the genome. pKK2 contains a green fluorescent protein (GFP) reporter gene with three calcineurin-dependent response elements (CDREs) following a nmt1 promoter without its cis element (Kume et al. 2011).

Plate assays for growth sensitivity to different compounds were performed by spotting the appropriate dilutions of log-phase-growing wild-type and mutant strains of S. pombe on YES or EMM solid media containing 2% (w/v) bacto-agar supplemented with the corresponding compounds MgCl₂, CaCl₂, KCl, hydrogen peroxide, sorbitol (Sigma-Aldrich), and FK506 (LC Laboratories). Plates were incubated at 28° for 3 days. All experiments were repeated at least three times.

Results

rho1-596 is a thermosensitive strain with severe cell wall defects

Rho1 is an essential GTPase that regulates cell wall biosynthesis and actin organization in S. pombe (Arellano et al. 1996, 1997). To further explore the in vivo functions of Rho1 we generated point mutant alleles of rho1⁺ by degenerate PCR of the ORF as previously described for cdc42⁺ (Martin et al. 2007). Mutated ORFs were integrated in the genome to replace rho1⁺ and selected for thermosensitive growth. DNA sequencing of the rho1 allele carried in one of the selected thermosensitive strains revealed a single mutation that substitutes the cysteine residue at position 17 for arginine. This residue maps to the P-loop domain implicated in GTP binding (Saraste et al. 1990). The cysteine is conserved in Rho2, Rho4, and Rho5 GTPases from S. pombe and in other Rho GTPases from different species, including Rho1 from S. cerevisiae and RhoA from mammalian cells. Interestingly, it is not present in S. pombe Rho3 or Cdc42. Cells bearing the rho1-596 allele showed a severe growth defect at temperatures over 32° and most cells were dead at 35° (Figure 1, A and B). The presence of an osmotic stabilizer such as sorbitol 1.2 M prevented cell death but not completely (Figure 1A). rho1-596 death resembled that caused by Rho1 depletion or overexpression of the negative allele rho1T20N (Arellano et al. 1996, 1997; Nakano et al. 1997), and it is different from the lysis of cell wall mutants such as mok1-664, carrying a defective thermosensitive α-glucan synthase (Katayama et al. 1999) at 34° (Figure 1B). rho1-596 cells become permeable to methylene blue staining but do not lose their intracellular content as mok1-664 cells do. To see whether the defects of the rho1-596 strain were caused by a decrease in Rho1 activity, we first transformed wild type and rho1-596 with the integrative plasmid pJK148 carrying HA-rho1⁺ or HA-rho1-596, respectively, expressed from the endogenous rho1⁺ promoter. Total HA-Rho1 or HA-Rho1-596 levels were detected by Western blot using anti-HA antibodies and the levels of active GTP-bound HA-Rho1 or HA-Rho1-596 were measured by GST-RBD pull-down. Our results showed that Rho1-596 levels were lower than wild-type Rho1 in cells grown at 28° and this difference was more dramatic after 90-min incubation at 35° (Figure 1C). There was also less GTP-bound Rho1-596 than wild-type Rho1 at both temperatures. Therefore rho1-596 is a hypomorphic thermosensitive allele with lower Rho1 activity even at the permissive temperature.

Figure 1.

Phenotype of S. pombe rho1-596 strain. (A) Growth of wild-type (wt) and rho1-596 cells spotted at an A₆₀₀ of 1.0 and 1/4 dilutions in YES and YES + 1.2 M sorbitol. Plates were incubated at different temperatures for 3 days. (B) Differential interference contrast of rho1-596 cells grown at 28° and 35° (upper panels), and bright-field microscopy of rho1-596 and mok1-664 cells stained with methylene blue to visualize live and dead cells (lower panels). Cells were grown at 34° for 3 hr. (C) Levels of GTP-bound Rho1 in wild-type and rho1-596 cells. Extracts from wild-type and rho1-596 cells grown at 28° and transferred at 28° or 35° for 90 min were precipitated with GST-RBD and blotted with anti-HA antibodies (upper panel). Total HA-Rho1 or HA-Rho1-596 in cell lysates was visualized by Western blot. Actin was used as loading control. (D) Cell wall composition of wt and rho1-596 strains. Cells were incubated at 25° and ¹⁴C-glucose was added 6 hr before harvesting the cells. Values are the mean of three independent experiments with duplicate samples. Error bars represent standard deviations for the total carbohydrate values. (E) β(1-3)-D-glucan synthase activity in wt and rho1-596 strains. Extracts were prepared from cells grown at 25° in YES for 14 hr. Specific activity is expressed as milliunits per milligram of protein. Values are the mean of at least three independent experiments with duplicated samples, and error bars represent standard deviations. (F) Growth of wt and rho1-596 cells transformed with the plasmid pALrho1⁺ carrying rho1⁺ under its own promoter. Cells were spotted at an A₆₀₀ of 4.0 and 1/4 dilutions in EMM and plates incubated at different temperatures for 3 days.

Since Rho1 regulates cell wall biosynthesis in fission yeast (Arellano et al. 1996, 1997), we also analyzed the cell wall composition in both rho1-596 mutant strain and wild-type cells grown at 25°. As shown in Figure 1D, there was a mild but significant decrease in the total amount of glucose incorporated into the cell wall of rho1-596 cells. This decrease was mainly due to a drop in β-D-glucan (from 22.7 to 17.2%) that was partially compensated for by a small increase in galactomannan. We also measured the in vitro activity of the β(1,3)-D-glucan synthase enzyme in wild-type and rho1-596 cells. This enzyme is regulated by GTP-bound Rho1 and its activity correlates directly with the level of active Rho1 GTPase (Arellano et al. 1996). As expected, a marked decrease of the enzymatic activity was observed in rho1-596 as compared to wild-type cells even at the permissive temperature (25°; Figure 1E). This difference was even more severe when GTP was added to the in vitro reaction.

As expected, rho1-596 thermosensitive growth defect was completely rescued by expression of wild-type rho1⁺ from pALrho1⁺, a multicopy plasmid with its own promoter (Figure 1F). Overexpression of rho1⁺ from pREP41X, a plasmid carrying a thiamine-inducible promoter (Forsburg 1993), also partially suppressed rho1-596 thermosensitive growth defect (Figure S1A). On the contrary, overexpression of the dominant negative allele rho1T20N did not suppress the thermosensitivity and was more deleterious to the mutant than to wild-type strain (Figure S1A). Interestingly, overexpression of the constitutively active allele rho1G15V did not suppress the thermosensitivity of the mutant (Figure S1A), suggesting that GTP-GDP cycling is required for the suppression.

We also asked whether an increase in the expression of the Rho1-GEFs could suppress the defect of cells carrying rho1-596. Overexpression of rgf1⁺, encoding the main Rho1-GEF (García et al. 2006), rescued the growth of rho1-596 mutant cells, but overexpression of rgf3⁺, encoding the cytokinesis specific Rho1-GEF (Tajadura et al. 2004; Morrell-Falvey et al. 2005; Mutoh et al. 2005), did not (Figure S1A). These results suggest that rho1-596 growth defects are not restricted to cytokinesis.

Overexpression of other Rho coding genes, such as rho4⁺, rho5⁺, and cdc42⁺, had no effect on the thermosensitivity of rho1-596 cells, whereas overexpression of rho2⁺, and mainly of rho3⁺, aggravated this phenotype (Figure S1B). Taken together, these results suggest that (a) rho1-596 thermosensitivity is due to a defect caused by a general decrease of Rho1 activity, (b) Rho1 function requires GTP-GDP cycling, and (c) other Rho proteins cannot replace Rho1.

Pmp1 phosphatase is a multicopy suppressor of rho1-596 growth thermosensitivity

To identify new effectors of Rho1 we performed a screening for multicopy suppressors of rho1-596 thermosensitive growth. S. pombe genomic DNA library pURSP1 was used to transform rho1-596 cells and transformant clones were selected for growth at 35°. Several clones contained a plasmid that included the pmp1⁺ gene coding for a dual specificity phosphatase, which negatively regulates the Pmk1 cell integrity MAPK pathway (Sugiura et al. 1998). We next cloned pmp1⁺ in a pREP41X plasmid and confirmed that overexpression of this gene partially suppressed rho1-596 thermosensitivity (Figure 2A) and that cell death at 35° was reduced from 95 to 40% (Figure 2B).

Figure 2.

Pmp1 is a rho1-596 suppressor. (A) Growth of wild-type (wt) and rho1-596 transformed with the plasmid pREP41X empty or carrying pmp1⁺ as insert. Cells were spotted at an A₆₀₀ of 4.0 and 1/4 dilutions in EMM without thiamine and incubated at 28° or 35° during 3 days. (B) Differential interference contrast images of rho1-596 cells transformed with the plasmid pREP41X or pREP41X-pmp1⁺. Cells were grown at 28° and shifted to 35° for 5 hr. Lower panel shows the percentage of cell death (n = 500 cells). (C) Growth of wt, rho1-596, pmp1Δ, and rho1-596 pmp1Δ double mutant strains at different temperatures. Cells were spotted at an A₆₀₀ of 1.0 and 1/4 dilutions in YES medium. (D) Differential interference contrast and fluorescent images of the same cells grown at 28° and stained with calcofluor. Right panel shows the percentage of cell death (n = 500 cells).

To corroborate the genetic interaction between rho1⁺ and pmp1⁺, we crossed rho1-596 strain with a strain lacking Pmp1. While pmp1Δ mutant strain is not thermosensitive, the double mutant rho1-596 pmp1Δ strain could not grow at temperatures >28°, and >60% of cells were dead at this temperature (Figure 2, C and D). These results confirmed the existence of a functional relationship between Pmp1 and Rho1 and suggest that Pmp1 function is necessary when Rho1 activity is compromised.

The Pmk1 cell integrity MAPK is antagonistic to Rho1

Pmk1 is the only known target for Pmp1 dual phosphatase (Sugiura et al. 1998). To study whether the positive effect caused in rho1-596 cells by overexpression of pmp1⁺ was due to Pmk1 inactivation or was mediated by an unknown Pmp1 target, we first constructed a rho1-596 pmk1Δ double mutant strain. This strain was able to grow even at 35°, implying that lack of Pmk1 improves rho1-596 survival (Figure 3A). To exclude the possibility that Pmp1 might exert other effects besides Pmk1 inhibition, we overexpressed pmp1⁺ in the rho1-596 pmk1Δ double mutant strain. In these conditions, the suppression caused by increasing Pmp1 production was not additive to that caused by pmk1Δ (Figure 3A). Additionally, we obtained a rho1-596 pmp1Δ pmk1Δ triple mutant strain and confirmed that elimination of the phosphatase in the absence of Pmk1 was not deleterious for rho1-596 cells (Figure 3B). Therefore our results suggest that Pmk1 MAPK activity is detrimental to rho1-596 strain.

Figure 3.

Pmk1 activation is deleterious for rho1-596 growth. (A) Overexpression of pmp1⁺ in wild-type (wt), rho1-596, and rho1-596 pmk1Δ double mutant strain. Cells were spotted at an A₆₀₀ of 4.0 and 1/4 dilution in EMM without thiamine and incubated at the indicated temperature during 3 days. (B) Growth of wt, rho1-596, rho1-596 pmk1Δ, rho1-596 pmp1Δ, and rho1-596 pmp1Δ pmk1Δ triple mutant strain. (C–E) Growth of wt, rho1-596, and several double mutant strain: rho1-596 rho2Δ (C), rho1-596 pyp1Δ (D), and rho1-596 sty1Δ (E). Cells were spotted at an A₆₀₀ of 1.0 and 1/4 dilution in YES medium and incubated at the indicated temperatures during 3 days.

To corroborate this conclusion we made several genetic experiments. First we made the double mutant rho1-596 rho2Δ. Rho2 GTPase is an upstream activator of the Pmk1 MAPK signaling pathway (Ma et al. 2006) but also regulates the cell wall α-glucan biosynthesis (Calonge et al. 2000). Elimination of Rho2 improved the growth of the rho1-596 mutant strain at 32° and 34° (Figure 3C). These results thus suggest that Pmk1 down-regulation improved growth of the rho1 hypomorphic mutant even when cell wall biosynthesis was still defective. Simultaneous deletion of other Rho GTPases such as Rho3, Rho4, or Rho5 did not cause any effect in rho1-596 cells (data not shown). We further corroborated these results by constructing a rho1-596 pyp1Δ double mutant strain. Pyp1 is a tyrosine phosphatase that negatively regulates both the cell integrity MAPK Pmk1, and the stress-activated MAPK Sty1 (Madrid et al. 2007). As can be seen in Figure 3D, the rho1-596 pyp1Δ double mutant was unable to grow over 28°. This negative effect was likely due to Pmk1 activation and not to Sty1 activation, since the rho1-596 sty1Δ double mutant strain was very sick and only grew at 25° (Figure 3E). Taken together, these results indicate that the cell integrity MAPK signaling is not beneficial but rather detrimental when Rho1 function is diminished.

Enhanced growth of rho1-596 cells in the absence of Pmk1 activity is independent of cell wall biosynthesis

To further explore the functional relationship between Rho1 and Pmk1 signaling, we analyzed whether the effect caused by Pmk1 was related to the cell wall defects present in rho1-596 cells. As previously shown in Figures 1A and 3A, either osmotic stabilization with sorbitol 1.2 M or Pmk1 deletion suppressed rho1-596 thermosensitivity only partially, and the cells were able to grow at 34°. However, rho1-596 pmk1Δ double mutant cells in the presence of 1.2 M sorbitol were able to grow up to 37°, as the wild-type cells (Figure 4A). Therefore the partial suppression caused by deletion of Pmk1 was additive with the suppression caused by sorbitol, suggesting that they act through independent mechanisms. Furthermore, cell wall polysaccharide analysis of the rho1-596 pmk1Δ double mutant strain did not show significant differences as compared to that of rho1-596 cells (Figure 4B), suggesting that the suppression caused by Pmk1 deletion is not due to an increase in cell wall polysaccharides. We also measured the levels of total and GTP-bound Rho1-596 in cells with or without Pmk1 and did not observe any significant changes (Figure 4C). Therefore, Rho1-596 activity does not seem to be affected by Pmk1 function.

Figure 4.

The absence of Pmk1 and osmotic stabilization independently improve rho1-596 growth. (A) Growth of wild-type (wt), rho1-596, pmk1Δ, and rho1-596 pmk1Δ cells in YES (upper panels) or YES supplemented with 1.2 M sorbitol (lower panels). Cells were spotted at an A₆₀₀ of 1.0 and 1/4 dilutions and incubated at the indicated temperature during 3 days. (B) Cell wall composition of wt pmk1Δ, rho1-596 and rho1-596 pmk1Δ strains. Cells were incubated at 25° and ¹⁴C-glucose was added 6 hr before harvesting the cells. Values are the mean of three independent experiments with duplicate samples. Error bars represent standard deviations for the total carbohydrate values. (C) Levels of total and GTP-bound HA-Rho1-596 in cells with or without Pmk1. Extracts from cells grown at 25° were precipitated with GST-RBD and blotted with anti-HA antibodies (middle panel). Total HA-Rho1-596 in cell lysates was visualized by Western blot (upper panel). Actin was used as loading control.

Basal Pmk1 phosphorylation is increased in rho1-596 cells

Overexpression of pmp1⁺ partially suppresses the thermosensitivity of rho1-596 cells by decreasing Pmk1 activity. To determine whether the cell integrity MAPK pathway is activated in the rho1-596 mutant strain, we examined the basal level of Pmk1 phosphorylation in rho1-596 cells using antiphospho-P44/42 antibodies that recognize only dually phosphorylated, and hence activated, Pmk1 (Sugiura et al. 1999). The results shown in Figure 5A revealed that rho1-596 cells had increased basal Pmk1 phosphorylation as compared to wild-type cells under normal conditions. This increase was almost completely abolished in rho1-596 rho2Δ or rho1-596 pck2Δ cells (Figure 5A). These data suggest that rho1-596 cells undergo constitutive cell integrity pathway activation likely due to the cell wall defects present in these cells, and this activation is mainly mediated by Rho2 and Pck2 (Figure 5A, Barba et al. 2008). The data could also suggest that Rho1 is a negative regulator of Pmk1 activity. However, if this were the case, Pmk1 activation in response to different stimuli would be enhanced in rho1-596 cells with respect to wild type, and we did not see significant differences in the kinetics of Pmk1 response to KCl, hypotonic stress, or oxidative stress (Figure 5B). Therefore Rho1 is not a negative regulator of the MAPK pathway.

Figure 5.

Cell integrity Pmk1 pathway is activated in rho1-596 cells. (A) Cells from wild-type (wt), rho2Δ, pck2Δ, rho1-596, rho1-596 rho2Δ, and rho1-596 pck2Δ, carrying a HA6His-tagged chromosomal version of pmk1, were grown in YES medium to mid-log phase and Pmk1-HA6His was purified by affinity chromatography. Activated Pmk1 was detected by blotting with anti-(phosho-p44/42) antibody, and total Pmk1 was detected with anti-HA antibody as loading control. (B) Stress-induced activation of Pmk1 wt and rho1-596 strains. Cells were grown as above and treated with 0.6 M KCl, or 1 mM hydrogen peroxide. For hypotonic stress treatment, cells were grown in YES medium plus 0.8 M sorbitol to mid-log phase and shifted to the same medium without sorbitol. At timed intervals either activated or total Pmk1 were detected by blotting as indicated above. (C) S. pombe rho1::ura4⁺ cells transformed with pREP41X-HA-rho1⁺ were grown at 25° in minimal medium without thiamine and transferred to medium with thiamine (time 0). At 2-hr intervals activated Pmk1 and HA-Rho1 were detected by blotting.

To validate that a decrease in Rho1 function is causing the increase in Pmk1 activity of rho1-596 cells and discard that it is not an allele-specific effect, we analyzed whether Rho1 depletion caused an increase in Pmk1 phosphorylation by using a strain carrying the thiamine-repressible pREP41X-HArho1 plasmid where the endogenous rho1⁺ has been deleted (Rincon et al. 2006). Repression of Rho1 by adding thiamine resulted in an increase of Pmk1 phosphorylation that was already observed after 2 hr (Figure 5C). Lethality in these cells was not observed before 8 hr at 25°. Therefore there is an increase in Pmk1 phosphorylation concomitant to Rho1 decrease.

rho1-596 cells are dependent on calcineurin activity

A typical phenotype of Pmk1 MAPK activation is the hypersensitivity to MgCl₂ and FK506, an immunosuppressant drug that causes inactivation of the highly conserved calcineurin phosphatase (Sugiura et al. 1998, 2002). We observed that rho1-596 cells were considerably more sensitive than wild-type cells to both MgCl₂ and FK506 (Figure 6, A and B). In both cases, the hypersensitivity was independent of osmotic stabilization and was suppressed by deletion of either rho2⁺ or pmk1⁺, which eliminates cell integrity MAPK signaling. These results led us to consider the hypothesis that Rho1 might have a function in concert with calcineurin that is antagonized by Pmk1 MAPK. We tried to obtain the double mutant of rho1-596 with either ppb1Δ or cnb1Δ cells, which lack the genes coding for the catalytic or the regulatory subunit of calcineurin, respectively. However, both double mutants were not viable even in the presence of an osmotic stabilizer (data not shown). Thus, calcineurin probably plays an essential role when the activity of Rho1 is decreased.

Figure 6.

rho1-596 cells are hypersensitive to MgCl₂ and FK506. Cells from wild-type (wt), rho1-596, rho1-596 rho2Δ, rho1-596 pmk1Δ, rho2Δ, and pmk1Δ were grown to mid-log phase and spotted at an A₆₀₀ of 2.0 and 1/4 dilutions onto YES or YES + 1.2 M sorbitol plates supplemented with (A) MgCl₂ and (B) FK506 at the indicated concentrations. Cells were incubated for 3 days at 28°. (C) Calcineurin activity estimated by immunoblotting with anti-GFP antibodies of wt and rho1-596 cells carrying the GFP gene under a promoter with three CDRE elements. Cells grown at 25° were incubated for 2 hr at 32° and, where indicated, FK506 was added for 1 hr to see the specific decrease in GFP protein. (D) Wt and rho1-596 cells transformed with the pREP3X plasmid carrying rho1⁺, ppb1⁺, and ppb1ΔC and grown in EMM medium. Cells were spotted at an A₆₀₀ of 4.0 and 1/4 dilutions onto the plates and were incubated for 3 days at the indicated temperatures.

We examined rho1-596 calcineurin activity in vivo by using a background strain carrying an integrated green fluorescent protein (GFP) reporter gene fused to three CDRE motifs (Kume et al. 2011). rho1-596 cells have four times increase over the wild-type level of calcineurin activity as measured by immunoblotting with anti-GFP antibodies (Figure 6C). We also analyzed whether increasing even more the calcineurin activity could compensate for Rho1 defects by transforming rho1-596 cells with a plasmid carrying a calcineurin version without the C-terminal portion, which is known to produce a constitutively active Ca²⁺/calmodulin-independent phosphatase (Sugiura et al. 1998). As we expected, ectopic increase in calcineurin activity partially compensated for the thermosensitivity of rho1-596 cells (Figure 6D). All together our results suggest that calcineurin signaling is activated and required for survival when Rho1 activity is compromised. Crz1 is the calcineurin responsive transcription factor (Stathopoulos and Cyert 1997) that orchestrates most of the calcineurin-dependent cellular responses in S. cerevisiae. In response to cell wall damage Crz1 induces the expression of FKS2, coding for one of the glucan synthase catalytic subunits (Zhao et al. 1998). However, deletion of the S. pombe ortholog to Crz1 transcription factor, Prz1, did not affect rho1-596 cells (Figure S2). Therefore, the functional interaction described here between Rho1 and calcineurin is not dependent on transcription activation by Prz1.

Discussion

Rho1 is an essential protein that regulates cell wall biosynthesis and actin cytoskeleton organization. In this work we have obtained and characterized for the first time a S. pombe mutant strain carrying a rho1 loss-of-function allele. Cells carrying the rho1-596 allele cannot grow at temperatures >32°. They show a very low GTP-bound Rho1-596 level, which in turn triggers a decrease in the activity of the β(1,3)-D-glucan synthase and consequent drop in the cell wall β-D-glucan content. However, while relevant, this is not the only defect elicited by the rho1-596 mutation because osmotic stabilization prevented cell death only partially. Death of rho1-596 is quite similar to that caused by decreasing the amount of Rho1 in a “switch-off” strain (Arellano et al. 1997) or by reducing Rho1 activity through overexpression of a rho1T20N dominant negative allele (Arellano et al. 1997). These strains do not die by rapidly losing the intracellular content, as occurs to mutants with defects in cell wall components (Cortes et al. 2005, 2012) or cell wall organization (De Medina-Redondo et al. 2010). While overexpression of rho1⁺ completely rescued rho1-596 growth at 36°, overexpression of other Rho GTPases coding genes like rho2⁺ or rho3⁺ could not rescue its thermosensitive growth and was deleterious, suggesting that both GTPases play different cellular roles than Rho1. Intriguingly, this is not the case in S. cerevisiae where Rho2 may share some functions with Rho1 (Schmelzle et al. 2002).

A multicopy suppressor screening, corroborated by a synthetic negative genetic interaction, revealed that function of Pmp1 dual specificity phosphatase is required when Rho1 activity is compromised. Since Pmk1 MAPK is the only known target described for Pmp1 (Sugiura et al. 1998), this suppressive effect is likely due to inactivation of the cell integrity MAPK pathway, as was confirmed by showing that elimination of Pmk1 improved the growth of the rho1-596 mutant strain at 34° even in the absence of Pmp1. The above finding is quite unexpected since MAP kinase orthologs in S. cerevisiae and other fungi are part of the signaling pathway required for maintenance of cell wall integrity (Levin 2005, 2011).

When S. cerevisiae cells are exposed to environmental stresses that perturb the cell wall, Rho1 activates Pkc1 and the MAPK Slt2/Mpk1, resulting in actin depolarization and cell wall remodeling. Moreover, disruption of signaling through this MAPK cascade compromises the integrity of the cell wall (Levin 2011). In striking contrast, elimination of the cell integrity MAPK pathway in fission yeast improved the survival of rho1-596 cells, which have severe cell wall defects, and the benefit of the cell integrity MAPK elimination for the survival of rho1-596 cells was additive to the presence of an osmotic stabilizer in the growth medium. Therefore Pmk1 activation is causing some effect that is deleterious for cells when Rho1 activity is defective.

Although unexpected, the above differences between S. pombe and S. cerevisiae MAPK orthologs are not unique. Even if only a fraction of the S. pombe genetic interaction network has been screened, a number of orthologous genes that have evolved different genetic relationships and play different or additional roles in fission vs. budding yeast have been described (Frost et al. 2012; Koch et al. 2012). The significant different features of fission yeast MAPK integrity pathway when compared to the S. cerevisiae pathway might imply that both cascades undergo different cellular functions. Indeed, S. cerevisiae Slt2/Mpk1 is only activated by cell wall stresses (mainly hypo-osmotic stress), whereas fission yeast Pmk1 becomes activated by multiple stressing situations and also during cell separation (Barba et al. 2008). Additionally, S. cerevisiae Slt2/Mpk1 regulates the production of various carbohydrate polymers of the cell wall as well as their polarized delivery to the site of cell wall remodeling (Levin 2011). This function has never been shown for fission yeast Pmk1 MAPK. Moreover, whereas Slt2/Mpk1 cooperates with calcineurin in the induction of the glucan synthase gene FKS2, fission yeast calcineurin and the Pmk1 MAPK signaling pathway play antagonistic roles in the regulation of Cl⁻ homeostasis and the Ca2⁺ signal transduction pathway (Sugiura et al. 1998, 2002; Ma et al. 2011b). Finally, Rho2, and not Rho1, is the major GTPase triggering the MAPK cascade in S. pombe (Ma et al. 2006). Indeed, we describe here that rho1-596 cells, with low Rho1 activity, showed increased basal Pmk1 phosphorylation as compared to the wild-type cells. This feature seems contradictory to previous results showing that overexpression of the constitutively active rho1G15V allele increased Pmk1 phosphorylation (García et al. 2009) and that Rgf1, described as specific Rho1 GEF, also can trigger activation of the cascade (García et al. 2009). However, all these observations can be explained considering that the Pmk1 pathway is activated both by hypertonic and hypotonic stresses, which are expected in gain-of-function or loss-of-function Rho1 alleles, respectively (Barba et al. 2008). Constitutively active rho1G15V alleles have severe cell wall perturbations (Arellano et al. 1996) and rho1-596 cells also have a defective cell wall that might be causing a sustained basal activation of the cell integrity pathway.

rho1-596 cells were as hypersensitive to MgCl₂ as cells lacking calcineurin (Sugiura et al. 1998, 2002). Certainly, rho1-596 cells were also hypersensitive to FK506 and abolition of Pmk1 signaling suppressed both phenotypes. We initially considered the possibility that rho1-596 cells had lower calcineurin activity and this would result in increased Pmk1 basal activity. On the contrary, we observed a higher calcineurin activity level in rho1-596 cells. Additionally, exposure of fission yeast cells to FK506 does not trigger an increase or decrease in basal phosphorylation of Pmk1 (Figure S3). With all the results presented here, we propose a model in which a decrease in Rho1 activity causes cell wall defects that increase basal Pmk1 activity and this is detrimental for the cells. At the same time, a decrease in Rho1 activity function causes a rise in calcineurin activity, which is required for survival (Figure 7).

Figure 7.

Scheme of Rho1 relationship with Pmk1 and calcineurin. Decrease of Rho1 function causes cell wall defects that activate the cell integrity MAPK Pmk1 signaling, and this is detrimental for the cells. At the same time, decrease of Rho1 function causes a rise in calcineurin activity that is required for cell survival.

In different yeasts, cell wall stress activates an ER stress-like response that requires calcineurin and calcium influx but not Crz1 transcription factor for survival (Bonilla and Cunningham 2003; Scrimale et al. 2009). In S. pombe, calcineurin is also essential to prevent lethal response to ER stress (Lustoza et al. 2011). Our data support that rho1-596 cells require calcineurin, but not Prz1 transcription factor, and high-affinity calcium channel for survival. Therefore, it is tempting to propose that chronic cell wall defects caused by low Rho1 activity can trigger an ER stress-like response in fission yeast. However, further studies will be required to prove this hypothesis since Rho1 participates in other processes besides cell wall biosynthesis like actin organization (Levin 2005; Park and Bi 2007; Perez and Cansado 2010), down-regulation of TORC1 activity (Yan et al. 2012), membrane fluidity homeostasis (Lockshon et al. 2012), and endocytosis (Prosser and Wendland 2012). Similarly, calcineurin participates in a plethora of physiological processes. A recent genome-wide screening identified 72 viable deletion strains hypersensitive to FK506 that belonged to many different gene ontology (GO) categories (Ma et al. 2011a). Fifteen of those deletion strains were hypersensitive to micafungin, an inhibitor of the (1,3)-β-D-glucan synthase. Further studies will be needed to characterize the possible participation of these genes in the common essential process that requires Rho1 and calcineurin.