QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Vol. 17, Issue 4, 1620-1631, April 2006

This Article Abstract Full Text (PDF) All Versions of this Article: E05-10-0933v1 17/4/1620    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by García, P. Articles by Sánchez, Y. Search for Related Content PubMed PubMed Citation Articles by García, P. Articles by Sánchez, Y.

Rgf1p Is a Specific Rho1-GEF That Coordinates Cell Polarization with Cell Wall Biogenesis in Fission Yeast

Instituto de Microbiología Bioquímica, CSIC/Universidad de Salamanca and Departamento de Microbiología y Genética, Universidad de Salamanca, Campus Miguel de Unamuno, 37007 Salamanca, Spain

Submitted October 7, 2005; Revised December 16, 2005; Accepted January 10, 2006
Monitoring Editor: Anne Ridley

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Rho1p regulates cell integrity by controlling the actin cytoskeleton and cell wall synthesis. We have identified a new GEF, designated Rgf1p, which specifically regulates Rho1p during polarized growth. The phenotype of rgf1 null cells was very similar to that seen after depletion of Rho1p, 30% of cells being lysed. In addition, rgf1⁺ deletion caused hypersensitivity to the antifungal drug Caspofungin and defects in the establishment of bipolar growth. rho1⁺, but none of the other GTPases of the Rho-family, suppressed the rgf1 phenotypes. Moreover, deletion of rgf1⁺ suppressed the severe growth defect in rga1⁺ null mutants (a Rho1-GAP, negative regulator). Rgf1p and Rho1p coimmunoprecipitated and overexpression of rgf1⁺ specifically increased the GTP-bound Rho1p; it caused changes in cell morphology, and a large increase in (1,3)-glucan synthase activity. These effects were similar to those elicited when the hyperactive rho1-G15V allele was expressed. A genetic relationship was observed between Rgf1p, Bgs4p ([1,3]-glucan synthase), and Pck1p (protein kinase C [PKC] homologue); Bgs4p and Pck1p suppressed the hypersensitivity to Caspofungin in rgf1 mutants. Rgf1p localized to the growing ends and the septum, where Rho1, Pck1p, and Bgs4p are known to function. Our results suggest that Rgf1p probably activates the Rho functions necessary for coordinating actin deposition with cell wall biosynthesis during bipolar growth, allowing the cells to remodel their wall without risk of rupture.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Fission yeast cells are rod-shaped and grow in a polarized manner at the cell ends. Immediately after cell division, the daughter cells initiate growth in a monopolar manner from the cell end that preexisted before cell division (the old end). After a point in G2, cells initiate growth from the new end (the end created by cell division) in a process known as new end take off (NETO), so that they grow in a bipolar mode up to mitosis (Mitchison and Nurse, 1985; Hayles and Nurse, 2001). Fission yeast is a useful model system for studying cell wall biosynthesis and how this fits in the complex morphogenetic processes required for the cell shape to be attained.

The Schizosaccharomyces pombe cell wall consists mainly of three polysaccharides, (1,3)-glucan, (1,3)-glucan, and galactomannoproteins, all of which form a large complex. Their coordinated synthesis represents an essential step in the assembly of a functional cell wall to ensure cell integrity (for a review, see Duran and Perez, 2004). Among the polysaccharides, (1,3)-glucans are the most prevalent (50–54% of the wall) and it is generally accepted that they are the main structural components responsible for cell wall rigidity (Manners and Meyer, 1977). (1,3)-glucan is the first polymer to be synthesized in S. pombe regenerating protoplasts (Osumi et al., 1989) and in the spore wall (Martin et al., 2000) and hence the regulation of this polysaccharide may be a key step in the sequential assembly of the other cell wall components. The enzymatic system that catalyzes the synthesis of this polysaccharide is (1,3)-glucan synthase (GS). GS is composed of at least two fractions: the catalytic moiety of the enzyme and the regulatory component. The catalytic subunit of GS is encoded by at least four genes: cps1⁺/bgs1⁺ (Le Goff et al., 1999; Liu et al., 2000b, 2002; Cortes et al., 2002), bgs2⁺ (Martin et al., 2000; Liu et al., 2000a), bgs3⁺ (Martin et al., 2003), and bgs4⁺ (Cortes et al., 2005). All of them code for essential proteins at different stages in the cellular life cycle. In addition to the catalytic subunit, the small GTP-binding protein Rho1p is an essential regulatory subunit (Arellano et al., 1996; Nakano et al., 1997). Rho1 acts as a binary switch by cycling between an inactive GDP-bound and an active GTP-bound conformational state. Rho1p stimulates GS in its GTP-bound prenylated form, providing a rationale for an understanding of the mechanism through which the cell can switch (1,3)-glucan synthesis on and off by interconverting the GDP and GTP forms of Rho1p.

To maintain intracellular osmolarity and to produce cell shapes other than spheres, cell wall expansion must be focused on particular regions. S. pombe uses both microtubules and the actin cytoskeleton for this purpose (for reviews, see Yarm et al., 2001; Chang and Verde, 2004; Gachet et al., 2004). It has been proposed that microtubules (MTs) would act to localize key proteins involved in setting up polarized growth or to localize secretion, or even to localize actin itself to the cortex. Actin is strictly required for cell growth and is assembled in three types of structures: actomyosin rings, actin cables, and actin patches. Both cables and patches are reorganized during the cell cycle and are focused around the areas of cell growth (Marks and Hyams, 1985). The cables serve as trackways along which both actin patches (Pelham and Chang, 2001) and presumably also myosin motors with their associated cargos move to the poles or the equator for cell growth (Win et al., 2001). Actin patches are dense membrane-associated structures possibly involved in localized cell wall synthesis. In fission yeast regenerating protoplasts, their localization coincides precisely with active sites of cell wall deposition (Kobori et al., 1989).

Rho1p provides a link between polarized growth and cell wall biosynthesis (Arellano et al., 1997; Nakano et al., 1997), and it belongs to a family of small GTPases that are key regulators in morphogenesis, polarity, movement, and division processes (reviewed in Jaffe and Hall, 2005). The fission yeast Rho family includes Cdc42p and Rho1p through Rho5p. Rho1p localizes to sites of polarized growth, the cell poles, and the septum (Arellano et al., 1997; Nakano et al., 1997) and activates the abovementioned cell wall-synthesizing enzyme GS (Arellano et al., 1996); Rho1 also regulates the organization of F-actin patches (Arellano et al., 1997), and it binds directly to the PKC family of protein kinases, Pck1p and Pck2p, functioning as a positive regulator of these (Arellano et al., 1999b; Sayers et al., 2000). However, little is known about the proteins that turn Rho1p on and off in the cell. Rho GTPase regulators such as GEFs (GDP-GTP exchange factors) modify the nucleotide-bound state of the GTPase and contain protein–protein interaction domains that could be important for GTPase localization, activation, and stabilization, and thus for interaction with its effectors (Gulli and Peter, 2001; Rossman et al., 2005). S. pombe contains seven genes bearing a Rho-GEF domain: scd1⁺, gef1⁺, gef2⁺, gef3⁺, rgf1⁺, rgf2⁺, and rgf3⁺ (Iwaki et al., 2003). Of these, scd1⁺ and gef1⁺ are Cdc42p-specific GEF(s) and Rgf3p has been described to function as a GEF for Rho1p. rgf3⁺ is an essential gene and regulates cell wall -glucan biosynthesis through the GTPase Rho1p, in particular during cytokinesis (Tajadura et al., 2004). Previous studies have shown that Rho1p depletion causes cell death that cannot be prevented by an osmotic stabilizer (Arellano et al., 1997). However, Rgf3p depletion was prevented by 1.2 M sorbitol (Tajadura et al., 2004). This intriguing phenomenon suggests that in the absence of Rgf3p, but in the presence of an osmotic support, Rho1p could be activated in some other way. Accordingly, we hypothesized that this function of Rho1p would be regulated by other GEF(s) (Iwaki et al., 2003). Here we demonstrate that Rgf1p specifically activates Rho1p. Our data support a model in which Rgf1p would coordinate actin deposition at polarized sites with cell wall biosynthesis, allowing the cells to remodel their wall without risk of rupture.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Media, Reagents, and Genetics
The genotypes of the S. pombe strains used in this study are listed in Table 1. The complete yeast growth medium (YES), selective medium (MM) supplemented with the appropriate requirements and sporulation medium (MEA) have been described elsewhere (Moreno et al., 1991). Caspofungin (Csp; Deresinski and Stevens, 2003) was stored at –20°C in a stock solution (2.5 µg/ml) in H₂O and was added to the media at the corresponding final concentration after autoclaving. Crosses were performed by mixing appropriate strains directly on MEA plates. Recombinant strains were obtained by tetrad analysis. For overexpression experiments using the nmt1 promoter, cells were grown in EMM containing 15 µM thiamine up to the logarithmic phase. Then, the cells were harvested, washed three times with water, and inoculated in fresh medium (without thiamine) at an OD₆₀₀ = 0.01.

Plasmid and DNA Manipulations
pYS8, containing the rgf1 ORF, was obtained by inserting a 7-kb EcoRI fragment from cosmid C645 into pAU-KS (Tajadura et al., 2004). An XhoI-NotI fragment from pYS8 (containing the 7-kb EcoRI fragment) was subcloned into the XhoI-NotI sites of pAL-KS, thus affording pAL-rgf1⁺. To tag Rgf1p at the C-terminus with enhanced green fluorescent protein (EGFP) and with the triple repeat of the influenza virus hemagglutinin (HA) epitope (Craven et al., 1998), pAL-rgf1⁺ was modified by site-directed mutagenesis. We destroyed the NotI site at the multiple cloning site and created a NotI site by site-directed mutagenesis just before the TAA stop codon of rgf1⁺ (pGR41). The GFP and HA epitopes were inserted in-frame at the NotI site of pGR41. pGR45 (pAL-rgf1⁺-GFP) and pGR46 (pAL-rgf1⁺-HA) fully complemented the rgf1 phenotypes. Strain PG40, with the rgf1⁺-GFP integrated under its own promoter, was constructed by subcloning the rgf1⁺ tagged with GFP (from plasmid pAL-rgf1-GFP) into the integrative vector pIJ148, resulting in pIJ148-rgf1⁺-GFP (pGR49). This plasmid was cut with Eco47III and integrated into the leu1 locus of strain VT14. The nmt1 promoter-containing vectors pREP3X and pREP41X (Forsburg and Sherman, 1997) were used to overexpress rho1⁺ to rho5⁺, cdc42⁺, and rgf3⁺. All GTPases of the Rho family were tagged with two HA epitopes at the 5' end (Calonge et al., 2003). The rho overexpression plasmids were kindly provided by P. Perez and P. M. Coll (Instituto de Microbiologia Bioquimica, Salamanca, Spain). pAL-bgs1⁺, pAL-bgs2⁺, pAL-bgs3⁺, and pAL-bgs4⁺ multicopy plasmids were used to overexpress the -GS catalytic subunits, each with their own promoter. pAL-bgs1⁺ was kindly provided by J. C. Cortes and J. C. Ribas (Cortes et al., 2002). pAL-bgs2⁺, pAL-bgs3⁺ and pAL-bgs4⁺ have been described previously (Martin et al., 2000, 2003; Cortes et al., 2005). To overexpress rgf1⁺, an XhoI-SmaI fragment containing the rgf1⁺ gene tagged with the HA epitope from plasmid pGR46 was ligated into the XhoI-SmaI sites of plasmid pREP41X (pGR57) and pREP3X (pGR58). pGR33 is pREP3X with an XhoI-SmaI fragment containing the rgf1⁺ ORF without the HA tag.

Rgf1 Deletion
The rgf1::his3 disruption construct was obtained in a two-step process. The 3'-flanking region (nt 4004–5350) was obtained by PCR, inserting SalI and ApaI sites into the same sites of the SK-his3⁺ vector to yield pVT16. Then, a PCR fragment of the 5' end of rgf1⁺ (nt–1490 to –62) carrying BamHI and NotI sites at the ends was digested with BamHI, treated with Klenow, and then digested with NotI and ligated into the SmaI and NotI sites of pVT16 to yield pVT2. Plasmid pVT2 was digested with ApaI and NotI, and the linear DNA inserted was used to transform a diploid and a haploid strain (YS165 and YS64, respectively). Correct deletion of the rgf1⁺ ORF was confirmed by PCR analysis using the following oligonucleotides: M22 (5'-GTGTTCGCTAATTGCGC-3') into the his3⁺ gene and R23-e (5'-CAAGGGTATGTGGTCTGG-3') downstream from the nt 5350 and therefore external to the deletion cassette. A diploid strain heterozygous for the rgf1::his3⁺ allele was subjected to tetrad analysis. his⁺/his^– segregation in tetrads was regular, indicating that rgf1⁺ is not essential for vegetative growth. Gene replacement was also confirmed by genomic Southern blotting of a tetrad (unpublished data). To make the rgf1::kan disruption construct (pGR59), pVT2 was cut with SalI and SpeI (to eliminate the his3 marker) and replaced it by the kanMX6 gene from pFA6a-kanMX6, (Bähler et al., 1998). Plasmid pGR59 was digested with ApaI and NotI and the linear DNA containing the cassette was used transform a haploid strain (YS64). rgf1::kan^R disruptants were selected as Kanamicin-resistant and Caspofungin-hypersensitive. cdc25-2 rgf1 and cdc10-129 rgf1 mutants were obtained by genetic crosses; the offspring were analyzed for cdc (ts phenotype) and for rgf1 kanamicin resistance and Caspofungin hypersensitivity. cwg1-1 rgf1 mutants were obtained by genetic cross of cwg1-1 (JCR132) and rgf1 (PG76) strains and selected from tetrads where NPD (nonparental ditypes) were produced.

View larger version (14K): [in this window] [in a new window]   Figure 1. Comparison of structural features of Rgf1, Rgf2, and Rgf3 analyzed by the SMART program (Letunic et al., 2002; http://smart.embl-heidelberg.de/). Domains are indicated: CNH, citron homology domain; DH, Dbl homology domain conserved among GEFs for Rho/Rac/Cdc42-like GTPases; PH, pleckstrin homology domain; DEP, domain of unknown function present in signaling proteins that contain PH, RasGEF, RhoGEF, RhoGAP, RGS, and PDZ domains.

Pulldown Assay for GTP-bound Rho Proteins
The expression vector pGEX-C21RBD (rhotekin-binding domain; Reid et al., 1996) was used to transform Escherichia coli cells. The fusion protein was produced according to the manufacturer's instructions and immobilized on glutathione-Sepharose 4B beads (Amersham). After incubation, the beads were washed several times, and the bound proteins were analyzed by SDS-PAGE and stained with Coomassie. The amount of GTP-bound Rho proteins was analyzed using the Rho-GTP pulldown assay modified from Ren et al. (1999). Briefly, wild-type, rgf1⁺-overexpressing, and rgf1 mutant cells were transformed with either pREP3X-HArho1⁺ or pREP3X-HArho4⁺ and grown for 18 h in minimal medium without thiamine. Extracts from 10⁸ cells were obtained as described previously (Arellano et al., 1997), using 500 µl of lysis buffer (50 mM Tris, pH 7.5, 20 mM NaCl, 0.5% NP-40, 10% glycerol, 0.1 µM dithiothreitol, 1 mM NaF, 2 mM MgCl₂, containing 100 µM p-aminophenyl methanesulfonyl fluoride, leupeptin, and aprotinin). GST-RBD fusion protein, 100 µg, coupled to glutathione-agarose beads was used to immunoprecipitate 1.5 mg of the cell lysates. The extracts were incubated with GST-RBD beads for 2 h. The beads were washed with lysis buffer four times, and bound proteins were blotted against 1:2000-diluted 12CA5 mAb as primary antibody to detect HA-Rho1p or HA-Rho4p. The total amounts of HA-Rho1p or HA-Rho4p levels were monitored in whole-cell extracts (10 µg of total protein), which were used directly for Western blot and were developed with 12CA5 mAb. Immunodetection was accomplished using the ECL detection kit (Amersham Biosciences).

View larger version (69K): [in this window] [in a new window]   Figure 2. Growth phenotypes of rgf1 null cells. (A) Left, growth curves of rgf1 cells (VT14) and the corresponding isogenic wild-type cells (HVP54). Log-phase cells grown at 28°C were diluted to the same optical density and further grown in YES medium. Right, percentage of colony forming units (cfu) of the mutant rgf1 compared with that of the wild-type isogenic strain. Cells prepared as above were diluted and counted, and the same number of cells were plated on YES medium and incubated for 3 d at 28°C. (B) Morphology of rgf1 mutant. Differential interference contrast (DIC) micrographs of S. pombe wild type (HVP54) and rgf1 (VT14) grown in YES liquid medium at 28°C; in right panel, rgf1 cells were grown in the presence of 1.2 M sorbitol for 6 h. (C) rgf1 mutant cells are hypersensitive to Caspofungin (Csp). Equal number of wild-type and rgf1 cells were diluted and (4 x 104, 2 x 104, 2 x 102, and 2 x 101 cells, respectively) were spotted onto YES plates with or without 0.1 µg/ml Csp (CANCIDAS). Colony formation was analyzed following 2–3 d of incubation at 28°C.

Microscopy Techniques
The localization of Rgf1p-GFP, Crn1p-GFP, and Atb2p-GFP was visualized in living cells. For Calcofluor staining, exponentially growing S. pombe cells were harvested, washed once, and resuspended in water with Calcofluor (Cfw) at a final concentration 20 µg/ml for 5 min at room temperature. After washing with water, cells were observed under a DMRXA microscope (Leica, Wetzlar, Germany). Actin staining was performed using AlexaFluor 488-phalloidin (Molecular Probes, Eugene, OR).

View larger version (61K): [in this window] [in a new window]   Figure 3. Rgf1p is required for bipolar growth. (A) Actin organization in rgf1 cells. Left, early-log phase cells rgf1+crn1+-GFP (JCR962) and rgf1 crn1+-GFP (PG92) grown in YES liquid medium at 28°C were collected and visualized for Calcofluor (Cfw) staining and GFP (actin-associated Crn1p) fluorescence. Cfw was added at 20 µg/ml, followed by immediate examination of the cells. Note that compared with the wild type, in the rgf1 mutant the actin patches often have a more monopolar distribution. The graphic represents the percentages of monopolar and bipolar actin growth patterns in rgf1+ (JCR962, n = 203) and rgf1 cells (PG92, n = 226) as examined by actin staining. Bottom right, rgf1 cells were fixed and stained with AlexaFluor-conjugated phalloidin to stain F-actin structures. Actin cables and patches are indicated. (B) cdc10-129 (MS168) and cdc10-129 rgf1 (PG88) grown at 25°C to OD600 0.15, shifted to 37°C for 4 h, and then grown at 25°C for 330 min. Aliquots of cells were harvested before and every 30 min after the shift to 25°C. The graphics (on the left) represents the percentage of bipolar cells () and septa () at each time point. On the right the percentage of lysed cells at each time point is represented. Micrographs show Cfw stained cdc10-129 and cdc10-129 rgf1 cells 60 min after the shift to 25°C. (C) cdc25-22 (NG669) and cdc25-22 rgf1 (PG43) cells grown at 25°C to an OD600 0.15, shifted to 37°C for 4 h, and then grown at 25°C for 300 min. Aliquots of cells were harvested before and every 20 min after the shift to 25°C. On the left, micrographs showing cdc25-22 and cdc25-22 rgf1 cells 100 min (top panels) and 200 min (bottom panels) after the shift to 25°C. The graphic on the right represents the percentage of lysis of cdc25-22-synchronized cells () or cdc25-22 rgf1-synchronized cells () at each time point. The septation index of cdc25-22 rgf1 cells is shown by the dashed line (–––).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

The rgf1⁺ gene encodes a protein of 1334 amino acids, with a predicted molecular size of 150.1 kDa. Structural analysis of Rgf1p revealed that it contains the putative Dbl homology domain (DH; amino acid residues: aa 625-807) and a pleckstrin homology domain (PH; aa 844-973) adjacent to the DH domain characteristic of most Rho-GEFs. The DH-PH tandem is responsible for the activation of Rho-family GTPases in response to diverse extracellular stimuli (reviewed in Gulli and Peter, 2001; Rossman et al., 2005). A DEP (Dishevelled, Egl-10, and Pleckstrin) domain (aa 424-497) and a CNH (Citron and NIK1-like kinase homology domain; aa 997-1293) were also found. Their function is not clear, but in most cases they act as regulatory domains involved in macromolecular interactions (http://www.genedb.org/genedb/pombe/index.jsp; Figure 1). There are seven genes that encode a putative Rho GEF domain in S. pombe (Coll et al., 2003; Hirota et al., 2003; Iwaki et al., 2003; Tajadura et al., 2004); among them, the closest homolog to rgf1⁺ is rgf2⁺. A computer search of the deduced amino acid sequence showed that the identity percentage observed between Rgf1p and Rgf2p was 39.4% and this rose to 63.4% upon comparing the GEF domains, whereas the identity between Rgf1p and Rgf3p (whole proteins) was 16 and 22.6% in the GEF domain.

rgf1 Null Cells Show Defects in Cell Integrity Similar to the Depletion of Rho1p
To characterize the relationship between Rgf1 and Rho proteins, we carried out a series of experiments to determine whether Rgf1p was acting upstream from any of the Rho proteins. First, we created a strain defective in rgf1 by replacing the rgf1 ORF with the his3⁺ marker, as detailed in Materials and Methods. The resulting strain, rgf1, showed a slow growth pattern at 28°C (Figure 2A), and the viability of the rgf1 cells was 55% compared with the wild-type isogenic strain. Curiously, the growth defect was less severe when rgf1 cells were grown at 37°C (unpublished data). We observed the morphology in the 25–28 to 32°C temperature range and found that regardless of the growth temperature 30–35% of the cells were lysed, whereas the rest of the cells exhibited the wild-type morphology (Figure 2B). The lysed cell phenotype of the rgf1 cells was similar to that observed in the ehs2-1 mutant (affected in the rgf3⁺ gene; Tajadura et al., 2004) and in cells depleted for Rho1p (Arellano et al., 1997). The same phenotype has also been reported in cells expressing the Rho1T20N dominant-negative mutant (Nakano et al., 1997). Lysis of the rgf1 mutants cells was suppressed by the addition of 1.2 M sorbitol (Figure 2B). These phenotypes prompted us to investigate whether the mutants had a defect in cell wall architecture. We examined the sensitivity of rgf1 null mutants to different concentrations of Csp (CANCIDAS, Merck), a lipopeptide antibiotic that inhibits (1,3)-glucan biosynthesis (Deresinski and Stevens, 2003). As shown in Figure 2C, rgf1 cells were unable to grow on plates supplemented with Csp (0.1 µg/ml), whereas the wild-type cells were able to withstand concentrations of up to 5 µg/ml (unpublished data). These results suggest that the rgf1 null mutant cells lose their integrity, probably because of defects in cell wall biosynthesis.

The rgf1 Mutation Causes Defect in Bipolar Growth
Activation of Rho-family GTPases leads to the assembly of contractile actin-myosin filaments (Jaffe and Hall, 2005). In fission yeast, actin is organized as longitudinal F-actin cables and cortical F-actin patches at the growing ends of interphase cells, where the cell wall is newly synthesized (Marks and Hyams, 1985). To determine whether Rgf1p plays a role in any of these events, we used Crn1p-GFP (coronin), a marker for actin patches (Pelham and Chang, 2001), and Atb2p-GFP (alpha-tubulin 2) for microtubule observation (Ding et al., 1998). As shown in Figure 3A, the rgf1 mutants showed a defect in actin organization in that they organized actin patches mostly at one end of the cell only (Figure 3A, photos and graphic). This cell end corresponded to the growing end in these monopolar cells as also assessed by Calcofluor staining. Actin organization at the cell division site and F-actin cable formation was not affected in rgf1 cells (Figure 3A). We also failed to detect any significant interphase MTs defects (unpublished data).

We next investigated the behavior of rgf1 cells in the G2 phase of growth and wondered whether the lysed cell phenotype of rgf1 null mutants was due to a defect in tip elongation. To this end, we constructed the double mutant cdc10-129 rgf1 (see Materials and Methods) and synchronized cells in G1 by arrest at 37°C. The areas in which new cell wall deposition, and hence growth, was occurring were visualized using the fluorescent dye Calcofluor white (Cfw). Eighty minutes after being shifted to the permissive temperature 55% of cdc10-129 cells displayed bipolar growth, whereas only 4% of cdc10-129 rgf1 cells were bipolar (Figure 3B). At 150 min after shifting, only 14% cdc10-129 rgf1 cells were bipolar. However in both strains, septation started and proceeded almost at the same time (Figure 3B, bipolar and septa plots). In this situation, the percentage of lysis in the cdc10-129 rgf1 mutant remained high (45–30%) during the first part of the time course, when bipolar growth takes place, and declined slightly in septating cells (25–15%; Figure 3B). In sum, rgf1 cells showed a defect in the activation of bipolar growth that coincided with the highest percentage of lysis.

To examine septation in rgf1 cells, we constructed the double mutant cdc25-22 rgf1 and synchronized cells in G2 by cdc25-22 arrest at 37°C. Cells were grown at 25°C to log phase, changed to 37°C for 4 h, and then returned to 25°C. Aliquots were taken at different times to count cells with septa and lysed cells. On shifting the cells to the permissive temperature, septation was initiated at the same time in cdc25-22 and cdc25-22 rgf1 cells. However, the second round of septation was slightly ahead in the rgf1 strain (unpublished data). The appearance and number of septa were similar in both strains (Figure 3C, top panels). This result suggested that septum formation and cell separation proceeded normally in the absence of Rgf1p. However, regarding cell lysis we found a peak of broken cells in the cdc25-22 rgf1 strain just before the second round of septation, corresponding to cells in the G2 phase (Figure 3C, lower panels and graph). Thus, rgf1 cells display several phenotypes that are consistent with a role of Rgf1p in actin reorganization during activation of bipolar growth, one of the major changes in polarized growth during the S. pombe morphogenetic cycle.

View larger version (62K): [in this window] [in a new window]   Figure 4. Rgf1p acts as a positive regulator of Rho1p. (A and B) The Caspofungin-hypersensitive growth phenotype of rgf1 mutants is suppressed by overexpression of rho1+. HVP54 (rgf1+) was transformed with pREP3X (empty vector), and VT14 (rgf1) was transformed with pREP3X-rho1 (rho1+), pREP3X-rho2 (rho2+), pREP3X-rho3 (rho3+), pREP3X-rho4 (rho4+), pREP3X-rho5 (rho5+), pREP3X-cdc42 (cdc42+), and pREP3X (empty vector). Transformants were spotted onto MM, MM plus thiamine, MM plus 2 µg/ml Caspofungin (Csp), and MM plus thiamine and 2 µg/ml Csp plates as serial dilutions (4 x 104 cells in the left row and then 2 x 104 cells and 2 x 102 in each subsequent spot) and incubated at 28°C for 3 d. (B) HVP54 (rgf1+) was transformed with pAL (empty vector) and VT14 (rgf1) was transformed with pAL-rho1 (rho1+) and pAL (empty vector). Serial dilutions of the transformant cultures were spotted onto MM and MM plus 2 µg/ml Csp (from left to right 4 x 104, 2 x 104, 2 x 103, 2 x 102, 2 x 10) and incubated as above. (C) Deletion of rgf1+ suppresses the growth defect in rga1 cells. Wild-type (wt) (PG75-4c), rgf1 (PG76-5a), rga1 (PG74-2b), and rgf1rga1 (PG73-1c) segregants were streaked onto YES medium and incubated at 28°C for 3 d. rgf1rga1, but not rga1, cells grew as wild-type cells. DIC (differential interference contrast) images of (wt) (PG75-4c), rga1 (PG74-2b), rgf1rga1 (PG73-1c)—from micromanipulation plates— grown in YES liquid medium at 28°C for 8 h are shown in the bottom panel.

To gain further evidence that Rgf1p was a GEF for Rho1p, we tested whether an rgf1 null mutation was able to counteract a mutation in Rga1p, a protein with GAP activity toward Rho1p and hence a negative regulator (Nakano et al., 2001). Lack of Rga1p produces small colonies and the cells show a swollen, multiseptated or branched shape; a phenotype similar to that seen in cells in which Rho1p is excessively activated (Figure 4C; Nakano et al., 2001). We replaced the rgf1⁺ gene with the his3⁺ marker in a diploid strain, rga1/rga1⁺, and the rgf1, rga1, rgf1rga1 segregants from 16 tetrads were analyzed. We found that the rgf1rga1 cells formed regularly sized colonies, like rgf1 cells (unpublished data). When rgf1, rga1, rgf1rga1 strains (respectively) were streaked out on rich medium (YES) at 28°C, the rga1 cells were severely impaired for growth, whereas rgf1rga1 exhibited a better growth pattern and resembled rgf1 cells. The rounded and branched shape seen in the rga1 mutant cells returned to the wild-type morphology in the double mutant rgf1rga1 cells (Figure 4C). Thus, rgf1⁺ and rga1⁺ indeed appear to antagonize each other, presumably acting on the same Rho-like GTPase.

Rgf1 Associates with Rho1p in S. pombe Cells and Promotes the GDP-GTP Exchange
To examine whether there was a direct interaction between Rgf1p and Rho1p in S. pombe cells, we performed coprecipitation experiments. We coexpressed HA-epitope–tagged Rgf1 protein (Rgf1-HA) together with either GST-Rho1 or GST in S. pombe cells. Cells were lysed, and the supernatant fractions of the lysates were incubated with glutathione-Sepharose beads to isolate GST complexes, which were analyzed by immunoblotting. As shown in Figure 5A, Rgf1-HA was found to be associated with GST-Rho1, but not with GST.

View larger version (33K): [in this window] [in a new window]   Figure 5. Rgf1p is a specific Rho1-GEF. (A) Coprecipitation of Rgf1p and Rho1p. Cell extracts from cells expressing, GST and Rgf1p-HA; GST-Rho1p and Rgf1p, and GST-Rho1p and Rgf1p-HA were precipitated with glutathione beads and blotted against 12CA5 monoclonal anti-HA antibody (top). Western blot with anti-HA antibody was performed on total extracts to visualize total Rgf1p-HA levels (bottom). (B) In vivo, the Rgf1p level modulates the amount of GTP-bound Rho1p. Wild-type (YS64) cells expressing pREP4X or pREP4X-rgf1, and rgf1 (VT14) mutant cells were transformed with either pREP3X-HA-rho1 or pREP3X-HA-rho4. GTP-Rho1p or GTP-Rho2p were pulled down from the cell extracts with GST-C21RBD and blotted against 12CA5, anti-HA mAb. Total HA-Rho1p and HA-Rho4p was visualized by Western blot. (C) Alignment of predicted amino acid sequence at the CR3 region of Rgf1p with the corresponding region of proteins with a Rho-GEF domain found in S. pombe. Multiple sequence alignments were performed using the ClustalW program. The amino acids deleted in the rgf1-PTTR mutant are marked with "black caps" over the predicted amino-acid sequence of Rgf1p. The proline (P) mutated in the rgf3 mutant (ehs2-1) is also highlighted and signaled by an arrow. Asterisks indicate identical amino acids among all identified gene products. (.) and (:) indicate well- and highly-conserved amino acids, respectively.

To further investigate the possible role of Rgf1p as a Rho1p activator, we analyzed the in vivo amount of GTP-bound Rho1p in cells with different amounts of Rgf1p. rgf1 mutant cells carrying the control plasmid pREP4X and wild-type cells carrying either pREP4X or pREP4X-rgf1⁺ were transformed with pREP3X-HA-rho1⁺. After induction of the nmt1 promoter for 18 h, the amount of Rho1p bound to GTP was analyzed by precipitation with GST-C21RBD, the rhotekin-binding domain (which had previously been obtained and purified from bacteria) and blotting with anti-HA antibody (Figure 5B). Western blots of whole extracts (25 µg protein) showed that the total amount of Rho1p was similar in all three strains with different amounts of Rgf1p (Figure 5B). The amount of active Rho1p increased considerably in the strain overexpressing Rgf1p compared with the wild-type strain. Moreover, only a minor amount of GTP-Rho1p was detected in the strain lacking Rgf1p. As a control, we also analyzed the amount of GTP-bound Rho4p in rgf1, wild-type, and cells overexpressing rgf1⁺ (Figure 5B, bottom panel). These cells were transformed with the plasmid pREP3X-HA-rho4⁺ and GTP-bound Rho4p was pulled down from the extract by binding to GST-C12RBD. No changes in the level of Rho4p bound to GTP were observed among the three strains (Figure 5B). These results provide evidence that Rgf1p interacts with Rho1p and acts as a specific Rho1p activator in S. pombe. To examine whether the GEF domain was essential for Rgf1p function, we created a deletion mutant in the RhoGEF domain of Rgf1p (rgf1-PTTR; Figure 5C). The DH domain contains three conserved blocks of sequences that have previously been referred to as conserved regions 1–3, or CR1–3. These three conserved regions form three long helices, H1a, H2b, and H8, which pack together to form the core of the DH domain. The four amino acids that were deleted in the rgf1-PTTR mutant (proline-threonine-threonine-arginine, PTTR) have been predicted to be located on helix H8 (CR3), which is the most highly conserved region of the DH domain and where many mutations that decrease nucleotide exchange activity map (Liu et al., 1998; Soisson et al., 1998). Moreover, a single change from a proline to a serine in that conserved region of Rgf3p is responsible for the thermosensitive lytic phenotype in the ehs2-1 mutant (Tajadura et al., 2004). We found that the rgf1-PTTR mutant integrated in a single copy in rgf1 strain maintained the lytic and the Csp-hypersensitive phenotype of the rgf1 null mutants, thus supporting the hypothesis that Rgf1p acts as a GEF.

rgf1⁺ Overexpression Causes Aberrant Morphology and Increases (1,3)-glucan Synthase Activity
It has previously been reported that overexpression of rho1⁺ or constitutively active rho1 mutants from the strong nmt1 promoter causes an aberrant morphology in S. pombe cells (Arellano et al., 1996; Nakano et al., 1997). If rgf1⁺ functions as a positive regulator of Rho1p, overexpression of rgf1⁺ would be expected to produce phenotypes similar to that of Rho1-overexpressing cells. The rgf1⁺ gene was cloned under the thiamine-repressible nmt1 promoter in the pREP3X vector. When thiamine was eliminated to enhance rgf1⁺ expression, the cells were unable to grow on plates (unpublished data). After 18 h of induction, cells were larger than the wild-type, round, or misshapen, with abnormal septa. These cells also showed a general increase in Cfw fluorescence, some containing aberrant depositions of Cfw-stainable material (see cells marked with an arrow and enlarged cells in Figure 6A).

View larger version (47K): [in this window] [in a new window]   Figure 6. Phenotypes of Rgf1p overexpression (OP). (A) Overexpression of rgf1+ causes cell growth arrest and an abnormal accumulation of cell wall material. DIC and Calcofluor-stained UV micrographs of wild-type cells transformed with pREP3X (empty plasmid) or pREP3X-rgf1 (Rgf1-OP) grown without thiamine for 18 h. (B) In vitro glucan synthase (GS) activity assayed with the membrane fraction of wild-type cells (HVP54) transformed with pREP3X (empty plasmid), pREP3X-rgf1 (Rgf1-OP), pREP3X-rho1 (Rho1-OP), or both pREP4X-rgf1 and pREP3X-rho1 (Rgf1-OP and Rho1-OP). Extracts were prepared from cells grown in MM without thiamine at 32°C for 18 h. Specific activity is expressed as milliunits mg–1 protein. Values are means of at least three independent experiments with duplicated samples, and error bars represent SDs.

As expected, GS activity increased during rgf1⁺ overexpression. This activity was fourfold higher than that observed in the wild-type strain (Figure 6B). To corroborate these results, we also studied the activity in cells that overexpressed rho1⁺ and rgf1⁺ at the same time (transformed with pREP3X-rho1 and pREP4X-rgf1 plasmid). As described previously, cells overexpressing rho1⁺ showed an increase in GS activity (Figure 6B; Arellano et al., 1996). This increase was considerably (10-fold) higher in cells that overexpressed rgf1⁺ at the same time (Figure 6B). These results clearly indicate that Rgf1p is involved in the regulation of (1,3)-glucan biosynthesis.

Genetic Evidence that Rgf1p Interacts Functionally with Bgs4p and Pck1p
It is known that Rho1p functions by activating -glucan biosynthesis, but the issues of which of the GS catalytic subunits it activates, remain unclear. In a previous work, we reported that a mutation in rgf3⁺ (ehs2-1 mutation) was suppressed by bgs3⁺, one of the putative (1,3)-GS subunits. Multiple copies of bgs3⁺ complemented the hypersensitivity to Ech and Cfw but not the temperature-sensitive phenotype (Martin et al., 2003). To define a possible relationship between Rgf1p and known Rho1p effectors, we first tested whether overexpression of any of the -GS subunits could suppress the hypersensitive phenotype of the rgf1 mutants. The rgf1 strain VT14 was transformed with the high-copy number plasmids pAL-bgs1⁺, pAL-bgs2⁺, pAL-bgs3⁺, and pAL-bgs4⁺, and transformants were monitored for growth in Csp. As shown in Figure 7A, only a moderate expression of bgs4⁺ restored growth of an rgf1 mutant in the presence of the antifungal agent. We also examined the consequences of overexpressing rgf1⁺ in cwg1-1 cells, which hold a nonlethal thermosensitive mutation in the essential bgs4⁺ gene (Cortes et al., 2005). When the cwg1-1 strain was transformed with the rgf1⁺ gene driven by the nmt1 promoter, neither the pREP3X-rgf1 with thiamine (promoter off) nor the same without thiamine (promoter on) suppressed lysis at 37°C of cwg1-1 cells (unpublished data). In addition, rgf1 cwg1-1 cells were phenotypically similar to cwg1-1 (bgs4) cells at 37°C (Figure 7B). This finding, combined with the above observations, suggests that Rgf1p specifically activates the Rho1p-Bgs4p GS complex.

View larger version (79K): [in this window] [in a new window]   Figure 7. The Caspofungin-hypersensitive growth phenotype of rgf1 mutants is suppressed by overexpression of bgs4+ and pck1+. (A) HVP54 (rgf1+) was transformed with pAL (empty vector) and VT14 (rgf1) was transformed with pAL (empty vector), pAL-bgs4 (bgs4+), pAL-bgs3 (bgs3+), pAL-bgs2 (bgs2+), and pAL-bgs1 (bgs1+). Transformants were spotted onto MM and MM plus 2 µg/ml Csp plates as serial dilutions (8 x 104 cells in the left row and then 4 x 104, 2 x 104, 2 x 103, 2 x 102, and 2 x 101 in each subsequent spot) and incubated at 28°C for 3 d. (B) PG76 (rgf1), JCR132 (cwg1-1), and PG82 (rgf1 cwg1-1) cells derived from a cross between strains PG76 (rgf1) and JCR132 (cwg1-1) were grown to log phase in rich medium at 28°C, shifted for 4 h to 37°C, and then visualized under differential interference contrast (DIC) microscopy. The relevant genotype of each strain is indicated at top of each panel. (C) HVP54 (rgf1+) was transformed with pDB248 (empty vector) and VT14 (rgf1) was transformed with pDB248 (empty vector), pDB248-pck1 (pck1+), and pDB248-pck2 (pck2+). Transformants were spotted onto MM and MM plus 2 µg/ml Casp plates and processed as in A.

Rgf1p Localizes to One or Both Poles during Cell Growth and to the Contractile Ring and Septum during Cytokinesis
To determine the subcellular localization of Rgf1p, the coding sequence of the green fluorescence protein (GFP) was fused in-frame before the rgf1⁺ stop codon. The GFP-rgf1⁺ rgf1 strain (GFP at amino acid 1334, integrated in single copy, with its own promoter and in the absence of original rgf1⁺ gene) completely restored the wild-type phenotype to rgf1 mutant cells. The cells were visualized using GFP fluorescence in order to detect Rgf1p and by Cfw staining. Rgf1p was found to localize to the growing ends and septum along the mitotic cycle, overlapping with Cfw staining (Figure 8A). When cell growth began, Rgf1p accumulated at the old growing end. During bipolar growth, Rgf1p also localized to the opposite pole. Finally, the GFP disappeared from both poles and localized only to the middle of the cell, concentrating as two faint dots on either side of the emerging septum. The GFP then moved to the inner border of the growing septum, forming a ring that moved centripetally with the edge of the growing septum (Figure 8B). After the septum wall had been completed, the GFP appeared in two separate bands (unpublished data). During cell division, Rgf1p remained on both sides of the septum until the two daughter cells were ready to separate or had already done so. To confirm these observations, confocal microscopy was used. The results of the 3-D reconstruction of the green fluorescence indicated that during septum formation Rgf1p-GFP was localized to a platelike structure; fluorescence was ring-shaped in the first stages, and as the ring was closing the fluorescence remained behind the edge and ended up distributed as a division plate (Figure 8C).

View larger version (117K): [in this window] [in a new window]   Figure 8. Rgf1p localizes to the growing regions: one or both poles, the medial ring, and the septum. (A) Rgf1p-GFP localization at different stages of the cell cycle. Early log phase cells containing the rgf1+-GFP fusion allele were visualized for GFP fluorescence and Cfw staining. Cfw was added at 20 µg/ml, followed by immediate examination of the cells (bottom panel). (B) Magnification of Rgf1p-GFP localization to the medial ring and along the plasma membrane during septum formation. The cells shown were chosen from among a population of >50 dividing cells. (C) Three-dimensional reconstruction of Rgf1p localization. Cells containing the rgf1+-GFP fusion allele observed under a confocal microscope and z-sections of 0.3 mm were taken. The image was reconstructed using LMS510 software.

Rgf1p localized to growing areas (septum and poles) and played an important function during bipolar growth; Rgf3p localizes and functions specifically during cytokinesis (Tajadura et al., 2004). We therefore investigated whether the role of Rgf1p in the regulation of Rho1p was overlapping that of Rgf3p. Previous work had shown that moderate expression of rgf1⁺ did not suppress lysis at 37°C of the rgf3 mutant (ehs2-1; Tajadura et al., 2004). We wondered if the opposite was also true; whether overexpression of rgf3⁺ was able to suppress the hypersensitive phenotype or the lysis of the rgf1 strain. rgf3⁺, driven either by its own promoter or by the nmt1 promoter, was not able to suppress the hypersensitivity in the presence of Csp nor the lysis of rgf1 cells. Moreover, disruption of rgf1⁺ in an rgf3 mutant (ehs2-1) produced viable cells at 28°C but not at 37°C, the temperature at which both mutants were able to grow on plates (unpublished data). These result support the hypothesis that Rgf1p and Rgf3p are not functionally interchangeable. Previous studies (Iwaki et al., 2003; Tajadura et al., 2004) and our own results suggest that both Rgf3p and Rgf1p are GEFs of Rho1p. The experiments reported here indicate that Rgf1p activates a Rho1p pathway (Rho1p-Bgs4p) other than that activated by Rgf3p.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Guanine nucleotide exchange factors (GEFs) are directly responsible for the activation of Rho-family GTPases in response to diverse stimuli and ultimately regulate many cellular responses such as proliferation, differentiation, and movement (Rossman et al., 2005). Seven Rho-GEFs belonging to the Dbl family of proteins have been identified in S. pombe (Iwaki et al., 2003). Here we studied the Rho-GEF, designated Rgf1, which like other Rho-GEFs contains the DH-PH tandem motifs required to activate Rho proteins (Schmidt and Hall, 2002).

Here we have shown that Rgf1p is likely to be a GEF for Rho1p. rgf1 cells are defective in cell integrity and lyse with a phenotype similar to cells devoid of Rho1 or Pck1/2 activity. Moreover, mutants lacking rgf1 display a defect in actin organization and in -glucan biosynthesis. The fact that both processes are controlled by Rho1p suggests that the main function of Rgf1p would be to regulate this GTPase. Consistent with this idea, the hypersensitivity to Csp and the lytic phenotype were suppressed by overexpression of rho1⁺ but not other rho genes. Additionally, we provide genetic and biochemical evidence to support the view that Rgf1p interacts functionally with and acts as a positive regulator of Rho1p: 1) Deletion of rgf1⁺ suppresses the slow growth defect of a null mutant in the rga1⁺ gene, encoding a GTPase-activating protein for Rho1 (Nakano et al., 2001). This finding suggests that Rgf1 may play a role antagonistic to that of Rga1p GAP. 2) Rgf1p specifically coprecipitated with Rho1p, and the level of Rgf1p modulated the level of GTP-Rho1p in vivo. 3) Overexpression of rgf1⁺ was lethal and caused a phenotype similar to that of the constitutively active allele Rho1G15V in wild-type S. pombe cells, whereas it was not deleterious when overexpressed in a GTPase-deficient Rho1p strain (Rho1F85I; S. Rincón and P. Pérez, unpublished data). Furthermore, we found that the GEF domain of Rgf1p was essential for its function; a deletion mutation in a highly conserved region of the Rgf1p-DH-domain produced a lack of function phenotype. We also found that a functional GEF domain was not necessary for its localization, because the mutated protein tagged with GFP localized correctly (unpublished data).

In S. pombe, Rho1p signaling is required to maintain cell integrity, regulating the biosynthesis of (1,3)-glucan and the cell wall in general, and it is also required for actin polymerization. The experiments reported in this study indicate that rgf1⁺ is involved in the regulation of cell wall biosynthesis. rgf1 mutant cells were unable to grow at 50-fold lower concentrations of the antifungal drug Csp than wild-type cells and showed a lytic phenotype that could be rescued by the presence of 1.2 M sorbitol. Cells that overexpressed rgf1⁺ showed aberrant depositions of Cfw-stainable material, accompanied by a GS activity that was 5-fold that of wild-type cells. Furthermore, cells overexpressing rgf1⁺ together with rho1⁺ showed a huge increase in GS activity (approximately 7- to 10-fold) compared with the wild-type level. Even without GTP added to the reaction, the GS activity of cells that overproduced rgf1⁺ was 20 times higher than in the wild type, indicating that an excess of Rgf1p had raised the intracellular pool of GTP-bound Rho1p (already activated). Furthermore, our results suggest that Rgf1p would activate the -GS complex containing the catalytic subunit Bgs4p. Rgf1p, Rho1p, and Bgs4p localize to growth areas, the septum and the poles (Arellano et al., 1997; Cortes et al., 2005). Individual mutants (in rgf1⁺ and bgs4⁺) showed similar cell wall–related phenotypes (lysis and hypersensitivity to antifungal drugs), and the double mutant rgf1 cwg1-1 was very similar to bgs4 (cwg1-1) single mutant. Moreover, overexpression of bgs4⁺ suppressed the rgf1 hypersensitive phenotype.

The interaction observed between Rgf1p and Pck1p is more intriguing because the role of Pck1p in cell wall integrity remains to be established. The patterns of cell wall regulation by Pck1p and Pck2p seem to be different. pck1, but not pck2, cells are hypersensitive to Ech and additional copies of Pck2p cannot suppress this phenotype (Arellano et al., 1999b), suggesting that in the absence of Pck1p the genes specifically involved in protection against antifungal drugs cannot be turned on. The fact that multiple copies of pck1⁺ (with its own promoter) are able to suppress Csp hypersensitivity in rgf1 mutants is in agreement with the notion of Pck1p kinase being an effector of Rho1p and suggests that Pck1p would be necessary for the activation of genes (probably bgs4⁺ or other GS) in response to signaling after cell wall damage. In fact, mild overexpression of either bgs4⁺ (our unpublished observations), bgs1⁺ or bgs2⁺ (Arellano et al., 1999b) was able to suppress the hypersensitive phenotype of pck1 mutants.

Activation of Rho family GTPases leads to the assembly of contractile actin:myosin filaments (Jaffe and Hall, 2005). In S. pombe, actin patch disassembly is one of the effects of Rho1p depletion. Interestingly, rgf1 cells showed a defect in the actin reorganization required for the transition from monopolar to bipolar growth. Among the genes required for NETO, tea1⁺ plays a critical function; the most penetrant phenotype of tea1 mutants is their failure to initiate growth at the new cell tip, such that these cells only grow in a monopolar manner (Verde et al., 1995; Mata and Nurse, 1997). Tea1p has been found to form a large protein complex; during NETO, the tea1p-complex at the cortex interacts with formins (and probably other polarity factors), triggering actin cable assembly and polarized cell growth (Martin et al., 2005). In a tea1 mutant, Rgf1p was not maintained at one of the new cell ends, and the cells did not grow at that end (unpublished data). Rgf1p may function downstream from Tea1p, because Tea1p is required to recruit Rgf1p to a new end. The identification of proteins that directly interact with Rgf1p will be necessary to understand how Rho1p participates in the transition in which monopolar cells initiate bipolar growth.

Mutants defective in monopolar growth, tea1, tea4, and bud6, grew at wild-type rates. However, a novel aspect of the rgf1 mutants is that their failure to initiate bipolar growth was accompanied by cell lysis. In a cdc10-129 rgf1 mutant, which at the restrictive temperature arrested in G1, before the activation of bipolar growth, 45% of the cells lysed 30 min after release from the restrictive temperature, whereas in a cdc25-22 rgf1 mutant, which at high temperature arrested in G2 with both ends growing, the highest percentage of lysis (55%) after release was seen after the first round of septation, coinciding in time with bipolar growth activation. Our current model is that activation of Rho1p, and in consequence -GS activation, during bipolar growth is not achieved properly in rgf1 mutants, producing cell wall weakness. To our knowledge, rgf1⁺ is the first gene that has been implicated in cell wall biogenesis and NETO and might well provide a link between these two processes.

Rgf1p is the second exchange factor identified for Rho1p. Why does Rho1p have multiple GEFs? A similar situation has been described for mammalian cells, where the number of Rho-GEFs (69 members) exceeds the number of Rho-type GTPases (so far 22 members; Rossman et al., 2005). An attractive hypothesis is that the GEF could determine the downstream signaling specificity of Rho GTPases. This has been suggested for Ras signaling in fission yeast, where two GEFs, Ste6p and Efc25p, differentially regulate two Ras pathways (Papadaki et al., 2002). In agreement with such a hypothesis, we propose that Rgf1p would specifically activate the Rho1-GS complex during the transition from monopolar to bipolar growth, whereas Rgf3p, the former Rho1-GEF, would accumulate at the contractile ring, probably activating the Rho functions that coordinate cell-wall biosynthesis to maintain cell integrity during septation (Tajadura et al., 2004). Moreover, our results suggest that Rho1-GEFs, Rgf1p and Rgf3p, are not functionally interchangeable; each single rgf1 and ehs2-1 mutant was able to grow on plates at 37°C, whereas the double rgf1 ehs2-1 mutant was not.

In conclusion, here we provide evidence that Rgf1p is a new Rho1-GEF that participates in the regulation of bipolar growth and we propose that Rgf1p may coordinate a growth polarity transition with cell wall biosynthesis to prevent losses of cell integrity and allow cell expansion.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
We thank P. Perez, P. Coll, H. Valdivieso, A. Duran, and J. C. Ribas for plasmids, strains, and help along this work. C. Roncero is acknowledged for his very helpful comments. We also thank C. Castro for help with confocal microscopy. P.G. and I G. were supported by a fellowship from the Junta de Castilla y León and V. Tajadura acknowledges support from a fellowship granted by the Ministerio de Educación y Ciencia, Spain. Text was revised by N. Skinner. This work was supported by Grant BIO2001–1663 from the Comisión Interministerial de Ciencia y Tecnología, Spain and CSI02C05 from the Junta de Castilla y León.

	Footnotes

 
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E05-10-0933) on January 18, 2006.

Address correspondence to: Yolanda Sánchez (ysm{at}usal.es).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Arellano, M., Duran, A., and Perez, P. ((1997). ). Localization of the Schizosaccharomyces pombe Rho1 GTPase and its involvement in the organization of the actin cytoskeleton. J. Cell Sci. 110, , 2547–2555.[Abstract]

Arellano, M., Durán, A., and Pérez, P. ((1996). ). Rho1 GTPase activates the (1–3)b-D-glucan synthase and is involved in Schizosaccharomyces pombe morphogenesis. EMBO J. 15, , 4