Cytokinesis in yeast meiosis depends on the regulated removal of Ssp1p from the prospore membrane
2007; Springer Nature; Volume: 26; Issue: 7 Linguagem: Inglês
10.1038/sj.emboj.7601621
ISSN1460-2075
AutoresPeter Maier, Nicole Rathfelder, Martin G. Finkbeiner, Christof Taxis, Massimiliano Mazza, Sophie Le Panse, Rosine Haguenauer‐Tsapis, Michael Knop,
Tópico(s)DNA Repair Mechanisms
ResumoArticle8 March 2007free access Cytokinesis in yeast meiosis depends on the regulated removal of Ssp1p from the prospore membrane Peter Maier Peter Maier EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Nicole Rathfelder Nicole Rathfelder EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Martin G Finkbeiner Martin G Finkbeiner EMBL, Cell Biology and Biophysics Unit, Heidelberg, GermanyPresent address: International Agency for Research on Cancer (IARC), Unit of Gene-Environment Interaction, 150, cours Albert Thomas, 69008 Lyon, France Search for more papers by this author Christof Taxis Christof Taxis EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Massimiliano Mazza Massimiliano Mazza EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Sophie Le Panse Sophie Le Panse EMBL, Cell Biology and Biophysics Unit, Heidelberg, GermanyPresent address: Institut Jacques Monod-CNRS, Universites Paris VI and VII, 2 Place Jussieu, 75251 PARIS Cedex 05, France Search for more papers by this author Rosine Haguenauer-Tsapis Rosine Haguenauer-Tsapis EMBL, Cell Biology and Biophysics Unit, Heidelberg, GermanyPresent address: Institut Jacques Monod-CNRS, Universites Paris VI and VII, 2 Place Jussieu, 75251 PARIS Cedex 05, France Search for more papers by this author Michael Knop Corresponding Author Michael Knop EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Peter Maier Peter Maier EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Nicole Rathfelder Nicole Rathfelder EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Martin G Finkbeiner Martin G Finkbeiner EMBL, Cell Biology and Biophysics Unit, Heidelberg, GermanyPresent address: International Agency for Research on Cancer (IARC), Unit of Gene-Environment Interaction, 150, cours Albert Thomas, 69008 Lyon, France Search for more papers by this author Christof Taxis Christof Taxis EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Massimiliano Mazza Massimiliano Mazza EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Sophie Le Panse Sophie Le Panse EMBL, Cell Biology and Biophysics Unit, Heidelberg, GermanyPresent address: Institut Jacques Monod-CNRS, Universites Paris VI and VII, 2 Place Jussieu, 75251 PARIS Cedex 05, France Search for more papers by this author Rosine Haguenauer-Tsapis Rosine Haguenauer-Tsapis EMBL, Cell Biology and Biophysics Unit, Heidelberg, GermanyPresent address: Institut Jacques Monod-CNRS, Universites Paris VI and VII, 2 Place Jussieu, 75251 PARIS Cedex 05, France Search for more papers by this author Michael Knop Corresponding Author Michael Knop EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany Search for more papers by this author Author Information Peter Maier1, Nicole Rathfelder1, Martin G Finkbeiner1, Christof Taxis1, Massimiliano Mazza1, Sophie Le Panse1, Rosine Haguenauer-Tsapis1 and Michael Knop 1 1EMBL, Cell Biology and Biophysics Unit, Heidelberg, Germany *Corresponding author. EMBL, Cell Biology and Biophysics Unit, Meyerhofstr. 1, 69117 Heidelberg, Germany. Tel.: +49 6221 387631; Fax: +49 6221 387512; E-mail: [email protected] The EMBO Journal (2007)26:1843-1852https://doi.org/10.1038/sj.emboj.7601621 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Intracellular budding is a developmentally regulated type of cell division common to many fungi and protists. In Saccaromyces cerevisiae, intracellular budding requires the de novo assembly of membranes, the prospore membranes (PSMs) and occurs during spore formation in meiosis. Ssp1p is a sporulation-specific protein that has previously been shown to localize to secretory vesicles and to recruit the leading edge protein coat (LEP coat) proteins to the opening of the PSM. Here, we show that Ssp1p is a multidomain protein with distinct domains important for PI(4,5)P2 binding, binding to secretory vesicles and inhibition of vesicle fusion, interaction with LEP coat components and that it is subject to sumoylation and degradation. We found non-essential roles for Ssp1p on the level of vesicle transport and an essential function of Ssp1p to regulate the opening of the PSM. Together, our results indicate that Ssp1p has a domain architecture that resembles to some extent the septin class of proteins, and that the regulated removal of Ssp1p from the PSM is the major step underlying cytokinesis in yeast sporulation. Introduction In the bakers' yeast Saccharomyces cerevisiae, vegetative cell division is accompanied by formation of a bud, which is connected to the mother by the bud neck. Subsequent division processes involve polarized growth of the plasma membrane as well as sequestration of cytoplasmic contents including organelles and half of the nucleus through this connection into the daughter. A completely different morphogenetic program, called sporulation, is performed during the meiotic cell division. This time, the four meiotic progeny, the spores, are constructed entirely in the cytoplasm of the mother cell (Moens and Rapport, 1971; Peterson et al, 1972; Zickler and Olson, 1975). This type of cell division is not restricted to meiosis, but occurs in many fungal species (e.g. Ashbya gossippii and Coccidioides immitis) (Wendland and Walther, 2005) as an alternative mode of vegetative cell division, and is also common in protists (e.g. Toxoplasma gondii) (Shaw et al, 2000; Morrissette and Sibley, 2002). It is termed intracellular budding or endodyogeny. Interestingly, many of these species are pathogens, and in some species intracellular budding is specifically associated with their pathogenic form (e.g. Coccidioides immitis) (Miyaji et al, 1985; Nemecek et al, 2006). During intracellular budding, the nuclear divisions (one, two or several) are uncoupled from the physical cell division process and the nuclei become enwrapped at the end of their divisions by new membranous compartments, one per nucleus. This process then leads to physical separation of the new cells from the mother cytoplasm. In yeasts, these membranes are termed prospore membranes (PSMs, S. cerevisiae) or forespore membranes (Schizosaccharomyces pombe) (Shimoda, 2004; Neiman, 2005) and encompass compartments that initially resemble flattened pouches. They become assembled at the spindle pole bodies (SPBs) early in meiosis II (Okamoto and Iino, 1982; Davidow and Byers, 1984; Knop and Strasser, 2000). With progression through meiosis II, the four membranes grow out through the cytoplasm around lobes of the nucleus. Throughout this process, the SPBs are connecting the PSMs with the nuclear envelope to ensure the faithful inheritance of the genomes into the newly formed compartments. Simultaneously, cytoplasmic content such as secretory organelles, or mitochondria that associate during meiosis II with the nuclear envelope (Gorsich and Shaw, 2004), become enwrapped by the PSMs. At the end of the meiotic divisions, each of the four new nuclei is formed through fission of the nuclear envelope and subsequently fully engulfed by one PSM. The process of spore formation then proceeds by closure of the PSM. This generates two membrane bilayers on top of each other. The intervening space is then filled up with different layers of macromolecular compounds that together constitute the spore wall (Briza et al, 1988; Coluccio et al, 2004). Two protein structures specific for PSM formation have been described (for reviews, see Moreno-Borchart and Knop, 2003; Neiman, 2005). The meiotic plaque at the cytoplasmic face of the SPB is required for initiation of membrane formation (Knop and Strasser, 2000). It substitutes the mitotic outer plaque of the SPB and consists of three essential components (Mpc54p, Mpc70p and Spo74p) and one non-essential protein (Ady4p) (Knop and Strasser, 2000; Bajgier et al, 2001; Nickas et al, 2003). In the absence of meiotic plaques, precursors of the PSM cannot be delivered to the SPBs and remain as clusters in the cytoplasm. The precursors are characterized by their content of the proteins Ssp1p/Spo3p, Ady3p and Don1p, and some of them were also found to contain membrane markers, the t-SNAREs Sso1p and Sso2p (Knop and Strasser, 2000; Moreno-Borchart et al, 2001). Another essential structure associated with PSMs is a coat that covers the leading edge of the growing membrane, termed the LEP coat. It is built of Ssp1p, Ady3p and Don1p during initiation of PSM formation (Knop and Strasser, 2000; Moreno-Borchart et al, 2001; Nickas and Neiman, 2002). Whereas Don1p and Ady3p are not essential for PSM and spore formation, deletion of Ssp1p completely abolishes the formation of spores, and no LEP coat can be found in these mutants. Initiation of PSM formation at the SPBs was unaltered, but it acquired irregular shapes and often formed tubular structures that tightly enwrapped nuclear fragments. Also, minicompartments encircled by PSM-like membranes were visible. From these results it was concluded that the LEP coat functions in maintaining the opening during PSM assembly (Moreno-Borchart et al, 2001). It is an open question how the shape of the spores is regulated or how equal and efficient sequestration of membranous material to the four PSMs is achieved and controlled. Furthermore, the mechanism of closure of PSMs during meiotic cytokinesis is not known. These questions are particularly puzzling because actin and microtubules are not required for any of these steps (Gordon et al, 2006; Taxis et al, 2006). Thus, new mechanisms that compensate for actin-mediated polar transport can be expected. Here, we report on the role of Ssp1p during PSM formation and meiotic cytokinesis. We found that Ssp1p plays a role early in the process, in regulation of the dynamics of precursors of the PSM. Late in the process, during cytokinesis at the end of meiosis II, active removal of Ssp1p from the PSM substitutes for the need of a contractile activity, such as an acto-myosin ring. Detailed biochemical and genetic analysis of Ssp1p revealed a certain degree of functional conservation of this protein with septins by several criteria: phosphoinositide (PIP) binding, mediated by an N-terminal cluster of basic amino-acid residues, SUMO (Smt3p) modification near the C-terminus, different protein–protein interaction domains and the overall domain architecture. Only the classical signature of septins, the GTPase domain, could not be identified. Together, our work describes how the different steps of spore plasma membrane de novo biogenesis are organized, as well as the role Ssp1p plays within these processes. Results Ssp1p mediates clustering of exocytic vesicles Ssp1p is expressed exclusively during meiosis (Moreno-Borchart et al, 2001). In order to gain insight into potential activities of Ssp1p to regulate housekeeping machinery present in mitotic and meiotic cell division processes, we expressed Ssp1p in mitotic cells under the control of the strong inducible GAL1 promoter. This revealed that Ssp1p is toxic and prevents vegetative growth of the cells (Figure 1A). In these cells, Ssp1p localized to the plasma membrane with a preference to areas of membrane growth (buds of dividing cells; Figure 1B). Additionally, granulose structures near or inside the buds were visible in all cells (arrows in Figure 1B). Electron micrographs of Ssp1p-overexpressing cells revealed a 5–8-fold accumulation of secretory vesicles in the area of the emerging bud in small budded cells. Furthermore, the accumulated vesicles appeared to be smaller in size (35–60 nm compared with approximately 70 nm in the control cells) and a striking package of the vesicles into clusters was apparent (Figure 1C). In order to analyze the functioning of the secretory pathway in the Ssp1p overexpression strain, we investigated the biosynthesis of various secretory marker proteins in these cells (Avaro et al, 2002). This revealed no specific defects on the transport of these proteins through the secretory pathway (ER to late Golgi and vacuolar sorting, data not shown). This result and the accumulation of exocytotic vesicles in the bud therefore points to a defect late in secretion. To address this further, we analyzed the dynamics of vesicle delivery and fusion with the plasma membrane using a Sec2p-GFP fusion. Sec2p is a guanidine exchange factor for the Rab-like protein Sec4p that localizes predominantly to secretory vesicles at the bud tip or at the bud neck (Ortiz et al, 2002). In control cells Sec2p-GFP was hardly visible on mobile vesicles moving towards the bud tip or bud neck. In Ssp1p-overexpressing cells, however, bright and mobile Sec2p-GFP structures were seen that moved toward the bud. Figure 1D shows some frames derived from movies made from control and Ssp1p-overexpressing cells (movie showing more cells provided as Supplementary Movies S1 and S2). The increased brightness of the vesicles further suggests that not single but clusters of mobile structures are transported in the Ssp1p overexpression strain. This is consistent with the observation of clustered vesicles by electron microscopy (EM) (Figure 1C) and may point to a function of Ssp1p in mediating the formation of vesicle clusters. Figure 1.Ssp1p promotes secretory vesicle cluster formation in mitosis and meiosis. (A) Overexpression of Ssp1p is toxic for vegetative cell growth. SSP1 was expressed under the control of the GAL1 promoter from a low-copy (CEN, pKS89) or high-copy plasmid (2μ, pKS116) in cells of strain ESM356-1. Serial dilutions of cells containing the indicated plasmids were spotted on either glucose- or galactose/raffinose-containing plates and photographs were taken following incubation at 30°C after 2 days (glucose) or 3 days (galactose). (B) Localization of Ssp1p upon medium (CEN) or strong (2μ) overexpression following 3 h of induction of the GAL1 promoter using immunofluorescence microscopy (strains of (A)). Arrows point to Ssp1p-stained aggregates in the areas of the bud/budneck. (C) Visualization of secretory vesicles in wild-type cells and cells expressing GAL1-SSP1 from a chromosomal location. Cells of strains ESM356-1 (i) and YKS207-14 (ii; strain ESM356-1 containing several copies of GAL1-SSP1 integrated in the URA3 locus) were grown in the presence of galactose for 3 h and processed for electron microscopy. SV, secretory vesicles. Bar, 400 nm. (D) Dynamics of secretory vesicles in small budded cells visualized using a Sec2p-GFP fusion (strain YMF178). Frames from a control cell (plasmid without SSP1) and a cell expressing SSP1 from the GAL1 promoter (plasmid pKS89) are shown. Arrows indicate a Sec2p-GFP containing cluster that moves toward the bud (see Supplementary Movies S1 and S2). (E) Analysis of 65–70 nm vesicle distributions in cells in meiosis II. OsO4-fixed and EPON-embedded cells of a Δmpc54 Δmpc70 strain (YKS65; i) or a Δmpc54 Δmpc70 Δssp1 strain (YKS135; ii and iii) were used for this experiment. Magnifications from areas around the SPBs are shown. (F) Histogram of all distances between observed vesicles in cells (n=8) of each of the two strains. Coordinates of secretory vesicles were recorded manually using Metamorph™ software. Download figure Download PowerPoint When we overexpressed Ssp1p in vegetative cells, we found that low-level expression from a low-copy number plasmid using a weakened GAL1 promoter (GALS) had no effect in wild-type cells, but it was lethal in cells that lack either one of the two t-SNARE genes (SSO1 and SSO2) (Supplementary Figure S1). These proteins are required for vesicle fusion with the plasma membrane. We furthermore noticed that this weak overexpression of Ssp1p led to a significant reduction of the restrictive temperature of temperature-sensitive sec4 and sec2 mutants (data not shown), another two proteins that function also during vesicle fusion at the plasma membrane. This indicates that Ssp1p has a dominant-negative function on the level of the core machinery that encompasses vesicle fusion with the plasma membrane. To investigate the vesicle clustering function of Ssp1p in meiosis, we compared the distribution of secretory vesicles (65–70 nm) in the Δmpc54 Δmpc70 mutant (where PSM assembly is blocked and precursor structures are accumulated in the cells; Knop and Strasser, 2000) with the situation in the Δssp1 Δmpc54 Δmpc70 mutant. We used EM (Figure 1E) and recorded the position of visible 65–70 nm vesicles from cells as depicted in the figure. The histograms of the distances between the vesicles revealed that the additional deletion of SSP1 in the Δmpc54 Δmpc70 mutant leads to a more uniform scattering of vesicles throughout the cells, whereas in the Δmpc54 Δmpc70 mutant, an increased frequency of short distances between the vesicles is apparent (Figure 1F). This result fits well with our previous report that meiotic Δssp1 Δmpc54 Δmpc70 cells showed an even distribution of the t-SNARES Sso1p and Sso2p throughout the cytoplasm instead of the dot-like staining pattern typical for precursor membranes in the Δmpc54 Δmpc70 strain (Moreno-Borchart et al, 2001). Together, these findings suggest that Ssp1p has a function in clustering vesicles and that it is able to block specifically the fusion of vesicles with the plasma membrane. Ssp1p binds to PI(4,5)P2 at the plasma membrane The observed localization of Ssp1p to membranes (Figure 1) indicates that Ssp1p interacts with lipids or specifically localized proteins, or both. To address this further, we tested the binding of Ssp1p to serial dilutions of all biologically relevant PIP species, as well as other lipids, spotted on membranes using a previously described overlay assay (Kanai et al, 2001). We found that the N-terminal fragment of Ssp1p (amino acids (aa) 1–269) as well as the full-length protein (not shown) did bind to PIPs (Figure 2A), but not to other phospholipids (data not shown). Ssp1p showed highest affinity to PI(4,5)P2. Figure 2.Protein–protein and protein–lipid interactions of Ssp1p. (A) Purified N-terminally 6HisT7-tagged Ssp1p (aa 1–269) was used to probe a nitrocellulose membrane containing spots of serial dilutions of the indicated lipids. Binding of Ssp1p was detected using a specific antibody that recognizes the T7 tag (Novagen). No signal was detected using an unrelated 6HisT7-tagged protein (data not shown). (B) Localization of GFP-Ssp1p1–131 and GFP-Ssp1p1–269 (in strain ESM356). (C) Localization of GFP-Ssp1p1–131 in WT (strain SEY6210) and the mss4-102 mutant (strain AAY202) at 26°C and following a shift to 38°C for 20 min. The pictures show sections acquired from the center of the cells using a spinning disc confocal microscope (Perkin-Elmer). (D) PIP binding of Ssp1p depends on a cluster of positively charged residues close to the N-terminus of the protein and not on the net charge of the protein. Experimental setup as for (A), but using equal amounts of a purified Ssp1p or Ssp1p* fragment governing aa 1–269 as a bait. (E) Plasma membrane binding but not binding to the structures inside or close to the bud is dependent on PIP binding of Ssp1p. Localization of full-length Ssp1p and the PIP binding-deficient mutant Ssp1p* following overexpression from a CEN-GAL1 plasmid in vegetative cells (of strain ESM356). Immunofluorescence microscopy was performed and pictures were taken using a confocal microscope. Download figure Download PowerPoint Overexpression of GFP-tagged Ssp1p fragments revealed that the first 131 aa are sufficient to localize to the plasma membrane (Figure 2B). In contrast to the full-length protein and the fragment spanning aa 1–269 this construct failed to localize to vesicles inside the bud and it no longer enriched in the area of the bud (Figure 2B). This indicates that different domains of the proteins mediate plasma membrane binding and binding to vesicles and that the PI(4,5)P2 binding domain resides in the N-terminal 131 aa. In order to investigate whether plasma membrane binding of GFP-Ssp1p1–131 depends on PI(4,5)P2 in vivo, we used a temperature-sensitive mss4-102 mutant, which is conditionally defective in the only PI4P-5-kinase (Audhya and Emr, 2002; Stefan et al, 2002). In this mutant, GFP-Ssp1p1–131 disappeared from the plasma membrane and localized to cytoplasmic structures within 10–20 min after a shift to the restrictive temperature (Figure 2C). A similar result was previously also obtained for the PI(4,5)P2-specific PH domain of PLCδ (Stefan et al, 2002). Direct binding of proteins to PIPs usually involves clusters of negatively charged amino-acid residues. We inspected the amino-acid sequence of the N-terminal domain (aa 1–131) of Ssp1p and found one such cluster. To test the involvement of these residues (K24, K26 and K27) in PIP binding, we substituted them with alanine. This mutant of Ssp1p (called Ssp1p*) was no longer able to bind to PIPs using the spot blot method. In contrast, a mutant with three Lys → Ala substitutions in the central domain of the protein exhibited unchanged lipid-binding properties (Figure 2D). This result clearly demonstrates that the N-terminal basic cluster of residues is mediating PIP binding and not the net charge of the protein (pI=5.6). In order to address the function of PI(4,5)P2 binding of Ssp1p in vivo, we tested whether overexpression of Ssp1p* is still toxic for the cells. This was still the case (data not shown), however, the cellular localization of Ssp1* was changed compared with the WT protein. The protein was no longer able to bind to the plasma membrane, whereas it still stained the structures present in the bud of dividing cells (Figure 2E). This confirms the previous notion that a separate function of Ssp1p mediates vesicle localization, independent of the PIP-binding capability of the N-terminal domain. In order to get further insight into the domain architecture of Ssp1p, we investigated the subclones of Ssp1p for toxic effects upon mitotic overexpression and two-hybrid interaction with its known binding partner Ady3p (Moreno-Borchart et al, 2001) (Figure 3). This analysis revealed that the N-terminal half (aa 1–269) of Ssp1p mediates self-interaction, whereas the C-terminus (aa 217–572) binds to Ady3p. Toxicity of Ssp1p requires aa 1–160. Together, our results demonstrate that distinct domains within Ssp1p mediate membrane binding, toxicity and protein–protein interaction. Figure 3.Subcloning of the different domains of Ssp1p. PIP binding was assayed using purified proteins and the blot technique of Figure 2A. Toxic growth effects were assayed as described in the legend to Figure 1A (using 2μ-GAL1 plasmids). Self interaction and interaction with Ady3p were determined using the two-hybrid system. Localization to the plasma membrane was performed using live cell imaging of GFP fusions and untagged constructs and immunofluorescence microscopy for all subcloned fragments of Ssp1p with two methods (GFP fusions in living cells (Figure 2B) and immunofluorescence microscopy (Figure 2E)) (ND, not determined). Download figure Download PowerPoint PI(4,5)P2 binding of Ssp1p influences the movement of precursor structures Previously, it has been shown that PI(4,5)P2 is present at the plasma membrane of sporulating cells and at the membranes of the spores, as soon as they become visible (Nakanishi et al, 2004; Rudge et al, 2004). Using the GFP-PLCδPH fusion we confirmed, that in fact PI(4,5)P2 can be found at the plasma membrane also during earlier stages of sporulation, before the PSMs become assembled (during late stages of meiosis I) (data not shown). At this stage, Ssp1p localizes to 10–30 punctuate structures inside the cells, which are the precursors of the PSMs (Moreno-Borchart et al, 2001). We recently investigated in detail the assembly of PSMs from these precursor structures and found that actin-dependent as well as Brownian movements of precursors occur, and that actin-dependent transport is restricted to areas underneath the plasma membrane of the cell (Taxis et al, 2006). It could therefore be that PI(4,5)P2 binding of Ssp1p mediates interaction of precursors with the plasma membrane and thereby influences the movements of precursors. To test this idea, we used time-lapse microscopy. We addressed whether precursor movements are changed in the strain that expresses Ssp1p* as compared with WT. To follow precursor movements, we used Don1p-GFP as a specific marker (Knop and Strasser, 2000). Don1p colocalizes with Ssp1p to precursors and the LEP coat of the PSMs (Moreno-Borchart et al, 2001). With frame rates of 1 frame/∼4 s and projections of the entire cells, we found that the precursors of the Ssp1p*-expressing strain exhibit ∼10% faster movements as compared with WT (Figure 4A). This difference, although not large, is significant because the analysis is based on more than 10 000 single measurements per strain (using automated object tracking) and in three independent measurements (P<0.001 using t-test analysis). In order to have an internal control, we additionally measured the movements of Don1p-GFP precursors at the SPBs in cells in early phases of meiosis II in the same movies (as an internal control). The SPBs can easily be recognized by their brighter decoration with Don1p-GFP (Knop and Strasser, 2000) and their pairwise movements, which is caused by the short metaphase spindles that connect them (Taxis et al, 2006). This revealed that the observed movements of the SPBs in the WT and the Ssp1p* strains were exactly the same (0.4% difference). Using high frame rates (1 frame/∼0.2 s) and single plane live cell recording, we noticed a 22% faster movement of the Ssp1p* precursors as compared with the precursors in the Ssp1p strain (Figure 4A). In this case, we could not identify SPBs (due to missing spatial information). Figure 4.PIP binding of Ssp1p is required for fast precursor movements and phosphorylation of Ssp1p. (A) Altered dynamics of precursor structures (visualized using Don1p-GFP) in meiotic cells expressing Ssp1p* as compared with wild type Ssp1p. Movies were recorded following 5–6 h after induction of sporulation. Don1p-GFP movies were analyzed using the automated object tracking function of Metamorph™. Movements of about 200–300 individual precursor structures per strain were recorded over 50 frames. Two clones per strain were analyzed. One movement corresponds to the movement of a precursor structure from one frame to the next in the movie. Whole-cell projection (3.77 s/frame) and single-section (0.21 s/frame) recordings were used. The dynamics of LEP coats was analyzed in whole-cell projections (n=30 cells per strain) (strains: YKS65 containing pRS41H-SSP1 (pMM80) or SSP1*). Bars indicate standard deviations of the mean velocities. (B) Electrophoretic mobility of Ssp1p and Ssp1p* in extracts of mitotic and meiotic cells (mitotic cells: strain ESM356 containing p416-GAL1-SSP1 (pKS89) or SSP1*; meiotic cells: strains of (A)). For meiotic cells, alkaline phosphatase (CIP)-treated extracts without and with inhibitors (50 mM 3-glycero phosphate, 50 mM NaF) were analyzed as well. Download figure Download PowerPoint Together, these results indicate that the PIP binding of Ssp1p reduces the movements of precursor structures. The observation of larger differences in faster movies is consistent with the idea that the movements are due to Brownian motion. Therefore, the difference between the movement of Ssp1p and Ssp1p* precursors may best be explained by weak interactions of Ssp1p with the plasma membrane, which inhibits Brownian movements. Next, we performed immunoblotting with meiotic cell extracts from the cells used for the analysis shown in Figure 4A and also from vegetative cells with overexpressed Ssp1p and Ssp1p* proteins. As can be seen in Figure 4B, protein levels for Ssp1p and Ssp1p* were comparable; however, in both the mitotic and the meiotic cells, WT Ssp1p showed an additional band with reduced mobility on the gel. Also, Ssp1p and Ssp1p* from meiotic cells showed multiple bands with reduced electrophoretic mobility compared with the mitotically expressed Ssp1p species. To test whether these different mobilities were due to phosphorylation, we treated meiotic extracts with alkaline phosphatase. In mitotic extracts, this shifted the bands of Ssp1p to the same position as the bands of Ssp1p*; however, in the meiotic extracts, both proteins still showed different bands (Figure 4B). This suggests that impaired PI(4,5)P2 binding concomitantly leads to reduced phosphorylation of Ssp1p. Additionally, this experiment revealed the presence of another modification of Ssp1p. Ssp1p is sumoylated In order to address the nature of the Ssp1p modification, we tested whether meiotically expressed Ssp1p as well as ectopically expressed Ssp1p from vegetative cells is modified by the small ubiquitin-like protein SUMO (Smt3p in yeast; Johnson and Blobel, 1999). As shown in Figure 5A, Ssp1p from meiotic cells was indeed sumoylated and migrated approximately 20 kDa above the unmodified version. In vegetative cells, only very little sumoylation was visible, suggesting meiosis-specific regulation of Ssp1p sumoylation. Figure 5.Ssp1p is modified by Smt3p/SUMO. (A) Smt3p/SUMO modification of Ssp1p in meiotic and vegetative cells (upon CEN-GAL1 expression). Im
Referência(s)