In budding yeast, two classes of post-Golgi secretory vesicles carrying different sets of cargoes typified by Bgl2p and invertase are delivered to the plasma membrane for secretion. The exocyst is implicated in tethering these vesicles to the daughter cell membrane for exocytosis. In this study, we report that mutations in the exocyst component Exo70p predominantly block secretion of the Bgl2p vesicles. Furthermore, a defect in invertase vesicle trafficking caused by vps1Δ or pep12Δ in the exo70 mutant background is detrimental to the cell. The secretion defect in exo70 mutants was most pronounced during the early budding stage, which affected daughter cell growth. The selective secretion block does not occur at the vesicle formation or sorting stage because the exocytic vesicles are properly generated and protein processing is normal in the exo70 mutants. Our study suggests that Exo70p functions primarily at early stages of the cell cycle in Bgl2p vesicle secretion, which is critical for polarized cell growth.
The budding yeast Saccharomyces cerevisiae undergoes polarized growth and provides an excellent model system for the study of cell polarity. Yeast cells are surrounded by a rigid cell wall, which protects the cells from the environment but also physically restrains membrane expansion. For budding, cells not only need to deliver proteins and lipids to the bud membrane but also need to secrete enzymes to remodel the cell wall in order to allow surface expansion. Harsay and Bretscher (1995) discovered that yeast cells have two distinct classes of exocytic vesicles carrying different sets of cargoes for secretion. One class carries plasma membrane proteins and cell wall modification enzymes such as Bgl2p (hereafter referred to as Bgl2p vesicles), whereas the other class carries proteins such as the periplasmic enzyme invertase (hereafter referred to as invertase vesicles). Several proteins are implicated in the generation of specific vesicles (David et al., 1998; Gall et al., 2002). However, whether the tethering or docking of these vesicles at the plasma membrane is differently regulated is unknown.
The exocyst is an octameric protein complex composed of Sec3p, Sec5p, Sec6p, Sec8p, Sec10p, Sec15p, Exo70p, and Exo84p. These proteins are localized to the bud tip, the site of active exocytosis and cell surface expansion, where they tether the post-Golgi secretory vesicles to the plasma membrane before fusion (for reviews see Guo et al., 2000; Hsu et al., 2004). In the present study, we report that Exo70p functions primarily at early stages of the cell cycle to regulate the secretion of specific vesicles that are critical for polarized cell growth.
Results And Discussion
The exo70 mutants accumulate post-Golgi secretory vesicles
Exo70p was initially identified through purification of the exocyst complex from yeast lysates (TerBush and Novick, 1995). To study the function of Exo70p, we generated exo70 mutant strains. The exo70-35 mutant has a growth defect that is more pronounced at temperatures ≤25°C, whereas exo70-38 is a temperature-sensitive mutant that grows normally at 25°C but fails to grow above 35°C (Fig. S1). To study secretion in these mutants, we examined the cells that were shifted to their restrictive temperatures by thin-section EM. As shown in Fig. 1 A, the exo70-35 and exo70-38 cells accumulated post-Golgi vesicles that were 80–100 nm in diameter.
These vesicles were preferentially distributed in the daughter cells. Consistent with the EM result, immunofluorescence staining of Sec4p, the Rab protein residing on the post-Golgi vesicles, was polarized to the bud in both mutants (Fig. 1 B). These results indicate that the polarized delivery of exocytic vesicles is not affected in the exo70 mutants, which is consistent with previous studies indicating that the exocyst functions at the vesicle-tethering step after vesicles are transported to the daughter cells (for reviews see Guo et al., 2000; Hsu et al., 2004).
exo70 mutants are primarily defective in the secretion of Bgl2p vesicles
All of the previously identified exocyst mutants were defective in invertase secretion (Novick et al., 1980; Zhang et al., 2005b). Therefore, we examined invertase secretion in the exo70 mutants. Surprisingly, the exo70-35 and exo70-38 mutants were able to secrete invertase at levels that were close to the wild-type cells (Fig. 2 A).
As controls, all of the other exocyst mutants showed substantial invertase secretion defects at the restrictive temperature (Fig. 2 A). We next examined the secretion of Bgl2p in these mutants. As shown in Fig. 2 B, exo70-35 accumulated Bgl2p at both 25 and 34°C, and exo70-38 accumulated Bgl2p at the restrictive temperature of 37°C. The newly accumulated Bgl2p in exo70-38 after shifting to 37°C was comparable with other temperature-sensitive exocyst mutants. The growth defect of exo70-38 is also similar to other exocyst mutants (Fig. S1 B). To better analyze the secretion of these cargoes, we performed pulse-chase experiments. As shown in Fig. 2 C, the wild-type cells secreted >90% of the newly synthesized Bgl2p within 30 min at 37°C, whereas exo70-38 only secreted ∼40% of Bgl2p. sec10-2 also showed the Bgl2p secretion defect, which was less severe than that in exo70-38. On the other hand, exo70-38 secreted >80% of the newly synthesized invertase within 30 min at 37°C, whereas sec10-2 only secreted ∼30% of invertase under the same condition (Fig. 2 D).
To further investigate the selectivity of vesicle block in exo70 mutants, we separated the vesicles by density gradients as previously described (Harsay and Bretscher, 1995; Harsay and Schekman, 2002). As shown in Fig. 2 (E–G), both exo70-35 and sec6-4 accumulated Bgl2p vesicles, whereas only a diminutive amount of invertase vesicles was detected in exo70-35. Our results confirmed an early prediction (Harsay and Bretscher, 1995; Harsay and Schekman, 2002) that certain secretory mutants may be specifically defective in one branch of the exocytic pathways. To date, all of the other sec mutants have been characterized by their block of both classes of vesicles (Novick et al., 1980; Zhang et al., 2005b; unpublished data). Thus, the exo70 mutants are unique in their selectivity in Bgl2p vesicle block. We do not totally exclude the possibility that new exo70 mutant alleles that have more detectable invertase secretion defects can be identified. However, analyses of the two exo70 mutant alleles that are different in nature (Fig. S1) strongly suggest that Exo70p primarily mediates the secretion of Bgl2p vesicles.
Cargoes in the invertase vesicles travel through endosomal compartments before reaching the plasma membrane. Vps1p is a dynamin implicated in the formation of vesicles that are destined to the endosomes from the TGN (Nothwehr et al., 1995). Pep12p is a SNARE protein involved in the fusion of Golgi or early endosomal vesicles with the late endosomes (Gerrard et al., 2000). In the vps1Δ or pep12Δ cells, invertase is delivered to the cell surface by routing through the Bgl2p pathway (Gurunathan et al., 2002; Harsay and Schekman, 2002). We found that exo70-35 was synthetic lethal with vps1Δ or pep12Δ (Fig. 2 H). This result suggests that in the exo70-35 cells that are constitutively defective in Bgl2p vesicle secretion, further disruption of the other exocytic route is detrimental to the cell.
We have also analyzed carboxypeptidase Y (CPY) processing and found that this protein is correctly sorted, and there is no kinetic delay for CPY traffic in the exo70 mutant (Fig. S2 A). The normal processing of CPY together with the evident accumulation of post-Golgi secretory vesicles suggest that protein sorting to vacuole and vesicle formation at the TGN is not defective in exo70 mutants. Rather, the selective block of Bgl2p vesicle exocytosis is likely imposed at the plasma membrane stage, which suggests that different exocytic pathways can be specifically regulated at the target membrane as well as at the TGN.
Cellular localization and structural integrity of the exocyst complex in
Using fluorescence microscopy, we examined exocyst localization in the exo70 mutants. As shown in Fig. 3 A, all of the exocyst subunits were polarized, including the exo70-35 mutant protein itself.
We also found that the immunoisolated exocyst complex was intact, and the exo70-35 protein was able to associate with other exocyst proteins in the complex (Fig. 3 B). Similar results were obtained from the exo70-38 cells (Fig. S2 B). Because the localization and composition of the exocyst complex are mostly unaffected in the exo70 mutants, the secretion defects are most likely caused by the malfunction of Exo70p itself.
exo70 mutant cells accumulate vesicles at early stages of the cell cycle
During our EM analyses of the exo70 cells, we noted that vesicle accumulation was much more prominent in small-budded cells than in large-budded cells. The correlation between bud size and vesicle number suggests that the secretion block in exo70 mutants mostly occurs during early stages of budding. In contrast, the other exocyst mutants accumulate secretory vesicles at various stages of the cell cycle (Fig. S3). Failure in secretion results in an ∼5% increase of cell density, which allows separation of the sec mutant cells and wild-type cells by density gradients (Novick et al., 1980). In the present study, wild-type and mutant cells were subjected to Percoll density gradients. As shown in Fig. 4 A, the wild-type cells were distributed to a single low density fraction, and the sec6-4 cells were accumulated at a single high density fraction.
However, the exo70-35 cells were distributed in both the low and high density fractions. We found that most of the large-budded cells (93%; n = 250) were present in the lighter fraction. In contrast, the denser fraction contained very few large-budded cells; instead, most were small- or medium-budded cells corresponding to the vesicle-accumulating population observed by EM. The exo70-35 cells with lower density were not caused by incomplete phenotypic penetration or spontaneous reversions because cells collected from the two density peaks were equally defective in growth on plates (unpublished data). Besides sec6-4 and exo70-35, we have also analyzed exo70-38 and other exocyst mutants, including sec3-2, sec10-2, and exo84-121. Although the exo70-38 cells were distributed to both heavy and light fractions, the other exocyst mutants were all distributed to a single high density fraction (Fig. 4 B, top). Further microscopic analysis of the cells obtained from the exo70-38 heavy fraction demonstrated that most of these cells were in their small- or medium-budded stage, whereas cells obtained from the other mutants were heterogeneous in budding stages (Fig. 4 B, bottom). This study clearly indicates that exo70 mutants block secretion primarily at early stages of the cell cycle.
It has been proposed that the two types of secretory vesicles play distinct roles in yeast cells (Harsay and Bretscher, 1995; Harsay and Schekman, 2002). The invertase vesicles carry proteins such as invertase, acid phosphatase, and possibly the general amino acid permease (Roberg et al., 1997), which are secreted under certain physiological conditions. The Bgl2p vesicles contain materials for plasma membrane expansion and cell wall remodeling. We examined whether block of the Bgl2p vesicle pathway in exo70 mutants could affect daughter cell growth. Budding of the wild-type and exo70-35 cells was compared at various time points after release from α-factor arrest at G1. As shown in Fig. 5 A, by 60 min, the wild-type cells exhibited much larger bud sizes than the mutant cells (1.5-fold; n = 150).
Clearly, block of the Bgl2p vesicle alone is sufficient to cause a delay in daughter cell growth. Furthermore, because secretion in the exo70 mutants is primarily blocked at early stages of the cell cycle, we speculated that growth defects of the mutant cells would be more prominent during early budding. To test this, we examined the growth of the temperature-sensitive exo70-38 mutant synchronized at different cell cycle stages. We first assessed the growth of cells at the early budding stage by examining bud formation from cells released from G1 at 37°C. As shown in Fig. 5 B, the majority of the wild-type cells were capable of generating buds after 60 min. In contrast, most of the exo70-38 cells were either unbudded or at the tiny-budded stage. We then tested whether growth of the cells at later stages of the cell cycle was affected in exo70-38. Cells released from G1 were first incubated at 25°C for 90 min to generate large-budded cells and were shifted to 37°C for various periods of time. We found that both the wild-type and exo70-38 cells were able to complete cytokinesis after 30 min of incubation at 37°C, as revealed by the disconnection of the cytoplasm between the mother and daughter cells. However, after cytokinesis, the exo70-38 cells were arrested at the next budding stage, whereas the wild-type cells were able to proceed to new rounds of the cell cycle (Fig. 5 C). On the other hand, the sec10-2 cells were defective in both budding and later stages of cell growth (Fig. 5, B and C).
The selective block of Bgl2p vesicles primarily during budding in exo70 mutants suggests that the function of Exo70p in exocytosis is spatially and temporally coupled to polarized cell growth. The exocyst complex is under the control of Rho GTPases (Novick and Guo, 2002). It was previously reported that a Cdc42p mutant, cdc42-6, specifically accumulated Bgl2p vesicles at early stages of the cell cycle (Adamo et al., 2001). The phenotypical similarity between cdc42-6 and the exo70 mutants suggests that Cdc42p and Exo70p function together during polarized cell growth. One possibility is that Cdc42p activates Exo70p during yeast budding. We have examined the interaction between Exo70p and Cdc42p in vitro using recombinant fusion proteins. However, the interaction was very weak, and there was no clear nucleotide dependence in the binding (unpublished data). Nevertheless, Cdc42p may regulate Exo70p function via intermediate molecules.
Exo70p functions in concert with the other exocyst components, which mediate the secretion of both the invertase and Bgl2p vesicles. Future studies may likely identify specific mutant alleles in those subunits that also preferentially affect Bgl2p secretion. Overall, characterization of the exo70 mutants revealed the intimate coupling between exocytosis and polarized cell growth. Future studies of the exo70 mutants will help us understand the nature of the Bgl2p vesicle pathway and how this pathway is linked to polarized cell growth and cell cycle progression.
Materials And Methods
Yeast strains and procedures
Standard methods were used for yeast media, growth, and genetic manipulations. The yeast strains used in this study are listed in Table S1 . The exo70-38 mutant was isolated by random mutagenesis. For generation of the exo70-35 mutant, site-directed mutagenesis on the C terminus of Exo70 was performed on a CEN TRP1 EXO70 plasmid. The mutant alleles were introduced into GY2810, in which the endogenous EXO70 gene was disrupted by HIS3 and supplemented with a CEN URA3 EXO70 plasmid. The transformants were replicated onto synthetic complete media (SC) plates containing 5-fluoroorotic acid (5-FOA) to select for the loss of the CEN URA3 EXO70 plasmid. To generate exo70-35 vps1Δ and exo70-35 pep12Δ double mutants, vps1Δ and pep12Δ strains in which the endogenous EXO70 was disrupted by HIS3 and supplemented with CEN URA3 EXO70 were transformed with a CEN LEU2 EXO70 or CEN LEU2 exo70-35 plasmid. The transformants were replicated on SC plates containing 5-FOA to select for the loss of the CEN URA3 EXO70 plasmid.
EM and light microscopy
Wild-type and exo70-35 cells were grown at 34°C in yeast extract/peptone/glucose (YPD) medium to early log phase and were shifted to 25°C for 3 h. The exo70-38 and sec10-2 cells were grown at 25°C and shifted to 37°C for 2 h. Cells were processed for thin-section transmission EM using a transmission electron microscope (model 1010; JEOL) at 100,000× as previously described (Zhang et al., 2005a). Fluorescence microscopy of the GFP-tagged exocyst and immunofluorescence observation of Sec4p were performed as previously described (Zhang et al., 2005b). The images were captured by a fluorescence microscope (DM IRB; Leica) using a 100× objective and a high resolution CCD camera (ORCA-ER; Hamamatsu) and were analyzed by OpenLab 3.1.4 software (Improvision).
Invertase and Bgl2p secretion assay
Invertase assays were performed as described previously (Zhang et al., 2005a). For exo70-38 and other exocyst mutants, the cells were first shifted to 37°C for 1 h and then were grown in low glucose medium (0.1% glucose) at the same temperature for 1 h. For exo70-35, cells were started at 34°C and were transferred to 25°C for 1 h before shifting to low glucose medium for 1 h and processing for the assay. Bgl2p assays were performed as described previously (Adamo et al., 2001). For exo70-38 and other exocyst mutants, the cells were shifted to 37°C for 2 h. For exo70-35, cells were started at 34°C and were transferred to 25°C for 2 h before being processed for the Bgl2p assay.
To analyze the secretion of newly synthesized Bgl2p, cells were cultured to early log phase at 25°C and were shifted to SC lacking methionine/cysteine at 37°C for 15 min. The cells were then labeled with [35S]methionine/cysteine for 10 min and chased for various times at 37°C. Bgl2p from external and internal fractions were immunoprecipitated using anti-Bgl2p polyclonal antibody and were analyzed by SDS-PAGE and autoradiography. Accordingly, to analyze the secretion of newly synthesized invertase, the wild-type, exo70-38, and sec10-2 cells were cultured to early log phase at 25°C and shifted to 37°C for 15 min. Cells were then shifted to low glucose medium for 10 min to derepress the invertase expression (pulse). The chase was initiated by adding 2% glucose to the culture to repress invertase expression. The secretion of newly synthesized invertase at various chase times was determined by invertase enzymatic assay.
Separation of vesicles by density gradients
Vesicle fractionation was performed as previously described (Harsay and Schekman, 2002). Yeast cells (100-ml culture per gradient) were grown in YPD medium to early log phase and shifted to YP plus 0.1% glucose at the restrictive temperatures to induce invertase expression. Cells were spheroplasted, lysed, and subjected to differential centrifugation. The pellets from the 100,000-g spin were fractionated on 20–55% Percoll step gradients (4 ml in TLA100.3 tubes [Beckman Coulter]), and 160-μl fractions were collected for invertase and Bgl2p assays.
Percoll density gradient separation of the cells
The wild-type and exo70-35 cells were grown at 34°C to early log phase and were shifted to 18°C for 3 h. The cells were washed and resuspended in 0.2 ml of synthetic minimal media. The cells were then layered on a 70% Percoll/synthetic minimal medium mixture and centrifuged for 20 min at 25,000 g in TLA100.3 tubes. For exo70-38 and other exocyst mutants, the cells were shifted for 2 h at 37°C before centrifugation.
Synchronization of cells by α-factor arrest
For synchronization of exo70-35 cells, early log-phase cells were arrested at G1 phase using 5 μM α factor for 2.5 h at 34°C. The cells were then released into fresh YPD at 25°C for various periods of time and were collected for light microscopy. To examine the growth of exo70-38 cells at different cell cycle stages, cells were treated with α factor for 2.5 h at 25°C and either directly released into fresh medium at 37°C or first released at 25°C for 90 min to generate large buds followed by a shift to 37°C for various times. Cells were then collected for light microscopy observation. The completion of cytokinesis was determined by staining with 1 mg/ml DAPI.
Online supplemental material
Table S1 lists the major yeast strains used in this study. Fig. S1 shows the growth properties of the exo70 mutants and the mutation sites on the protein. Fig. S2 A shows CPY processing in exo70-35 by pulse-chase assay. Fig. S2 B is the immunoprecipitation experiment using radiolabeled yeast lysate showing that the isolated exocyst complex is mostly intact in the exo70-38 mutant. Fig. S3 displays the EM images of representative exo70-35, exo70-38, and sec10-2 cells at their small-budded and large-budded stages.
This work is supported by the National Institutes of Health (grant RO1-GM64690), American Cancer Society, and the Pew Scholars Program (grants to W. Guo).
B. He and F. Xi contributed equally to this paper.
Abbreviations used in this paper: 5-FOA, 5-fluoroorotic acid; CPY, carboxypeptidase Y; SC, synthetic complete media; YPD, yeast extract/peptone/glucose.