Rewiring a Rab regulatory network reveals a possible inhibitory role for the vesicle tether, Uso1

Significance

Members of the Rab family act as molecular switches to control distinct stages of vesicular traffic into and out of a eukaryotic cell. Many Rabs are linked to the adjacent Rab by their regulators, generating circuits that direct transport along the pathway. We have rewired one such regulatory circuit by exploiting a chimeric Rab that acts early on the pathway but can be activated by a regulator that is situated late on the pathway. Expression of this chimeric Rab leads to a block late on the pathway. Our findings confirm the importance of Rabs in establishing membrane identity.

Abstract

Ypt1 and Sec4 are essential Rab GTPases that control the early and late stages of the yeast secretory pathway, respectively. A chimera consisting of Ypt1 with the switch I domain of Sec4, Ypt1-SW1^Sec4, is efficiently activated in vitro by the Sec4 exchange factor, Sec2. This should lead to its ectopic activation in vivo and thereby disrupt membrane traffic. Nonetheless early studies found that yeast expressing Ypt1-SW1^Sec4 as the sole copy of YPT1 exhibit no growth defect. To resolve this conundrum, we have analyzed yeast expressing various levels of Ypt1-SW1^Sec4. We show that even normal expression of Ypt1-SW1^Sec4 leads to kinetic transport defects at a late stage of the pathway, with secretory vesicles accumulating near exocytic sites. Higher levels are toxic. Toxicity is suppressed by truncation of Uso1, a vesicle tether required for endoplasmic reticulum–Golgi traffic. The globular head of Uso1 binds to Ypt1 and its coiled-coil tail binds to the Golgi-associated SNARE, Sed5. We propose that when Uso1 is inappropriately recruited to secretory vesicles by Ypt1-SW1^Sec4, the extended coiled-coil tail blocks docking to the plasma membrane. This putative inhibitory function could serve to increase the fidelity of vesicle docking.

Rab GTPases are master regulators of vesicular transport, each typically controlling a particular stage of membrane traffic. Rabs are activated by guanine nucleotide exchange factors (GEFs) that catalyze the exchange of GDP for GTP and inactivated by GTPase-activating proteins (GAPs) that stimulate the slow intrinsic rate of GTP hydrolysis (1). Since GDP- but not GTP-bound Rabs are extracted from membranes by GDP dissociation inhibitor (2), the localization of a GEF plays a key role in defining the localization of its substrate Rab (3–5). Once in their GTP-bound state, each Rab recruits a distinct set of effectors, such as, motors that drive vesicle movement, tethers that direct vesicle-target recognition, and regulators of the SNAREs that catalyze fusion (6). By defining when and where each Rab enters and exits its GTP-bound form, GEFs and GAPs help to establish the functional identity and directionality of vesicular carriers (1).

While each Rab displays a distinctive pattern of localization, adjacent Rabs are often functionally linked to each other through their regulators (1). Here we focus on the yeast secretory pathway in which Ypt1, a Rab1 homolog, controls transport from the endoplasmic reticulum (ER) to the Golgi (7, 8); Ypt31 and Ypt32, redundant Rab11 homologs, control exit from the Golgi (9, 10), and Sec4, a Rab 8 homolog, promotes delivery of Golgi-derived secretory vesicles to sites of exocytosis (11). Sec2 is an efficient and specific GEF for Sec4 and, like Sec4, it is concentrated on the surface of Golgi-derived secretory vesicles (4, 12). The association of Sec2 with those vesicles relies upon its interaction with the GTP-bound forms of either of the upstream Rabs, Ypt31 and Ypt32 (13, 14). This relationship constitutes a Rab GEF cascade in which one Rab recruits the GEF that activates the next Rab. A directly analogous relationship has been established for the mammalian homologs in which Rab11-GTP recruits Rabin 8, a Sec2 homolog, which then activates Rab8 (15). Additional Rab GEF cascades have been identified on other trafficking pathways (16–21). Several Rab GAPs have been implicated in opposing GAP cascades (22–25). On the yeast secretory pathway Gyp1, a GAP that inactivates Ypt1 (26), is recruited to the late Golgi by binding to Ypt32-GTP, thus limiting the extent of overlap between the Ypt1 and Ypt32 domains (25). GEF and GAP cascades, working together in a counter-current fashion, could lead to a series of abrupt transitions from one Rab to the next as membrane flows along a pathway (1). In principle, by coupling localization domains that recognize the active form of one Rab to catalytic domains that either activate a downstream Rab or inactivate an upstream Rab, a defined program of Rab transitions could be specified that would lead to ordered changes in membrane identity (1).

Here we probe this model by exploiting a fascinating allele of YPT1 to rewire the Rab regulatory circuit on the yeast secretory pathway. In early studies, various domains of Ypt1 and Sec4 were swapped and the resulting chimeras analyzed in vivo to establish the structural basis of Rab specificity (27, 28). One such chimera was originally named Ypt1-EF^Sec4, to denote the replacement of the putative effector domain of Ypt1 with that of Sec4 (27), but subsequently renamed Ypt1-SW1^Sec4 to describe more accurately the structural element (switch 1) that had been swapped (29). When expressed at normal levels from the endogenous promoter, this chimera was found to function as the sole copy of the essential YPT1 gene without any significant growth defect or inhibition in the export of the cell wall enzyme invertase (27). This result implies that Ypt1-SW1^Sec4 must be able to interact productively with all of the essential Ypt1 regulators and effectors.

The Ypt1-SW1^Sec4 chimera resurfaced in the context of a crystallographic study of Sec2 bound to its substrate, Sec4 (29). The structure indicated that the main contacts with Sec2 involved the switch 1 and switch 2 domains of Sec4. As the switch 2 domain of Ypt1 is nearly identical to that of Sec4, this led to the prediction that replacing the switch 1 domain of Ypt1 with that of Sec4 would allow the chimera to interact with Sec2. The prediction held up: although Sec2 has no activity with Ypt1, it efficiently activates Ypt1-SW1^Sec4 in vitro (29). Taken together, these two studies present a conundrum: Ypt1-SW1^Sec4 fulfills all of the essential Ypt1 roles early on the secretory pathway, yet it can be activated in vitro by a GEF that resides at the end of the pathway. If Rabs define the identity of the membrane with which they are associated, then Ypt1-SW1^Sec4 expression should trigger havoc in the cell by leading to Ypt1 activity on the wrong part of the secretory pathway. Nonetheless, phenotypic analysis suggested that this was not the case (27). Here we consider three possible resolutions to this conundrum: (i) Sec2 does not actually activate Ypt1-SW1^Sec4 in vivo, despite the exchange activity observed in vitro; (ii) Ypt1-SW1^Sec4 expression is not as benign as it first appeared; or (iii) the proposal that Rabs define membrane identity is seriously flawed. Our studies not only confirm the importance of Rabs to membrane identity, they have also revealed a possible inhibitory function for the long, coiled-coil vesicle tether Uso1. This tether normally acts in response to Ypt1-GTP to form the initial link between ER-derived vesicles and the Golgi apparatus (30). We find that, when inappropriately recruited to post-Golgi secretory vesicles by Ypt1-SW1^Sec4, Uso1 acts to inhibit their fusion with the plasma membrane.

Results

To determine if Sec2 can activate Ypt1-SW1^Sec4 in vivo as it does in vitro (29), we conducted both genetic tests and localization studies. Ypt1 is normally activated by the TRAPP complexes and the Bet3 subunit is critical for this GEF activity (31). If Ypt1-SW1^Sec4 can be activated by Sec2p in vivo, its expression might alleviate the growth defect of a bet3 mutant. Fig. 1A shows that bet3-1 cells fail to grow at temperatures higher than 30 °C. Expression of Ypt1 from the GPD promoter allows somewhat stronger growth at 30 °C, but not at 33 °C, while expression of Ypt1-SW1^Sec4 confers strong growth at 33 °C. Immunoblot analysis indicates similar levels of expression of Ypt1 and Ypt1-SW1^Sec4 from the GPD promoter (Fig. S1A). This result is consistent with the hypothesis that Sec2 activates Ypt1-SW1^Sec4 in vivo.

Fig. 1.

(A) Expression of Ypt1-SW1^Sec4 suppresses bet3-1. Wild-type (NY834) (top three rows) or bet3-1 (SFNY380) cells (bottom four rows) were transformed with plasmids expressing either Ypt1 (yielding NY3188 and NY3190) or Ypt1-SW1^Sec4 (yielding NY3189 and NY3191) from the GPD1 promoter on a CEN vector and then grown in SC-URA medium. The parental strains and transformants were spotted in serial dilutions onto plates and incubated at the indicated temperatures for 3 d. Ypt1-SW1^Sec4, but not Ypt1, suppresses the bet3-1 growth defect at 33 °C. Inhibition of wild-type cell growth is seen upon Ypt1-SW1^Sec4 expression from this promoter at 37 °C. (B) Overexpression of Sec4 does not suppress the growth defect resulting from expression of Ypt1-SW1^Sec4. Wild-type expressing Ypt1 or Ypt1-SW1^Sec4 from the GPD1 promoter on a CEN vector (NY834, NY3188, NY3189) were transformed with a 2-μ plasmid expressing Sec4 (yielding NY3299 NY3300, NY3301). The growth test was carried out using the same conditions as in A. (C–F) Ypt1-SW1^Sec4 localizes to polarized sites and colocalizes with Sec2p in yeast cells. (C) Yeast cells expressing mCh-Ypt1 (NY3192) or mCh-Ypt1-SW1^Sec4 (NY3193) were grown to early log phase in SC-LEU medium at 25 °C. Cells were harvested and imaged by confocal fluorescence microscope. (Scale bar, 5 μm.) (D) The percentage of cells with polarized mCherry signal was quantified. Error bars represent SD from four independent experiments. At least 200 cells were scored for each strain. *P < 0.0005 Student’s t test. (E) Cells coexpressing Sec2-GFP and mCh-Ypt1 (NY3194) or mCh-Ypt1-SW1^Sec4 (NY3195) were grown to early log phase in synthetic medium at 25 °C. Next, 500 µL of cells were collected and images were captured using a confocal fluorescence microscope. (Scale bars, 5 µm.) (F) Quantification of colocalization of Sec2-GFP and mCh-Ypt1 or mCh-Ypt1-SW1^Sec4.

Fig. S1.

Examination of protein expression levels. All strains were grown to early log phase and an equal amount of cells were collected. Cells were lysed by the alkali lysis method. (A) The expression level of Ypt1 or Ypt1-SW1^Sec4 in the strains used for the growth test in Fig. 1A was evaluated using anti-Ypt1 antibody. Strains NY834, NY3188, NY3189, SFNY380, NY3190, and NY3191 correspond to lanes 1 to 6 (from left to right). Lanes 6 and 7 represent two independent transformants. Adh1 was detected as a loading control. (B) Comparison of the expression level of Ypt1 or Ypt1-SW1^Sec4 under control of the GAL or GPD promoters. Strains NY3196 and NY3197 containing GAL inducible plasmids were grown in YP +2% galactose for 5 h. Cells containing GPD expression plasmids are the same as in A. Control strains are wild-type. Equal amounts of cells were collected for all strains and lysates were prepared by the alkali lysis method. Cell lysates were then diluted as indicated to achieve a similar signal on Western blots. (C) Expression of GAL-Ypt1SW1^Sec4 had no effect on the expression level of Sec15 or Sec4. Strains were induced for 5 h as in B and lysed using the same method. Sec15 and Sec4 were examined using anti-Sec15 and anti-Sec4 antibodies. Adh1 was used as a loading control.

The localization of a GEF plays a major role in determining the localization of its substrate Rab (3, 5). Sec2 is associated primarily with secretory vesicles (12). These vesicles are delivered to sites of polarized cell surface growth, such as the tips of small buds or the necks of large budded cells where, after a brief delay, they fuse with the plasma membrane (32, 33). Sec2, as well as its substrate Rab, Sec4, are therefore highly concentrated at these polarized growth sites (12). If Sec2 activates Ypt1-SW1^Sec4 in vivo, we would expect the localization of Ypt1-SW1^Sec4 to at least partially overlap with that of Sec2. We expressed mCherry-tagged Ypt1 (mCh-Ypt1) or Ypt1-SW1^Sec4 (mCh-Ypt1-SW1^Sec4) in wild-type cells. mCh-Ypt1 is associated with Golgi cisternae, which appear as dispersed puncta with less than 4% of the cells exhibiting a concentration at sites of polarized surface growth (Fig. 1 C and D). In contrast, about 80% of the cells expressing mCh-Ypt1-SW1^Sec4 exhibit a prominent concentration at either the tip of a small bud or the neck of a large-budded cell, mimicking the distribution of Sec2 or its substrate, Sec4 (Fig. 1 C and D). Partial colocalization with Sec2-GFP was observed, further supporting the hypothesis that Sec2 activates Ypt1-SW1^Sec4 in vivo (Fig. 1 E and F). The ability of Ypt1-SW1^Sec4 to suppress bet3-1 more efficiently than Ypt1 (Fig. 1A) likely reflects the existence of a significant soluble pool of Sec2 (12) that could lead to the activation of a fraction of Ypt1-SW1^Sec4 on the early secretory pathway, where it could fulfill the essential roles of Ypt1.

In a prior study we found that Ypt1-SW1^Sec4 expressed from the endogenous YPT1 promoter can function as the sole copy of YPT1 with no obvious growth defect (27). However, the mislocalization of mCh-Ypt1-SW1^Sec4 led us to examine the effects of varying levels of Ypt1-SW1^Sec4 expression on growth and membrane traffic. When we introduced a plasmid expressing Ypt1-SW1^Sec4 from the GPD promoter into wild-type cells, the transformants grew well at intermediate temperatures, but exhibited slow growth at 37 °C (Fig. 1B). We therefore constructed a plasmid expressing Ypt1-SW1^Sec4 from the strong, inducible GAL1 promoter. Transformants grew well on medium containing glucose, a repressing carbon source, but failed to grow on inducing galactose medium (Fig. 2A). Control strains expressing Ypt1 under GAL regulation grew well on both media (Fig. 2A). Therefore, YPT1-SW1^Sec4 is a dose-dependent, dominant-negative allele. Western blots indicate that the GPD promoter led to approximately twofold overexpression of Ypt1-SW1^Sec4, while the GAL promoter led to ninefold overexpression (Fig. S1B).

Fig. 2.

Overexpression of Ypt1-SW1^Sec4 is toxic and inhibits Bgl2 secretion. (A) Yeast cells harboring GALpYpt1(NY3196) or GALpYpt1-SW1^Sec4 (NY3197) expression constructs were grown in YPD overnight. Next, 1.25 OD₆₀₀ units of cells from each strain were harvested for a growth test on YPD medium or washed once in sterile water and spotted onto a YPGAL plate. Plates were incubated at 25 °C for 2 d. (B) Overexpression of Ypt1-SW1^Sec4 blocks secretion of the Bgl2 glucanase. GALpYpt1 (NY3196), GALpYpt1-SW1^Sec4 (NY3197), and the control strains sec6-4 (NY17) and bgl2Δ (NY3072), were grown at 25 °C in YP +2% raffinose medium overnight to early log phase, shifted to YP +2% galactose medium for 4 h at 25 °C, and then an additional 90 min at either 25 °C or 37 °C, as indicated. Internal and external fractions were prepared as described in Materials and Methods. Ten-fold more of the internal fraction was loaded relative to the external fraction. Bgl2 was visualized by Western blotting using anti-Bgl2 rabbit polyclonal antibody and Adh1 was visualized as a loading control. (C) Bgl2 secretion is also inhibited in strains expressing Ypt1-SW1^Sec4 from the endogenous promoter, as the sole copy of YPT1. Wild-type (NY1210), YPT1 (NY1048), YPT1-SW1^Sec4 (NY1052), sec6-4 (NY17), and bgl2Δ (NY3072) strains were grown in YPD medium overnight. Each culture was divided into two aliquots and grown for an additional 90 min at 25 °C or 37 °C, as indicated. Bgl2 and Adh1 were detected as described above. (D) Quantitation of internal Bgl2 in cells of B. Error bars represent SD, n = 5. *P < 0.0228, **P < 0.0310. (E) Quantification of internal Bgl2 in cells of C. Error bars represent SD, n = 5. *P < 0.0008, **P < 0.0013.

Cells expressing Ypt1-SW1^Sec4 from the GAL1 promoter began to slow in growth rate 7 h after a shift from raffinose medium to galactose medium (Fig. S2). We therefore examined secretory pathway function after 5 h of induction. The cell wall glucanase Bgl2 is exported efficiently by wild-type cells with only a minor internal pool of transport intermediates (Fig. 2 B and D). A temperature-sensitive sec6-4 strain, used as a control, accumulated a large internal pool and a reduced external pool of Bgl2 at its restrictive temperature, 37 °C. Cells expressing Ypt1-SW1^Sec4 from the GAL promoter exhibited a large internal pool at both 25 °C and 37 °C, indicating a severe secretory defect (Fig. 2 B and D).

Fig. S2.

Yeast cells harboring GALpYpt1 (NY3196) or GALpYpt1-SW1^Sec4 (NY3197) were grown in YP +2% raffinose overnight then switched to YP +2% galactose at t = 0. OD_{600 nm} was measured at the indicated times.

To determine the site of the block along the secretory pathway, we examined cells by thin section electron microscopy (EM). Cells expressing Ypt1 from the GAL promoter exhibited a normal morphology (Fig. 3A), while most cells expressing Ypt1-SW1^Sec4 accumulated a large number of 100-nm vesicles (Fig. 3B). Some cells also contained several cup-shaped structures that likely represent aberrant forms of Golgi cisternae (34). These phenotypes are consistent with a defect late on the secretory pathway.

Fig. 3.

Thin-section EM indicates that Ypt1-SW1^Sec4 expression leads to secretory vesicle accumulation. (A) Cells expressing GALpYpt1 (NY3196) or (B) GALpYpt1-SW1^Sec4 (NY3197) were grown in YP +2% galactose for 7 h and prepared for EM, as described in Materials and Methods. Arrow shows a vesicle. Arrowhead shows a Golgi-related structure. (C–F) Cells expressing Ypt1 (NY1048) (C and E) or Ypt1-SW1^Sec4 (NY1052) (D and F) at normal levels from the endogenous promoter as the sole copy of YPT1. E and F show expanded boxed regions from C and D, respectively. Scale bars as indicated. (G) Quantitation of secretory vesicles. The number of vesicles per cell was scored from EM images. Sixty cells were analyzed for wild-type (NY1048) and 44 cells for Ypt1SW1^Sec4 expressed from the endogenous promoter (NY1052).

The exocytic v-SNARE, Snc1, is delivered in secretory vesicles to the plasma membrane, then internalized into endosomes and the trans-Golgi network, from which it enters a new round of vesicles in an ongoing cycle (35). At steady-state, GFP-Snc1 is predominantly at the plasma membrane with only a few internal puncta, representing endosomes and the trans-Golgi network (Fig. 4A). Following induction of Ypt1-SW1^Sec4, prominent GFP-Snc1 concentrations were observed at bud tips and bud necks, internal to the cell surface, indicating a slowing of the traffic cycle at a point after vesicles have been delivered to exocytic sites, but before their fusion with the plasma membrane.

Fig. 4.

Expression of Ypt1-SW1^Sec4 leads to intracellular GFP-Snc1 accumulation. (A) Wild-type (NY1210), GALpYpt1 (NY3204), and GALpYpt1-SW1^sec4 (NY3205) strains harboring a CEN plasmid expressing GFP-Snc1 (SFNB1223) were grown in SC +2% raffinose medium to early log phase, and then shifted to medium containing 2% galactose for 4 h at 25 °C. Images were captured by confocal microscope. (B) Strains expressing Ypt1(NY3198) or Ypt1-SW1^sec4 (NY3199) as the sole copy of Ypt1 from the endogenous promoter were transformed with a CEN plasmid expressing GFP-Snc1 (SFNB1223). Localization of GFP-Snc1 was examined. (Scale bars, 5 um.) (C) Quantification of cells with GFP patches from B. Error bars represent SD; n = 3, *P < 0.000012.

We also examined the secretory pathway function in cells expressing Ypt1-SW1^Sec4 at normal levels as the sole copy of Ypt1. Although these cells grow well, the internal pool of Bgl2 was increased several-fold with respect to that of wild-type cells, indicating a modest secretory defect (Fig. 2 C and E). Thin-section EM revealed the accumulation of 100-nm secretory vesicles (Fig. 3 D, F, and G). In cells that had been sectioned through the mother–bud axis, the vesicles were highly concentrated toward the tips of the bud. No accumulation of cup-shaped, Golgi-related structures was evident, unlike cells expressing Ypt1-SW1^Sec4 from the GAL promoter. GFP-Snc1 was concentrated at tips of small buds and near the necks of large budded cells expressing Ypt1-SW1^sec4 from the endogenous promoter (Fig. 4 B and C). In total, the results indicate that, even at normal levels of expression, when the growth rate is not significantly affected, Ypt1-SW1^Sec4 leads to a kinetic block of the secretory pathway at a point after secretory vesicles have been delivered to exocytic sites, but before their fusion with the plasma membrane. The Golgi-related structures that appear in some cells expressing Ypt1-SW1^Sec4 from the GAL promoter could result from the strong, protracted block in the pathway that in time leads to a back-up at an earlier stage, as has been observed in various late-blocked sec mutants (36).

In a prior study we did not observe a significant defect in the secretion of the cell wall enzyme invertase in cells expressing Ypt1-SW1^Sec4 from the YPT1 promoter (27). We have confirmed these observations, finding a small but not statistically significant reduction in the efficiency of invertase secretion (Fig. S3A). Invertase is delivered to the cell surface by vesicles of a higher buoyant density than those used for export of Bgl2, and the invertase vesicles (37) are thought to bud from endosomes rather than the Golgi (38). To determine if the apparent insensitivity of invertase secretion to Ypt1-SW1^Sec4 expression reflects a fundamental mechanistic difference, we examined the efficiency of invertase secretion in cells expressing Ypt1-SW1^Sec4 from the stronger GPD1 promoter. A somewhat larger, statistically significant defect was observed (Fig. S3B), indicating that secretion of invertase, like Bgl2, is subject to inhibition by Ypt1-SW1^Sec4.

Fig. S3.

Invertase assays were carried out as described in Materials and Methods. (A) Invertase secretion in strains expressing Ypt1 (NY3198) or Ypt1-SW1^Sec4 (NY3199) as the sole copy of Ypt1, with sec6-4 (NY17) used as a control. Cells were grown in YP medium. Error bars represent SD; n = 3, P < 0.08. (B) Invertase secretion in strains expressing GPDp-Ypt1(NY3188) or GPDp-Ypt1SW1^Sec4 (NY3189) from the GPD promotor. Wild-type strain (NY3235) harbored an empty vector as a control. Cells were grown in SC selection medium. Error bars represent SD; n = 7, *P < 0.0085.

The formation of secretory vesicles from the Golgi requires Ypt32 or its paralog Ypt31 (9, 10) and these Rabs are incorporated into the newly formed vesicles. However, as vesicles mature, Ypt31/32 is normally lost and Sec4 is recruited (33). We examined the localization of Ypt32 and Sec4 in strains expressing Ypt1-SW1^Sec4. In cells expressing Ypt1, GFP-Ypt32 exhibits only a minor accumulation at sites of polarized cell surface growth, with most of the protein localized to puncta distributed throughout the cytosol (Fig. 5 A and C). In cells expressing Ypt1-SW1^Sec4, GFP-Ypt32 exhibits very prominent concentrations in small buds and the necks of large budded cells (Fig. 5 A and C). GFP-Sec4 is also concentrated at bud tips and necks in cells expressing either Ypt1 or Ypt1-SW1^Sec4 (Fig. 5 B and D). Thus, the vesicles accumulating as a result of Ypt1-SW1^Sec4 expression appear to carry both Ypt32 and Sec4, although it is also possible that two vesicle populations accumulate at sites of polarized growth, one carrying Ypt32, the other carrying Sec4.

Fig. 5.

Localization of GFP-Ypt32p and GFP-Sec4p in cells expressing Ypt1 or Ypt1-SW1^Sec4 as the sole copy of Ypt1 from the endogenous promoter. (A) Images of GFP-Ypt32 in Ypt1 cells (NY3202) or Ypt1-SW1^Sec4 cells (NY3203). (B) Images of GFP-Sec4 in strains expressing Ypt1 (NY3200) or Ypt1-SW1^Sec4 (NY3201). Cells were grown in selection medium to log phase at 25 °C. Next, 500 µL of the culture was collected and imaged by confocal microscope. (Scale bar, 5 µm.) (C) Quantification of the percentage of cells exhibiting a polarized distribution of GFP-Ypt32 or (D) GFP-Sec4. Error bars represent SD; n = 3, *P < 0.001 for Ypt32 (C) and P < 0.005 for Sec4 (D).

Ypt1-SW1^Sec4 expression might be expected to drive the ectopic recruitment of Ypt1 effectors to secretory vesicles. We examined the localization of three different components of the secretory machinery that have each been implicated as Ypt1 effectors: Cog3, Sec7, and Uso1 (30, 39, 40). All of these components are normally associated with Golgi cisternae and therefore appear in wild-type cells as punctate structures, with little evident polarization toward bud tips or necks. In the cases of Cog3 and Sec7, there was little difference in localization between cells expressing Ypt1 or Ypt1-SW1^Sec4 (Fig. 6 A, B, and D). In the case of Uso1, however, a striking change was seen. A shift of Uso1-3xGFP to bud tips and necks was observed in a large fraction of cells expressing Ypt1-SW1^Sec4 (Fig. 6 C and D).

Fig. 6.

GALpYpt1-SW1^Sec4 expression affects a subset of Ypt1 effectors. Cells were grown in selection medium containing 2% galactose for 5 h at 25 °C. GFP fusion proteins were visualized by confocal fluorescence microscope. The overlay panels show a merge of the fluorescence image and a DIC image. (A) Yeast cells coexpressed Cog3-3xGFP (pNB1620) with GALpYpt1 (NY3206) or GALpYpt1SW1^Sec4 (NY3207). (B) Yeast cells coexpressed Sec7-EGFP (SFNB797) with GALpYpt1 (NY3208) or GALpYpt1SW1^Sec4 (NY3209). (C) Yeast cells coexpressed Uso1p-3xGFP (pNB1621) with GALpYpt1 (NY3210) or GALpYpt1SW1^Sec4 (NY3211). (Scale bar, 5 µm.) (D) Quantification of cells exhibiting a polarized distribution of the indicated proteins. Error bars represent SD; n = 3.

We have used suppressor analysis to address the mechanism of inhibition of vesicle traffic by Ypt1-SW1^Sec4. Since Sec2 can recognize and activate Ypt1-SW1^Sec4, it is plausible that Ypt1-SW1^Sec4 competes against the normal substrate, Sec4, for a limiting pool of Sec2. This would reduce the level of Sec4-GTP and therefore inhibit the secretory pathway at a late stage. Another possible mechanism of inhibition would be interference by Ypt1-SW1^Sec4 in the interaction of Sec4-GTP with one of its effectors. We overexpressed either Sec2 or Sec4, from high copy-number plasmids, in cells expressing Ypt1-SW1^Sec4 from the GAL promoter. No restoration of growth on galactose medium was observed in these strains (Fig. 7A). We also overexpressed Sec4 in cells expressing Ypt1-SW1^Sec4 from the weaker GPD promoter. The growth defect observed at 37 °C was unaffected by increased Sec4 expression (Fig. 1B). Therefore, it appears unlikely that competition against Sec4 represents the mechanism of vesicle traffic inhibition by Ypt1-SW1^Sec4. Furthermore, expression of Ypt1-SW1^Sec4 from the GAL promoter has no effect on the expression level of either Sec4 or its effector Sec15 (Fig. S1C).

Fig. 7.

The toxicity of high-level Ypt1-SW1^Sec4 expression is suppressed by truncation of Uso1, but not by overexpression of Sec2 or Sec4. (A) Strains expressing GALpYpt1 (NY3196) or GALpYpt1-SW1^Sec4 (NY3197) were transformed with a high copy-number plasmid carrying Sec2 (pNB134) generating strains NY3213 and NY3214 respectively or with a high copy-number plasmid carrying Sec4 (pNB1622), generating strains NY3215 and NY3216, respectively. Cells were grown to early stationary phase in selection medium, serially diluted, and spotted onto YPGAL or YPD plates, as indicated. Plates were incubated at 25 °C for 2 d. (B) Wild-type (NY3196, NY3197), uso1-1 (NY3217, NY3218), uso1-10 (NY3219, NY3220), sec7-1 (NY3223, NY3224), and sec22-3 (NY3225, NY3226) cells expressing Ypt1 or Ypt1-SW1^Sec4 from the GAL1 promoter were grown to early stationary phase in YPD medium at 25 °C, washed once, spotted onto YPGAL plates in fivefold serial dilutions, and then incubated at the indicated temperature for 3 d. Both uso1 truncations, but neither sec7-1 nor sec22-3, suppress the toxicity of Ypt1-SW1^Sec4. (C) Quantification of cells exhibiting an internal patch of GFP-Snc1 (Fig. S4 for representative images). Error bars represent SD; n = 3. *P < 0.00015, **P < 0.0011.

Another plausible mechanism involves the ectopic recruitment of Ypt1 effectors to the secretory vesicle membrane by Ypt1-SW1^Sec4. The inappropriate presence of one or more Ypt1 effectors on the vesicle surface might inhibit its ability to dock and fuse with the plasma membrane. Since Uso1 redistributes to exocytic sites in response to Ypt1-SW1^Sec4 expression, while Cog3 and Sec7 do not, we tested the ability of two loss-of-function alleles of uso1 to suppress the galactose-dependent toxicity of GALpYpt1-SW1^Sec4 expression. The uso1-1 mutation truncates the 1,790-residue-long protein at amino acid 950. This results in temperature-sensitive defects in growth and secretion (41). However, at permissive temperatures the uso1-1 strain was significantly resistant to the galactose-dependent toxicity of GALp Ypt1-SW1^Sec4 expression (Fig. 7B). The uso1-10 allele truncates the protein at amino acid 1,123. This less-severe truncation does not cause growth defects; however, it too suppressed the galactose-dependent toxicity of GALp Ypt1-SW1^Sec4 expression (Fig. 7B). As controls, we examined the effects of the sec7-1 and sec22-3 mutations on the galactose-dependent toxicity of GALp Ypt1-SW1^Sec4 expression. Sec7 acts in the formation of secretory vesicles from the Golgi, while Sec22 is needed, like Uso1, for transport from the ER to the Golgi. The sec7-1 mutation conferred only slight suppression of the galactose sensitivity, at 30 °C or 34 °C, while no suppression was seen by sec22-3 at any temperature (Fig. 7B).

We next examined the effects of uso1-10, sec7-1, and sec22-3 on GFP-Snc1 localization in cells expressing Ypt1-SW1^Sec4 from the GAL promoter. Substantial suppression of the polarized GFP-Snc1 localization phenotype was observed in uso1-1 and uso1-10 cells. This was evident as a significant decrease in the number of cells exhibiting a polarized patch of GFP-Snc1 and a reappearance of GFP-Snc1 at the plasma membrane in some cells (Fig. 7C and Fig. S4). No suppression was observed in sec7-1 or sec22-3 cells (Fig. 7C and Fig. S4). We also confirmed partial suppression by uso1-10 of the Bgl2 secretion defect of cells expressing Ypt1-SW1^Sec4 from the GAL promoter (Fig. S5). The ability of uso1 truncations to suppress the toxicity and transport defects resulting from GALp Ypt1-SW1^Sec4 expression suggests that the recruitment of full-length Uso1 to the surface of secretory vesicles by Ypt1-SW1^Sec4 blocks their ability to dock and fuse with the plasma membrane.

Fig. S4.

WT (NY3204 or NY3205), uso1-1 (NY3227 or NY3228), uso1-10 (NY3229 or NY3230), sec7-1 (NY3231or NY3232), and sec22-3 (NY3233 or NY3234) cells expressing either Ypt1 or Ypt1SW1^Sec4 from the GAL promoter were transformed with a plasmid expressing GFP-Snc1. Arrows point to restored plasma membrane localization in uso1-1 and uso1-10 strains. Cells were grown at 25 °C and then imaged by fluorescence microscopy. Quantification is shown in Fig. 7C. (Scale bars, 5 μm.)

Fig. S5.

Inhibition of Bgl2 secretion by Ypt1-SW1^Sec4 is partially suppressed in uso1-10 cells. Assays were carried out as described in Materials and Methods. Wild-type cells expressing GALpYpt1(NY3196) or GALpYpt1-SW1^Sec4 (NY3197) and uso1-10 mutant cells expressing GALpYpt1(NY3219) or GALpYpt1-SW1^Sec4 (NY3220) were collected after 3 h of galactose induction. Two independent transformants were tested. Bgl2 was detected using anti-Bgl2 antibody.

Uso1 is the yeast homolog of the mammalian p115 protein. Like p115, Uso1 has an amino-terminal globular head followed by a long, homo-dimeric, coiled-coil tail (42) (Fig. 8A). The head of p115 has been shown to bind Rab1 (43) and we find that the homologous region of Uso1 (amino acids 1–246) binds to Ypt1 (Fig. 8 B and C). The coiled-coil region of p115 includes a section that is SNARE-like in sequence. We found that the first half of the coiled-coil region of Uso1 (amino acids 726–1,246) binds to the syntaxin-related SNARE, Sed5 (Fig. 8D), but not to the other SNAREs involved in transport from the ER to the Golgi (Fig. S6 B–D). The globular head showed no affinity for Sed5 (Fig. S6A). The loss of the SNARE-like domain in uso1-1 correlates with the acquisition of temperature-sensitive growth defects. The uso1-10 allele includes both the globular head domain that binds to Ypt1, as well as the SNARE-like region implicated in binding to Sed5 and this allele confers normal growth. Because uso1-10 suppresses the galactose-dependent toxicity of GALp Ypt1-SW1^Sec4 yet confers no growth defect on its own, and the Uso1-10 protein still has the capacity to bind both of its ligands, Ypt1 and Sed5 in vitro, it is unlikely that suppression is the result of an upstream inhibition of the secretory pathway.

Fig. 8.

Uso1 interacts with Ypt1 and Sed5. (A) A schematic diagram of the Uso1 protein. AD is an acidic domain; CC1–CC4 are coiled-coil regions; H1 and H2 are regions of homology between Uso1 and its mammalian homolog p115. A SNARE-like domain is highlighted in yellow within the CC2 region. The uso1-1 mutation truncates the protein at aa 950, while uso1-10 truncates the protein at amino acid 1,123. Hinge regions I and II were identified by Nakajima et al. (41). (B and C) The globular head domain of Uso1 interacts with Ypt1. GST-tagged Uso1 domains were immobilized on glutathione beads and then incubated with increasing amounts of purified His₆-Ypt1 proteins. GST and GST-Sec13 were used as controls. After washing, the bound His₆-Ypt1 was detected. (D) The CC1 and CC2 coiled-coil region of Uso1 (amino acids 726–1,246) binds to Sed5. GST-Sed5 was immobilized on glutathione beads and then incubated with increasing amounts of the His₆-Uso1 coiled-coil region (amino acids 726–1,246). After washing, the bound His₆-Uso1 was detected as described in Materials and Methods.

Fig. S6.

(A) The amino terminal region (amino acids 1–860) of Uso1 does not bind to Sed5. The CC1 and CC2 coiled-coil regions of Uso1 (amino acids 726–1,246) do not bind to the SNAREs, Bos1 (B), Sec22 (C), or Bet1 (D). GST-tagged SNAREs were immobilized on glutathione beads and then incubated with the His₆-Uso1 amino terminal region (amino acids 1–860) (A) or the coiled-coil region (amino acids 726–1,246) (B–D). All in vitro binding experiments were performed in binding buffer for 4 h at 4 °C. Bound proteins were detected as described in Materials and Methods.

Discussion

Many Rabs are linked to adjacent Rabs through their GAPs and GEFs, forming regulatory circuits that are thought to trigger Rab transitions and thereby promote changes in the functional identity of the membrane as it flows along a transport pathway (1). We have probed the effect of rewiring one such regulatory circuit on the yeast secretory pathway. Early analysis of the Ypt1-SW1^Sec4 chimera had indicated apparently normal Ypt1 function (27), yet the subsequent demonstration that Ypt1-SW1^Sec4 could be activated efficiently by the Sec4 exchange factor, Sec2 (29), caused us to reexamine this situation. We considered the possibility that Sec2 might not actually activate Ypt1-SW1^Sec4 in vivo as it does in vitro (29), however the suppression of a bet3-1 mutation by Ypt1-SW1^Sec4 expression as well as the colocalization of Ypt1-SW1^Sec4 with Sec2 at sites of polarized exocytosis argues against this possibility.

Ectopic activation of Ypt1-SW1^Sec4 on secretory vesicles by Sec2 would be expected to lead to a Rab “identity crisis,” in apparent conflict with our earlier study showing little phenotypic effect (27). We now report that Ypt1-SW1^Sec4 expression does indeed lead to trafficking defects, the severity of which correlates with the level of expression. Even normal expression of Ypt1-SW1^Sec4, which does not slow growth, causes a measurable inhibition of the final stage of the secretory pathway, after the Golgi-derived vesicles have been concentrated at polarized sites of exocytosis but before their fusion with the plasma membrane. Higher levels of expression potently block both secretion and cell growth. These results confirm the importance of Rabs to the identity of the membranes with which they are associated.

A key question is just how extensive is the reorganization of the secretory machinery in response to Ypt1-SW1^Sec4 expression. A literal interpretation of the Rab cascade model might predict that the entire Golgi would relocalize to polarized exocytic sites in response to Ypt1-SW1^Sec4 expression. The middle Rab on the pathway, Ypt32, did shift toward exocytic sites in agreement with the cascade model; however, of the three known Ypt1 effectors that we tested, only Uso1 exhibited a clear shift toward exocytic sites. While it is possible that the failure of the other two, Cog3 and Sec7, to relocalize could reflect a reduced affinity toward Ypt1-SW1^Sec4, the ability of YPT1-SW1^Sec4 to replace YPT1 argues against this interpretation. It appears more likely that Cog3 and Sec7 rely on additional interactions for their localization.

The observation that truncations of uso1 suppress the toxicity and transport defects associated with Ypt1-SW1^Sec4 expression points to a probable mechanism of vesicular traffic inhibition. Uso1 forms a 155-nm-long homo-dimeric coiled-coil rod with a globular head at the amino terminal end (42). We show that the globular head binds to Ypt1 and the tail binds to the syntaxin-related Golgi tSNARE, Sed5. While truncation of the Uso1 rod before the Sed5 binding site (uso1-1) impairs ER–Golgi traffic (41), a less-severe truncation (uso1-10) does not; nonetheless, both truncations suppress Ypt1-SW1^Sec4 toxicity. Furthermore, neither the sec22-3 mutation, which inhibits ER–Golgi transport, nor the sec7-1 mutation that blocks export from the Golgi, suppresses the toxicity. Thus, it is not the slowing of ER–Golgi traffic that leads to suppression. We propose that when Ypt1-SW1^Sec4 recruits full-length Uso1 by binding to its globular head, the long coiled-coil tail protrudes outward from the surface of secretory vesicles and sterically hinders their ability to dock and fuse with the plasma membrane. The truncations shorten the rod so that docking can proceed.

This model raises the interesting question of why Uso1 doesn’t normally block the docking of ER-derived vesicles with the Golgi membrane by the same mechanism. The coiled-coil rod of Uso1 might bend to allow a closer approach of the vesicle to the Golgi. Rotary shadow EM revealed two hinge regions that correspond to breaks in the coiled-coil motif between residues 800–900 and 1,200–1,400 (42). This would be analogous to the related coiled-coil tethers EEA1 and the Golgin GCC185. In the case of EEA1, the rod collapses in response to binding Rab5-GTP (44), while in the case of GCC185, flexibility of a hinge region is essential for its function in endosome-Golgi traffic (45). It will be interesting to investigate if the conformation of Uso1 is regulated through its interaction with either Ypt1 or Sed5. Regulated bending could ensure that Uso1 not only promotes docking of vesicles to the correct target, but physically blocks docking to the wrong target. This would serve to increase the fidelity of vesicular traffic.

Materials and Methods

Yeast and Plasmid Strains Construction.

The yeast strains and plasmids used in this study are listed in Tables S1–S3. To generate the various Ypt1-SW1^Sec4 expression plasmids, full-length Ypt1-SW1^Sec4 was PCR-amplified from NRB483 and inserted into the p416-GDP vector to generate (NRB1612) or into an integrative pRS306 based vector to generate NRB1615 at XbaI and XhoI sites. The plasmid was linearized with StuI and integrated into the yeast genome at the URA3 locus. To generate GAL-inducible plasmids (NRB1617, NRB1619), the Ypt1-SW1^Sec4 PCR fragment was inserted behind the GAL1-10 promoter in pNB1383 or pNB1384 at BamHI/XbaI sites [NRB1383 pRS305 with Cyc1 term (SacII/SacI) + Adh term (ApaI/XhoI) + GAL1-10 promoter (SmaI/BamHI), and NRB1384 pRS306 with Cyc1 terminator (SacII/SacI) + Adh terminator (ApaI/XhoI) + GAL1-10 promoter (SmaI/BamHI)]. The plasmid expressing mCh-Ypt1-SW1^Sec4 (NRB1613) was generated by replacing the Ypt1 gene in pNR1326 with the Ypt1-SW1^Sec4 PCR fragment. This plasmid was linearized with AflII and integrated into the yeast genome at the LEU2 locus. In each case, Ypt1 expression plasmids were generated in parallel to use as controls (NRB1611, NRB1614, NRB1616, NRB1618). Fusions of Uso1 or Cog3 protein to 3xGFP tags were constructed by PCR amplification of the C-terminal 1,000-bp fragment of USO1 (NRB1621) or COG3 (NRB1620) from wild-type yeast genomic DNA with SalI and BamHI sites, and cloned into the plasmid pPG5-3xGFP (NRB1302) (46) followed by linearization with ClaI or SpeI, accordingly, and transformed into yeast cells. All plasmids constructed in this study were verified by DNA sequencing. C-terminal truncations of Uso1 were constructed in SFNY 1841 as described in Longtine et al. (47). Truncations were confirmed by PCR analysis.

Table S1.

Yeast strains

NY no.	Genetype	Source
NY17	MATa, ura3-52, sec6-4	P.N. collection
NY834	MATα, ura3-52 his4-619 leu2-3,112	P.N. collection
NY1048	MATα, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1::LEU2]	P.N. collection
NY1052	MATa, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1-SW1sec4:: LEU2]	P.N. collection
NY1210	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200	P.N. collection
NY3072	MATa, his3∆1 leu2∆0 met15∆0 ura3∆0 [bgl2∆::KanMX4]	P.N. collection
NY3188	MATα, ura3-52 his4-619 leu2-3,112 [p416-GPDp-YPT1::URA3]	Present study
NY3189	MATα, ura3-52 his4-619 leu2-3,112 [p416-GPDp-YPT1-SW1sec4::URA3]	Present study
NY3190	MATα, ura3-52 his4-619 leu2-3,112 bet3-1 [p416-GPDp-YPT1::URA3]	Present study
NY3191	MATα, ura3-52 his4-619 leu2-3,112 bet3-1 [p416-GPDp-YPT1-SW1sec4::URA3]	Present study
NY3192	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 [ADHp-mCherry-YPT1::LEU2]	Present study
NY3193	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 [ADHp-mCherry-YPT1-SW1sec4::LEU2]	Present study
NY3194	MATa, GAL+, ura3-52 his3-Δ200 s2-1Δ:: HIS3 [SEC2-GFP::LEU2] [GPDp-mCherry-YPT1::URA3]	Present study
NY3195	MATa, GAL+, ura3-52 his3-Δ200 s2-1Δ::HIS3 [SEC2-GFP::LEU2] [GPD-mCherry-YPT1-SW1sec4::Ura3]	Present study
NY3196	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GAL1-10p-YPT1::URA3]	Present study
NY3197	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GAL1-10p-YPT1-SW1sec4::URA3]	Present study
NY3198	MATα, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1::LEU2] [GFP1-SNC1::URA3, CEN]	Present study
NY3199	MATa, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1-SW1sec4::LEU2] [GFP-SNC1::URA3, CEN]	Present study
NY3200	MATα, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1::LEU2] [GFP-SEC4::URA3]	Present study
NY3201	MATa, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1-SW1sec4::LEU2] [GFP-SEC4::URA3]	Present study
NY3202	MATα, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1::LEU2] [GFP-YPT32:: URA3]	Present study
NY3203	MATa, his3-Δ200, ura3-52 ypt1Δ::HIS3 [YPT1-SW1sec4::LEU2] [GFP-YPT32::URA3]	Present study
NY3204	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GAL1-10p-YPT1::LEU2] [GFP-SNC1::URA3,CEN6]	Present study
NY3205	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GAL1-10p-YPT1-SW1sec4::LEU2] [GFP-SNC1::URA3,CEN6]	Present study
NY3206	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [COG3-3xGFP::URA3] [GAL1-10p-YPT1::LEU2]	Present study
NY3207	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [COG3-3xGFP::URA3] [GALp-YPT1-SW1sec4::LEU2]	Present study
NY3208	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [SEC7-EGFP::URA3] [GALp-YPT1::LEU2]	Present study
NY3209	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [SEC7-EGFP::URA3] GALp-YPT1-SW1sec4::LEU2]	Present study
NY3210	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [USO1-3xGFP::URA3] [GAL1-10p-YPT1::LEU2]	Present study
NY3211	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [USO1-3xGFP::URA3] [GAL1-10p-YPT1-SW1sec4::LEU2]	Present study
NY3213	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GALp-YPT1::LUE2] [Sec2::URA3 2μ]	Present study
NY3214	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GAL-YPT1- SW1sec4::LUE2][SEC2::URA3 2μ]	Present study
NY3215	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GAL-YPT1::URA3] [SEC4::LUE2 2μ]	Present study
NY3216	MATa, GAL+, ura3-52 leu2-3,112 his3-Δ200 [GAL-YPT1-SW1sec4:: URA3] [SEC4::LEU2, 2μ]	Present study
NY3217	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-1::KanMX6 [aa 1–950] [GAL1-10p-YPT1::URA3]	Present study
NY3218	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-1::KanMX6 [aa 1–950] [GAL1-10p-YPT1-SW1sec4::URA3]	Present study
NY3219	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-10::KanMX6 [aa 1–1,123] [GAL1-10p-YPT1::URA3]	Present study
NY3220	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-10::KanMX6 [aa 1–1,123] [GALp-YPT1-SW1sec4::URA3]	Present study
NY3221	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s7-1	Present study
NY3222	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s22-3	Present study
NY3223	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s7-1 [GAL1-10p-YPT1::URA3]	Present study
NY3224	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s7-1 [GAL1-10p-YPT1-SW1sec4::URA3]	Present study
NY3225	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s22-3 [GAL1-10p-YPT1::URA3]	Present study
NY3226	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s22-3 [GAL1-10p-YPT1-SW1sec4::URA3]	Present study
NY3227	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-1::KanMX6 [aa 1–950)] [GAL1-10p-YPT1::LEU2][GFP-SNC1::URA3 CEN6]	Present study
NY3228	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-1::KanMX6 (aa 1–950)] [GAL1-10p-YPT1-SW1sec4::LEU2][GFP-SNC1::URA3 CEN6]	Present study
NY3229	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-10::KanMX6 [aa 1–1,123)] [GAL1-10p-YPT1::LEU2][GFP-SNC1::URA3 CEN6]	Present study
NY3230	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-10::KanMX6 [aa 1–1,123][GAL1-10p-YPT1-SW1sec4::LEU2][GFP-SNC1::URA3 CEN6]	Present study
NY3231	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s7-1 [GAL1-10p-YPT1::URA3] [GFP-SNC1::URA3 CEN6]	Present study
NY3232	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s7-1 [GAL1-10p-YPT1-SW1sec4::URA3] [GFP-SNC1::URA3 CEN6]	Present study
NY3233	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s22-3 [GAL1-10p-YPT1::URA3] [GFP-SNC1::URA3 CEN6]	Present study
NY3234	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200 s22-3 [GAL1-10p-YPT1-SW1sec4::URA3] [GFP-SNC1::URA3 CEN6]	Present study
NY3235	MATα, ura3-52 his4-619 leu2-3,112 [p416-GPDp-CYC1term]	Present study
NY3299	MATα, ura3-52 his4-619 leu2-3,112 ] [pRS425-Sec4::LEU2]	Present study
NY3300	MATα, ura3-52 his4-619 leu2-3,112 [p416-GPDp-YPT1::URA3] [pRS425-Sec4::LEU2]	Present study
NY3301	MATα, ura3-52 his4-619 leu2-3,112 [p416-GPDp-YPT1-SW1sec4::URA3] [pRS425-Sec4::LEU2]	Present study
SFNY380	MATα, ura3-52 his4-619 leu2-3,112 bet3-1	S.F.-N. collection
SFNY1841	MATa, GAL+, ura3-52 leu2-3 112 his3-Δ200	S.F.-N. collection
SFNY2447	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-1::KanMX6 (aa 1–950)	S.F.-N. collection
SFNY2448	MATa, GAL+, ura3-52 leu2-3,112 his3∆200 uso1-10::KanMX6 (aa 1–1123)	S.F.-N. collection

Table S3.

Plasmids for bacteria expression

Strain no.	Plasmid description	Source
SFNB1033	pGEX2T-GST-BOS1	S.F.-N. collection
SFNB1042	pGEX2T- GST-Sed5	S.F.-N. collection
SFNB1043	pGEX2T- GST-Bet1	S.F.-N. collection
SFNB2099	pGEX6P1-GST-Uso1(amino acids 1–246)	S.F.-N. collection
SFNB2100	pGEX6P1-GST-Uso1(amino acids 1–724)	S.F.-N. collection
SFNB2131	pGEX6P1-GST-Uso1 (amino acids 726–1,246)	S.F.-N. collection
SFNB2132	pET15b-His-Uso1 (amino acids 726–1,246)	S.F.-N. collection
SFNB2140	pET15b- His-Uso1 (amino acids 1–860)	S.F.-N. collection

Table S2.

Plasmids for yeast transformation

NRB no.	Plasmid description	Source
NRB134	SEC2 [SEC2 2 μ URA. Sec2 ClaI-BamHI fragment (4.5 kb)]	P.N. collection
NRB1302	pPG5-3xGFP	P.N. collection
NRB1312	pRS306-GFP-SEC4- CYC1term (Sph I digestion to integrate into URA3 locus)	P.N. collection
NRB1325	pRS306-ADH1p-GFP-YPT32-CYC1term (StuI digestion to integrate into URA3 locus)	P.N. collection
NRB 1326	YIplac128-ADHp-mCherry-YPT1- CYC1term (AflII digestion to integrate into LEU2 locus)	P.N. collection
NRB1383	pRS305 with Cyc1 term (SacII/SacI) + Adh term (ApaI/XhoI) + GAL1-10 prom (SmaI/BamHI)	P.N. collection
NRB1384	pRS306 with Cyc1 term (SacII/SacI) + Adh term (ApaI/XhoI) + GAL1-10 prom (SmaI/BamHI)	P.N. collection
NRB1611	p416GPDp-YPT1-CYC1term, CEN6	Present study
NRB1612	p416GPDp-YPT1-SW1sec4-CYC1term, CEN6	Present study
NRB1613	YIplac128-ADHp-mCherry-YPT1-SW1sec4- CYC1term (AflII digestion to integrate into LEU2 locus)	Present study
NRB1614	pRS306-GPDp-mCherry-YPT1-CYC1tern	Present study
NRB1615	pRS306 GPDp-mCherry-YPT1-SW1sec4-CYC1tern	Present study
NRB1616	pRS305-GALp-YPT1-CYC1term (AflII digestion to integrate into LEU2 locus)	Present study
NRB1617	pRS305-GALp- YPT1-SW1sec4-CYC1term (AflII digestion to integrate into LEU2 locus)	Present study
NRB1618	pRS306-GALp-YPT1-CYC1term (StuI digestion to integrate into URA3 locus)	Present study
NRB1619	pRS306-GALp- YPT1-SW1sec4-CYC1term (StuI digestion to integrate into URA3 locus)	Present study
NRB1620	HEMIZAP pRS306-COG3-3xGFP-CYC1term (SpeI digestion to integrate into COG3 locus)	Present study
NRB1621	HEMIZAP pRS306-USO1-3XGFP- CYC1term (ClaI digestion to integrate into USO1 locus)	Present study
NRB1622	pRS425-SEC4 (2 μ LEU2)	Present study
SFNB797	pUSE-SEC7-EGFP-URA (SpeI digestion to integrate into SEC7 locus)	S.F.-N. collection
SFNB1223	pRS416-TPIp-GFP-SNC1,CEN6	S.F.-N. collection

Growth Tests.

Yeast cells were grown in yeast extract peptone dextrose (YPD) medium to stationary phase. Cells were washed once with sterile water and spotted on YPD plates in fivefold serial dilutions starting with an OD₆₀₀ of 5. In the case of yeast strains harboring a CEN or 2-μm circle-based plasmid, cells were spotted on synthetic complete (SC) selection plates. Plates were left at the specified temperature for the indicated time. For GAL induction experiments, yeast cells were grown in YP medium containing 2% raffinose overnight and then spotted on YPGAL plates.

Fluorescence Microscopy and Quantitative Localization Analysis.

Yeast strains harboring a GFP- or mCherry-tag were grown at 25 °C to early log phase (OD₆₀₀ 0.4–0.6) in selective SC medium. Next, 500 μL of cells were pelleted and resuspended in growth medium. Fluorescence imaging was performed as described previously (48). In brief, images were acquired with a 100× oil-immersion objective lens (α Plan Apochromat 100×/1.46 oil DIC lens; Carl Zeiss) on a spinning disk confocal microscopy system (Yokogawa Electric Corporation), connected to a microscope (Observer Z1; Carl Zeiss) equipped with an electron multiplying charge-coupled device (CCD) camera (Cascasde II; Photometrics). Excitation of GFP or mCherry was achieved using 488-nm argon and 568-nm argon/krypton lasers, respectively. For each sample, z-stacks with a 100- or 200-nm slice distance were generated. Images were analyzed using AxioVision software 4.8 (Carl Zeiss). For quantitation studies, three independent transformants were examined and at least 100 cells for each condition were scored. Three separate experiments were used to calculate the SD.

Invertase Secretion Assay.

The invertase secretion assay was carried out as previously described by Shen et al. (48). Briefly, yeast cells were grown at 25 °C in YP or SC selection medium containing 5% glucose overnight to early log phase (0.3–0.5 OD/mL). Next, 1 OD₆₀₀ unit of cells were transferred to two sets of 15-mL tubes, and pelleted. One set of cells were washed once with 0.5 mL sterile water and resuspended in 1 mL 10 mM NaN3, and kept on ice. This was for the 0-h samples. Another set of cells were washed once with sterile water, resuspended in 1 mL of YP or SC selection medium containing 0.1% glucose, and incubated at 37 °C with shaking for 1 h. Cells were pelleted and resuspended in 1 mL of 10 mM NaN₃. This set of cells was used for 1-h samples. For each 0- and 1-h sample, the external invertase was measured directly. The internal invertase was measured after 0.5 mL of cells were treated with 50 μg/mL zymolyase at 37 °C for 45 min to generate spheroplasts and lysed with 0.5 mL of 0.5% Triton X-100. The percentage of invertase secretion was calculated by [Ext (1 h) − Ext (0 h)]/{[Ext (1 h) − Ext (0 h)] + [Int (1 h) − Int (0 h)]} (18).

Bgl2 Secretion Assay.

The Bgl2 secretion assay was performed as previously described (49). In brief, 50 mL of yeast cells were grown at 25 °C in YPD medium overnight to early log phase (∼0.3 OD₆₀₀/mL). About 5 OD_{600 nm} units of cells were transferred to a sterile flask. Two sets of flasks were prepared for each strain. One set was incubated at 25 °C and the other set was at 37 °C. After 90-min incubation with shaking, both sets of cell cultures were harvested by centrifugation at 900 × g for 5 min. Cell pellets were resuspended in 1 mL of ice-cold 10 mM NaN₃ and 10 mM NaF, and then incubated on ice for 10-min. The suspension was transferred to microfuge tubes, pelleted, and resuspended in 1 mL of fresh prespheroplasting buffer (100 mM Tris⋅HCl, pH 9.4, 50 mM β-mercaptoethanol, 10 mM NaN₃, and 10 mM NaF), and incubated on ice 15-min. Cells were then pelleted and washed with 0.5 mL of spheroplast buffer (50 mM KH₂PO₄-KOH, pH 7.0, 1.4 M sorbitol, and 10 mM NaN₃), and resuspended in 1 mL of spheroplast buffer containing 167 µg/mL zymolyase 100T (Nacasai Tesque). Cells were incubated in a 37 °C water bath for 30 min. Spheroplasts were spun down at 5,000 × g for 10 min, and 100 µL of the supernatant from each tube was transferred into a new tube and mixed with 34 µL of 4× SDS sample buffer (the external pool). All of the remaining supernatant was discarded and the pellet (spheroplast) was resuspended in 100 µL 1× SDS sample buffer (the internal pool). Samples were loaded on a 10% SDS/PAGE gel. Bgl2 was visualized by Western blotting with anti-Bgl2 rabbit polyclonal antibody at 1:5,000 dilution (provided by the laboratory of Randy Schekman, University of California, Berkeley). As a loading control, Adh1 was visualized by Western blotting with anti-Adh1 rabbit polyclonal antibody (AB1202; EMD Millipore) at 1:10,000 dilution. For quantitation of Bgl2, samples used for immunoblotting were diluted accordingly to achieve approximately similar level of Bgl2 protein and loaded onto the same SDS/PAGE gel. Serial dilutions of a control sample were run in parallel to establish a standard curve. The electrophoretic bands were quantitated using ImageJ software (https://imagej.nih.gov).

Electron Microscopy.

Yeast cells expressing Ypt1 (NY1048) or Ypt1SW1^Sec4 (NY1052) at the normal level were grown at 25 °C in YPD to an OD_{600 nm} of ∼0.5 and then processed for EM as previously described (50). In brief, ∼10 OD_{600 nm} units of cells were collected using a 0.22-µm filter apparatus, washed with 10 mL of 0.1 M cacodylate (pH 6.8), then resuspended in 10 mL of fixative (0.1 M cacodylate, 4% glutaraldehyde, pH 6.8). Cells were fixed at room temperature for 1 h and then moved to 4 °C for 16 h. The next day, cells were washed twice with 50 mM KPi (pH 7.5), and then resuspended in 2 mL of 50 mM KPi buffer containing 0.25 mg/mL Zymolyase 100T. Cells were then incubated for 40 min at 37 °C in a water bath with gentle shaking. After the incubation, the cells were washed twice with ice-cold 0.1 M cacodylate buffer using a 0.22-µm filter apparatus, and resuspended in 1.5 mL of cold 2% OsO₄ in 0.1 M cacodylate buffer. Cells were then incubated for 1 h on ice, washed three times with water, and then incubated in 1.5 mL of 2% uranyl acetate at room temperature for 1 h. Cells were dehydrated by a series of ethanol washes and incubated overnight in Spurr resin. Cells were embedded in fresh Spurr resin and baked at 80 °C for at least 24 h. Sections were stained with lead citrate and uranyl acetate, and images were acquired using a transmission electron microscope (Tecnai G2 Spirit; FEI) equipped with a CCD camera (UltraScan 4000; Gatan). For yeast cells harboring a GAL10 expression plasmid (NY3196, NY3197), cells were grown in a YP medium containing 2% raffinose overnight and then switched to a medium containing 2% galactose for 7 h at 25 °C and then processed for EM.

Immobilization of GST Fusion Proteins.

For purification of GST fusion proteins, BL21(DE3) cells were incubated at 18 °C overnight with 0.5 mM isopropyl β-d-1-thiogalactopyranoside to induce protein expression. Cells were collected and resuspended in 1× PBS with 1 mM DTT and protease inhibitors. Cells were sonicated for 2 min total with 15-s on/off bursts on ice. Triton X-100 was added to a final concentration of 1% and lysates were incubated on ice for 15 min. Lysates were cleared through a 15-min centrifugation at 27,000 × g. The supernatant was incubated with 1 mL of 50% glutathione Sepharose beads (GE Healthcare) that had been prewashed with PBS for 1 h at 4 °C with rotation. The beads were washed extensively with PBS and stored at 4 °C.

Purification of His₆ Fusion Proteins.

For purification of His₆-tagged fusion proteins, cells were incubated overnight at 18 °C with 0.5 mM isopropyl β-d-1-thiogalactopyranoside to induce protein expression. Cells were collected and resuspended in sonication buffer (20 mM Hepes pH 7.4, 150 mM NaCl, 1 mM DTT, 15 mM imidazole and protease inhibitors). Cells were sonicated for 2 min total with 15-s on/off bursts on ice. Lysates were cleared through a 15-min centrifugation at 27,000 × g. The supernatant was incubated with 2 mL of 50% Ni-NTA resin (Qiagen) for 1 h at 4 °C with rotation. The Ni-NTA resin was prewashed with sonication buffer. Following binding the resin was washed extensively with sonication buffer. Fusion proteins were eluted off the resin with sonication buffer with 250 mM Imidazole. Proteins were buffer exchanged with 20 mM Hepes (pH 7.4), 150 mM NaCl. Glycerol was added to a final concentration of 30% and proteins were stored at −20 °C.

In Vitro Bindings with Recombinant Proteins.

Equimolar amounts (0.2 µM) of immobilized GST fusions were incubated with increasing amounts of bacterial purified His₆-Uso1 in binding buffer (25 mM Hepes pH 7.4, 150 mM NaCl, 2% Triton X-100, 1 mM MgCl_2, 1 mM EDTA, 1 mM DTT, protease inhibitors) for 4 h at 4 °C with rotation. Beads were washed three times with binding buffer and eluted in 25 µL of sample buffer by heating for 5 min at 100 °C. Binding studies with bacterially purified Ypt1-His₆ were performed as described above except with binding buffer II (25 mM Hepes pH 7.4, 150 mM NaCl, 0.5% Triton X-100, 1 mM MgCl_2, 1 mM EDTA, 1 mM DTT, protease inhibitors).

Bound proteins were resolved on SDS/PAGE gels and detected with mouse monoclonal anti-His.