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

Vol. 20, Issue 23, 4985-4996, December 1, 2009

This Article Abstract Full Text (PDF) Supplemental Materials All Versions of this Article: E09-04-0324v1 20/23/4985    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Barfield, R. M. Articles by Schekman, R. PubMed PubMed Citation Articles by Barfield, R. M. Articles by Schekman, R.

The Exomer Coat Complex Transports Fus1p to the Plasma Membrane via a Novel Plasma Membrane Sorting Signal in Yeast

Robyn M. Barfield^*, J. Christopher Fromme^*,, and Randy Schekman^*

*Department of Molecular and Cell Biology and Howard Hughes Medical Institute, University of California, Berkeley, Berkeley, CA, 94720; and Department of Molecular Biology and Genetics, Weill Institute for Cell and Molecular Biology, Cornell University, Ithaca, NY 14853

Submitted April 22, 2009; Revised September 22, 2009; Accepted September 28, 2009
Monitoring Editor: Benjamin S. Glick

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Sorting of transmembrane cargo proteins to different cellular compartments is mediated by sorting signals that are recognized by coat proteins involved in vesicle biogenesis. We have identified a sorting signal in the yeast cell fusion protein Fus1p that is required for its transport from the trans-Golgi compartment to the plasma membrane. Transport of Fus1p from the trans-Golgi to the cell surface is dependent on Chs5p, a component of the multisubunit exomer complex. We show that Fus1p transport is also dependent on the exomer components Bch1p and Bud7p. Disruption of the clathrin adaptor protein complex 1 (AP-1) restores Fus1p localization to the cell surface in the absence of exomer, possibly by promoting an alternate, exomer-independent route of transport. Mutation of an IXTPK sequence in the cytosolic tail of Fus1p abolishes its physical interaction with Chs5p, results in mislocalization of Fus1p, and therefore causes a cell fusion defect. These defects are suppressed by disruption of AP-1. We suggest that IXTPK comprises a novel sorting signal that is recognized and bound by exomer leading to the capture of Fus1p into coated vesicles en route to the cell surface.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Sorting of transmembrane cargo proteins to the various compartments within the secretory pathway is mediated by specific sorting signals recognized by selective coat protein complexes involved in transport and vesicle budding. Coat complexes as well as the sorting signals they recognize have been identified for many of the transport steps in the secretory pathway. For example, the COPII coat recognizes diacidic and hydrophobic based sorting signals to mediate transport from the endoplasmic reticulum (ER) to the Golgi complex, and the COPI coat recognizes a dilysine-based sorting signal to mediate transport from the Golgi to the ER (Lee et al., 2004). The clathrin coat, along with clathrin adaptor proteins (APs) and the GGAs recognize dileucine and tyrosine based sorting signals to mediate transport between the trans-Golgi network (TGN), endosomes and the vacuole (Bonifacino and Traub, 2003). Much less is understood about the coat complexes that mediate TGN-to-plasma membrane transport and about the sorting signals these coats recognize.

In polarized epithelial cells, examples of sorting signals that target cargo to the basolateral membrane include tyrosine- and dileucine-based motifs (Folsch, 2008) as well as the PDZ-binding ligand of syndecan-1 (Maday et al., 2008). However, a protein coat that recognizes these signals has not been identified. Targeting of cargo to the apical membrane may involve glycosylation as well as association with lipid rafts (Folsch, 2008). In yeast and mammalian systems, diacylgycerol levels, which are maintained in part by the phosphatidylinositol-transfer protein Sec14p and the peripheral Golgi protein Nir2, are critical for protein transport to the plasma membrane (Corda et al., 2002; Litvak et al., 2005). Phosphatidylinositol-4-phosphate is also involved in TGN-to-plasma membrane transport in yeast (Hama et al., 1999; Walch-Solimena and Novick, 1999; Audhya et al., 2000). Recently, the protein complex exomer was identified as a coat complex involved in TGN-to-plasma membrane transport in yeast (Sanchatjate and Schekman, 2006; Trautwein et al., 2006; Wang et al., 2006), although a sorting signal recognized by exomer was not defined.

Exomer is a protein complex important for the transport of select membrane proteins from the TGN to the plasma membrane in yeast (Santos and Snyder, 1997, 2003). It is composed of five subunits: Chs5p and a family of four related proteins—Chs5p-Arf1p binding proteins (ChAPs) (Sanchatjate and Schekman, 2006; Trautwein et al., 2006; Wang et al., 2006). The four ChAPs include two pairs of closely related paralogues: Chs6p and Bch2p; Bch1p and Bud7p (Sanchatjate and Schekman, 2006; Trautwein et al., 2006; Wang et al., 2006). All five exomer components are cytosolic proteins that colocalize with late Golgi markers and associate with each other both in vivo and in vitro (Santos and Snyder, 1997, 2000; Sanchatjate and Schekman, 2006; Trautwein et al., 2006; Wang et al., 2006). Chs5p may play a structural role in organizing exomer and the ChAPs may confer cargo specificity (Sanchatjate and Schekman, 2006). When baculovirus-purified exomer is combined with Escherichia coli-purified Arf1p-guanosine 5'-O-(3-thio)triphosphate, exomer formed an electron-dense protein coat on synthetic liposomes reminiscent of the COPII protein coat. However, in the case of exomer, coated buds or coated vesicles were never observed as they were with COPII (Matsuoka et al., 1998; Wang et al., 2006). Therefore, exomer alone is not a complete protein coat sufficient to deform membranes and induce bud formation.

The exomer component Chs5p is known to be important for the plasma membrane localization of two transmembrane cargo molecules: Chs3p, which is required at the cell surface for synthesis of the yeast cell wall component chitin; and Fus1p, which is required at the cell surface for cell fusion during the yeast mating response (Trueheart et al., 1987; Trueheart and Fink, 1989; Santos and Snyder, 1997, 2003; Nolan et al., 2006). In chs5 mutants, neither Chs3p nor Fus1p localizes to the plasma membrane and instead both are found only in intracellular patches that seem to correspond to a TGN or endosomal compartment (Santos and Snyder, 1997, 2003). In addition to Chs5p, Chs3p also requires Chs6p or Bch1p and Bud7p for its plasma membrane localization (Ziman et al., 1998; Sanchatjate and Schekman, 2006; Trautwein et al., 2006).

We hypothesized that one or more of the ChAPs would play a role in Fus1p localization. We have identified Bch1p and Bud7p as the ChAPs required for localization of Fus1p to the plasma membrane. In addition, we found that the exomer-dependent localization of Fus1p to the plasma membrane relies on a novel IXTPK sorting signal found within the cytosolic tail of Fus1p and that exomer physically interacts with Fus1p in a sorting signal dependent manner.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Yeast Strains, Media, and Growth Conditions
The yeast strains used in this study are listed in Table 1. Yeast strains were grown in standard rich media (YPD) or synthetic dextrose media (SD) supplemented with appropriate amino acids. All strains were cultured at 30°C. Strains were constructed either by integration of disruption cassettes replacing the entire coding region (Brachmann et al., 1998; Longtine et al., 1998), by integration of insertions at the COOH-terminal codon (Longtine et al., 1998) or by tetrad dissection of sporulated diploid strains. Allelic replacements or insertions were confirmed by polymerase chain reaction (PCR). chs6::HIS3, chs6::TRP1, bud7::KANMX4, and fus1::KANMX4 disruption cassettes were generated by PCR amplification from total genomic DNA isolated from TSY49, TSY52, EUROSCARF YOR299w::KANMX4, and EUROSCARF YCL027w::KANMX4, respectively.

For colocalization experiments, FUS1^151-200-GFP (pRB210) or FUS1^{176-ALAAA-180}-GFP (pRB254) was transformed into YRB218 for colocalization with Sec7-RFP or cotransformed into YRB191 with pFUS1-KEX2-RFP (pRB277), pFUS1-RFP-TLG1 (pER21), or pADH1-RFP-TLG1 (pER22). Percentage of Fus1p-GFP punctae that colocalized with the TGN/endosomal red fluorescent protein (RFP) markers was calculated by selecting clear, distinct Fus1p-GFP punctae and then determining the percentage of those punctae that colocalized with Sec7p-RFP, Kex2p-RFP, or RFP-Tlg1p. There was slightly higher colocalization of Fus1p-GFP with RFP-Tlg1p driven by the ADH1 promoter compared with the FUS1 promoter.

Protease Protection Assays
MATa bar1 FUS1–13Myc::TRP1 cells were grown in YPD to OD₆₀₀ 0.3–0.4. The culture was treated with 50 ng/ml -factor pheromone peptide for 90 min and then treated on ice with NaN₃ for 10 min. Nine OD₆₀₀ cell equivalents were centrifuged at 2000 x g for 3 min. Cell pellets were resuspended in 600 µl of 100 mM Tris, pH 9.4, and 10 mM dithiothreitol and incubated at room temperature for 10 min. Cells were centrifuged at 2000 x g for 2 min, washed once with 1 ml of buffer B (1 M sorbitol, 40 mM HEPES, pH 7.5, 160 mM KoAC, and 5 mM MgCl₂) and resuspended in 300 µl of buffer B. Cells were split into three equal aliquots and incubated at room temperature for 10 min: one aliquot was an untreated control, one aliquot was treated with 0.5 mg/ml proteinase K (Roche Diagnostics, Indianapolis, IN), and one aliquot was treated with 0.5 mg/ml proteinase K in the presence of 0.5% Triton X-100. Samples were precipitated with 17% trichloroacetic acid for 30 min on ice, rinsed with acetone, resuspended in 100 µl of sample buffer (8 M urea, 4% SDS, 125 mM Tris, pH 6.8, 20% glycerol, and 5% β-mercaptoethanol), agitated with glass beads for 7 min and solubilized at 55°C for 10 min. Aliquots (10 µl) of a 1:10 dilution of each sample were resolved by SDS-polyacrylamide gel electrophoresis (PAGE), transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA), and immunoblotted with anti-Myc antibodies (diluted 1:20,000; Cell Signaling Technology, Danvers, MA).

To quantify the amount of Fus1p-Myc accessible to proteinase K, we calculated the percentage of full-length Fus1p-Myc relative to the total Fus1p-Myc (the two prominent species added together) in each lane. We then divided the percentage of the full-length Fus1p-Myc species in the protease treated lane by the percentage in the untreated lane and subtracted this from 100%. Proteins were quantified using Photoshop.

Quantitative Cell Fusion Assays
Cell fusion assays were performed essentially as described previously (Heiman and Walter, 2000), with MAT fus1 cells expressing soluble cytosolic GFP (pDN291; Ng and Walter, 1996). Cells were incubated on YPD for 135 min and fixed on ice in 4% paraformaldehyde for at least 1 h before being washed twice in 1x phosphate-buffered saline and visualized by fluorescence microscopy. For cell fusion assays quantified in Table 4, we mated MAT fus1 cells with MATa fus1 cells expressing the various FUS1-GFP constructs.

Yeast Two-Hybrid
The two-hybrid strain PJ69-4A was cotransformed with pGAD-C1 (empty vector or pGAD-FUS1) and pGBDU-C1 (empty vector or pGBDU-CHS5) (James et al., 1996). Cells were grown to saturation in SD lacking leucine and uracil and 4 µl of 1:5 serial dilutions were spotted on SD lacking leucine and uracil and SD lacking leucine, uracil, and histidine.

Chs5p-TAP Purification
Tandem affinity purification of Chs5p, including cross-linking with 5 mM dithiobis(succunimidylpropionate) (DSP), was performed essentially as described previously (Sanchatjate and Schekman, 2006). The Chs5-TAP strain was transformed with wild-type or mutant pMET25-FUS1–3xHA and grown in SD lacking uracil and methionine.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
BCH1 and BUD7 Are Redundantly Required for Fus1p Plasma Membrane Localization
Fus1p is an O-glycosylated type I membrane protein that localizes to the plasma membrane at the tip of shmoos, or mating arrested cells, during the yeast mating response (Trueheart and Fink, 1989). Fus1p is required for cells of opposite mating type to fuse (Trueheart et al., 1987; Nolan et al., 2006). In addition to its cell surface localization, Fus1p is found in intracellular punctae thought to represent a late Golgi or post-Golgi compartment (Santos and Snyder, 2003; Figure 1Aa). The cell surface localization of Fus1p is dependent on the exomer complex component Chs5p. In chs5 cells, Fus1p-GFP is not found at the cell surface and is only seen intracellularly (Santos and Snyder, 2003; Figure 1Ab). We hypothesized that in addition to Chs5p, Fus1p shmoo tip localization would also require one of the four ChAPs that are part of the exomer complex. To test this hypothesis, we examined Fus1p-GFP localization in ChAP single mutant cells treated with -factor for 90 min, conditions that induce expression and localization of Fus1p to the shmoo tip. We found that each of the four ChAP single mutants exhibited Fus1p-GFP tip localization similar to wild-type cells (Table 3 and Figure 1Ad). Even triple mutant cells where BCH2, CHS6, and BUD7 were deleted showed Fus1p-GFP shmoo tip localization indistinguishable from wild-type cells (Table 3 and Figure 1Ac). However, when BCH1 was deleted in combination with any of the other three ChAPs, Fus1p-GFP was at least partially mislocalized (Table 3). In bch1 bud7 double mutants, the mislocalization defect was as severe as in a chs5 mutant; very few cells showed Fus1-GFP tip localization but instead demonstrated a punctate intracellular localization pattern similar to chs5 mutants (Table 3 and Figure 1A, b and e). This mislocalization defect was not due to a decrease in Chs5p expression; expression of Chs5p was similar in all ChAP mutant combinations (Supplemental Figure 1).

View larger version (46K): [in this window] [in a new window]   Figure 1. BCH1 and BUD7 are redundantly required for Fus1p plasma membrane localization. (A) Fus1p-GFP localization in (a) wild-type, (b) chs5, (c) bch2 chs6 bud7, (d) bch1, and (e) bch1 bud7 cells. All cells were also mutant for BAR1. Cells were treated with -factor for 90 min and then visualized by fluorescence microscopy. (B) Fus1p-Myc protease accessibility in exomer mutants. Asterisk (*) indicates endogenously cleaved Fus1p-Myc and exogenously cleaved Fus1p-Myc (see text for details). Quantifications are the average of five or more independent experiments. Error bars represent the 95% confidence interval.

In wild-type cells treated with proteinase K, almost 45% of Fus1p-Myc was at the cell surface and accessible to protease (Figure 1B). These data indicate that although the majority of wild-type cells have cell surface expression of Fus1p-GFP by microscopy, a large portion of the total Fus1p is localized intracellularly. In contrast to wild-type cells, <5% of Fus1p-Myc was accessible to protease in chs5 mutant cells, consistent with the Fus1p-GFP microscopy data (Table 3 and Figure 1B). bch2 chs6 bud7 triple mutants had wild-type levels of protease accessible Fus1p-Myc, and bch1 single mutants showed a slight decrease in protease accessibility (Figure 1B). However, deleting BCH1 and BUD7 together had a much more dramatic effect on protease accessibility; 19% of Fus1p-Myc was accessible in bch1 bud7 double mutants, less than half of the wild-type level (Figure 1B). Surprisingly, this accessibility was still well above what we measured in chs5 cells. These data indicate that although the percentage of cells with Fus1p localized to the shmoo tip (as measured by fluorescence microscopy) was similar in chs5 and bch1 bud7 cells, the proportion of Fus1p molecules that is localized to the shmoo tip (as measured by protease protection) was much less in chs5 cells compared with bch1 bud7 cells. Even the ChAP quadruple mutant (bch2 chs6 bch1 bud7) showed 13 ± 7% protease accessibility, suggesting that Chs5p can partially function without the four ChAPs in the transport of Fus1p.

Because Fus1p must be present at the cell surface to promote cell fusion, we predicted that bch1 bud7 double mutants would have a cell fusion defect similar to fus1 mutants. To test this hypothesis, we conducted filter mating assays to measure the ability of bch1 bud7 cells to function in cell fusion. A severe defect in cell fusion is seen only when both mating types are mutant for FUS1 (Trueheart et al., 1987); therefore, we combined MAT fus1 mutants carrying a soluble cytosolic GFP marker (Ng and Walter, 1996) with MATa bch1 bud7 mutants. After concentrating the cells on filter paper and mating on YPD, we fixed the cells and used fluorescence microscopy to score mating pairs as fused if the cytosolic GFP marker was observed in both halves of the mating pair or unfused if the cytosolic GFP marker was restricted to one-half of the mating pair. As predicted, bch1 bud7 double mutants showed a cell fusion defect similar to fus1 mutants (28 ± 11 and 16 ± 7% relative to wild-type, respectively), whereas bch1 and bud7 single mutants showed similar cell fusion to wild-type (98 ± 26 and 87 ± 13% relative to wild-type, respectively). One caveat however, is that bch1 bud7 double mutants are also defective in localizing Chs3p to the cell surface (Sanchatjate and Schekman, 2006; Trautwein et al., 2006) and CHS3 is also required for cell fusion (Trilla et al., 1999). Therefore, the cell fusion defect seen in bch1 bud7 double mutants could be due to the inability of these cells to properly localize Fus1p, Chs3p, or both. However, based on our microscopy data and our protease protection experiments, we conclude that BCH1 and BUD7, the two most highly similar ChAPs, are redundantly required for the localization of Fus1p to the plasma membrane.

AP-1 Is Important for the Intracellular Localization of Fus1p
In addition to its localization at the cell surface, Fus1p localizes to intracellular punctae (Santos and Snyder, 1997; Figure 1Aa) reminiscent of the intracellular compartments where another exomer cargo, Chs3p, resides (Chuang and Schekman, 1996; Santos and Snyder, 1997). The intracellular stores of Chs3p seem to be maintained by recycling between the TGN and early endosome. This recycling event is dependent on the clathrin adaptor AP-1, although the direction in which AP-1 transports cargo is not clear. Disruption of the AP-1 complex results in relocalization of Chs3p to the cell surface in the absence of exomer (Valdivia et al., 2002). Our lab has suggested that in the absence of AP-1 and exomer, Chs3p accesses an alternate pathway to the plasma membrane (Valdivia et al., 2002).

We considered the possibility that Fus1p and Chs3p might share an alternate path to the cell surface in an exomer AP-1 double mutant. Fus1p-GFP localization was examined in chs5 and bch1 bud7 cells also deleted for APL2, the large subunit of the AP-1 complex. These combined mutations restored Fus1p tip localization to wild-type levels (Table 3 and Figure 2A, c and e). To further validate these findings, we conducted protease protection assays and found that the Fus1p-Myc protease inaccessibility phenotype seen in chs5 and bch1 bud7 mutants was suppressed by a mutation in APL2 (Figure 2B). The cell fusion defect seen in chs5 and bch1 bud7 mutants was also suppressed by a mutation in APL2. chs5 apl2 mating pairs fuse 65 ± 5% of the time compared with 9 ± 5% for chs5 cells, relative to wild type. bch1 bud7 apl2 mating pairs fuse 62 ± 23% of the time compared with 22 ± 13% for bch1 bud7 cells, relative to wild type. These results suggest that the intracellular localization of Fus1p is maintained by recycling between the TGN and early endosome in an AP-1–dependent manner. When this recycling is not maintained, Fus1p accesses the cell surface by an exomer-independent mechanism.

View larger version (43K): [in this window] [in a new window]   Figure 2. AP-1 is important for the intracellular retention of Fus1p in chs5 and bch1 bud7 mutants. (A) Fus1p-GFP localization in (a) wild-type, (b) chs5, (c) chs5 apl2, (d) bch1 bud7, and (e) bch1 bud7 apl2 cells. All cells were also mutant for BAR1. Cells were treated with -factor for 90 min and then visualized by fluorescence microscopy. (B) Fus1p-Myc protease accessibility in exomer AP-1 mutants. Asterisk (*) indicates endogenously cleaved Fus1p-Myc and exogenously cleaved Fus1p-Myc (see text for details). Quantifications are the average of three or more independent experiments. Error bars represent the 95% confidence interval.

View larger version (11K): [in this window] [in a new window]   Figure 3. Fus1p-GFP constructs. Green residues indicate critical residues for Fus1p-GFP shmoo tip localization. (A) Cartoon of Fus1p with transmembrane domain (TM), proline-rich domain (P), and Src homology 3 domain (SH3). Fus1p-GFP construct is indicated on the left. Top, the 50-112-amino acid deletion constructs are depicted. Bottom, residues 151-200 are shown and the small deletion constructs within this region and the 176ILTPK180 to 176ALAAA180 construct are depicted. Ability to rescue fus1 cell fusion defect (cell fusion) and GFP localization to the shmoo tip are indicated on the right. ++, +, and – indicate 67–100, 34–66, and 0–33%, respectively. (B) Protein sequence alignment of Fus1p-like proteins from S. cerevisiae, Vanderwaltozyma polyspora, and Zygosaccharomyces rouxii. Identical residues are indicated by an asterisk (*), very similar residues by a colon (:), and somewhat similar residues by a period (.).

100–150

Table 4 shows the percentage of fus1 mating pairs that fused when the different FUS1-GFP constructs were expressed. We normalized the cell fusion percentage of fus1 cells expressing wild-type FUS1-GFP to 100%. FUS1^301-350-GFP and FUS1^351-400-GFP constructs complemented the fus1 cell fusion defect as well as wild-type. FUS1^201-250-GFP and FUS1^401-450-GFP constructs partially complemented the fus1 cell fusion defect. FUS1^151-200-GFP, FUS1^251-300-GFP and FUS1^401-512-GFP constructs failed to complement the fus1 cell fusion defect (Table 4 and Figure 3).

In Table 4, the percentage of cells that have Fus1p-GFP localized to the shmoo tip is indicated. Of the five deletion constructs that did not fully complement the fus1 cell fusion defect, only Fus1p^151-200-GFP was as severely mislocalized as wild-type Fus1p-GFP in chs5 and bch1 bud7 mutants (Tables 3 and 4). Cells expressing Fus1p^151-200-GFP showed a nearly complete lack of GFP localization at the tip and had a very clear intracellular punctate localization pattern (Figure 4Ab). Fus1p^201-250-GFP partially localized to the vacuolar membrane (Figure 4Ac), Fus1p^251-300-GFP either localized to the plasma membrane or was undetectable (Figure 4Ad) and Fus1p^401-450-GFP and Fus1p^401-512-GFP localized predominantly to the shmoo tip (Figure 4Ae). These data suggest that cells expressing FUS1^201-250-GFP, FUS1^251-300-GFP, FUS1^401-450-GFP, and FUS1^401-512-GFP may be defective in some functional aspect of cell fusion whereas cells expressing FUS1^151-200-GFP are most likely defective in Golgi-to-cell surface traffic.

View larger version (47K): [in this window] [in a new window]   Figure 4. Fus1p 176IXTPK180 is necessary for the plasma membrane localization of Fus1p. (A) Localization of (a) wild-type Fus1p-GFP, (b) Fus1p151-200-GFP, (c) Fus1p201-250-GFP, (d) Fus1p251-300-GFP, (e) Fus1p401-512-GFP, and (f) Fus1p176-ALAAA-180-GFP in wild-type cells and (g) Fus1p151-200-GFP and (h) Fus1p176-ALAAA-180-GFP in apl2 cells. (B) Colocalization of (a–c) Fus1p151-200-GFP and (d–f) Fus1p176-ALAAA-180-GFP with pFUS1-RFP-Tlg1p. Arrows indicate examples of colocalized punctae. All cells in A and B were also mutant for endogenous FUS1 and BAR1. Cells were treated with -factor for 90 min and then visualized by fluorescence microscopy.

151-200

151-200

151-200

151-200

151-200

151-200

Fus1p ₁₇₆IXTPK₁₈₀ Is Necessary for the Exomer-dependent Plasma Membrane Localization of Fus1p
To refine the sorting signal region, we made four smaller deletions within amino acids 151-200 in the context of Fus1p-GFP and tested the ability of these deletion constructs to complement the fus1 cell fusion defect (Figure 3A). FUS1^151-163-GFP and FUS1^190-200-GFP fully complemented fus1 (Table 4). FUS1^164-176-GFP partially complemented fus1 whereas FUS1^164-174-GFP complemented almost as well as wild type (Table 4). FUS1^177-189-GFP complemented 41% relative to wild type (Table 4). Thus, we focused on the region of Fus1p from residues 175-189.

FUS1^184-189-GFP showed reduced complementation in fus1, but of the alanine substitutions of residues 175-189, only four—I176A, T178A, P179A, and K180A—showed reduced complementation in fus1 (Table 4). Surprisingly, FUS1^I176A-GFP was more defective than FUS1^164-176-GFP, a construct that deletes I176. However, we noticed that in deleting residues 164–176, we juxtaposed I163 to L177, creating the wild-type IL sequence. Although individual alanine substitutions of residues I176, T178, P179, and K180 displayed partial effects, a combined mutant, ₁₇₆ILTPK₁₈₀ to ₁₇₆ALAAA₁₈₀, was as severely defective in fusion as FUS1^151-200-GFP (Table 4).

We found that the shmoo tip localization of the mutant forms of Fus1p correlated with the cell fusion defect (Table 4). For example, Fus1p^{176-ALAAA-180}-GFP, which had the most severe cell fusion defect, was as severely mislocalized as wild-type Fus1p-GFP in chs5 or bch1 bud7 mutants (Tables 3 and 4 and Figure 4Af). Furthermore, the cell fusion and mislocalization defects were suppressed by a mutation in APL2 (Table 4 and Figure 4Ah).

Although the size, number and intensity of Fus1p-GFP intracellular punctae vary from cell to cell in cells expressing wild-type and mutant FUS1-GFP (probably due to expression from a CEN plasmid), punctae seen in cells expressing FUS1^{176-ALAAA-180}-GFP or FUS1^151-200-GFP are indistinguishable from each other and resemble TGN/endosomal punctae. We coexpressed these mutant FUS1-GFP constructs with each of three late Golgi/endosomal markers—Sec7p-RFP, Kex2p-RFP, or RFP-Tlg1p—and determined that Fus1p^{176-ALAAA-180}-GFP and Fus1p^151-200-GFP had similar colocalization patterns with all three markers. Fus1p^151-200-GFP punctae showed 15% colocalization with Sec7p-RFP, 27% colocalization with Kex2p-RFP, and 32 or 52% colocalization with RFP-Tlg1, depending on the promoter used (Figure 4B, a–c). Fus1p^{176-ALAAA-180}-GFP punctae showed 19% colocalization with Sec7p-RFP, 30% colocalization with Kex2p-RFP and 40 or 46% colocalization with RFP-Tlg1, depending on the promoter used (Figure 4B, d–f). Therefore, Fus1p^151-200-GFP and Fus1p ^{176-ALAAA-180}-GFP are indistinguishable in their inability to rescue cell fusion, their mislocalization, and the colocalization of their intracellular punctae with TGN/endosomal markers.

An alignment between Saccharomyces cerevisiae Fus1p and Fus1p from other yeasts show that IXTPK is conserved whereas residues 184-189 are not (Figure 3B). Because residues 184-189 are not individually important for cell fusion (Table 4) and are not conserved among other Fus1p species, whereas the IXTPK signal is important for cell fusion and Fus1p tip localization and is conserved, we suggest that residues 184-189 may serve a structural role for the Fus1p ₁₇₆IXTPK₁₈₀ sorting signal.

Fus1p ₁₇₆IXTPK₁₈₀ Is Required for the Physical Interaction between Fus1p and Exomer
We have shown that Chs5p, Bch1p, Bud7p, and Fus1p residues ₁₇₆IXTPK₁₈₀ are important for the plasma membrane localization of Fus1p. Others have shown that the cytosolic domain of Fus1p physically interacts with the exomer component Chs5p in a yeast two-hybrid assay (Nelson et al., 2004). Based on these observations, we predicted that Fus1p ₁₇₆IXTPK₁₈₀ would be important for a physical interaction with Chs5p. To test this hypothesis, we examined the ability of IXTPK mutant versions of the Fus1p cytosolic domain (97–512) to bind to Chs5p in a yeast two-hybrid assay. We found that mutation of any of the residues critical for Fus1p-GFP plasma membrane localization (T178A, P179A, K180A, and ₁₇₆ILTPK₁₈₀ to ₁₇₆ALAAA₁₈₀) abolished binding to Chs5p, whereas mutation of L177 or deletion of other residues within amino acids 151-200, which do not affect Fus1p-GFP plasma membrane localization, showed wild-type levels of binding to Chs5p (Figure 5A).

View larger version (50K): [in this window] [in a new window]   Figure 5. Fus1p 176IXTPK180 is necessary for the physical interaction between Fus1p and Chs5p. (A) Yeast two-hybrid. Wild-type Fus1p (97-512) or a version of Fus1p (97-512) containing a small deletion or a mutation(s) in the 176ILTPK180 region was fused to the C terminus of the GAL4 activation domain (LEU). Chs5p (1-264) was fused to the C terminus of the GAL4 binding domain (URA). –leu –ura plates measure growth, whereas –leu –ura –his plates measure binding of the GAL4 activation domain to the GAL4 binding domain. (B) In vivo cross-linking and coimmunoprecipitation of Chs5p-TAP and Fus1p. Chromosomally TAP-tagged Chs5p was purified from cells expressing either wild-type (WT) or a 176ILTPK180 to 176ALAAA180 mutant (M) version of pMET25-FUS1-HA in the presence (+) or absence (–) of 5 mM DSP cross-linker. Purified Chs5p, which copurifies with all four ChAPs, was visualized in the final elution fraction with sypro staining. Fus1p-HA was visualized in input and final elution fractions by immunoblot analysis using an anti-HA antibody (diluted 1:1000; HA.11, Covance Research Products, Princeton, NJ). Fus1p-HA WT without DSP and Fus1p-HA mutant with DSP were purified 15 and 29% as well as Fus1p-HA WT with DSP, respectively, when factoring Fus1p-HA in the input fraction and purified Chs5p in the final elution fraction. Asterisk (*) indicates endogenously cleaved Fus1p-HA species (see text for details).

The Exomer-dependent IXTPK Sorting Signal Functions in a Restricted Context
We showed that Fus1p ₁₇₆IXTPK₁₈₀ is necessary for the interaction between Fus1p and exomer and for its transport to the cell surface, but is it sufficient? Some signals, such as an ER export signal, can be transplanted to mobilize a neutral reporter protein (Ma et al., 2001). To test whether IXTPK is sufficient for plasma membrane localization, we made chimeric proteins containing the lumenal N terminus of Kex2p, a resident TGN protein (Redding et al., 1991), and the transmembrane and cytosolic C terminus of Fus1p. This Kex2p-Fus1p chimera contained GFP at its C terminus and its expression was driven by the FUS1 promoter. The Kex2p-Fus1p-GFP fusion protein trafficked to the cell surface whereas a truncated fusion lacking the C terminus of Kex2p or Fus1p was mostly retained intracellularly (Figure 6A, a–c). However, the plasma membrane localization of this Kex2p-Fus1p-GFP chimera was not dependent on the exomer component CHS5 (Figure 6A, c and d).

View larger version (60K): [in this window] [in a new window]   Figure 6. The exomer-dependent IXTPK sorting signal functions in a restricted context. (A) GFP localization of (a) full-length Kex2p-GFP (percentage of cells with plasma membrane GFP: 2 ± 2), (b) C-terminally deleted Kex2p-GFP (24 ± 5), and (c) Kex2p-Fus1p-GFP (89 ± 7) in wild-type cells and (d) Kex2p-Fus1p-GFP (76 ± 13) in chs5 cells. All constructs contained the Fus1p transmembrane domain and were under control of the FUS1 promoter. (B) GFP localization of (a) full-length Mid2p-GFP (88 ± 9), (b) C-terminally deleted Mid2p-GFP (ND), and (c) Mid2p-Fus1p-GFP (58 ± 0) in wild-type cells and (d) Mid2p-Fus1p-GFP (60 ± 10) in chs5 cells. All constructs contained the Mid2p transmembrane domain and were under control of the FUS1 promoter. All cells in A and B were also mutant for endogenous FUS1 and BAR1. Cells were treated with -factor for 90 min and then visualized by fluorescence microscopy. Numbers in parentheses are the percentage of shmooing cells with GFP localized to the plasma membrane ± the 95% confidence interval. Numbers are the average of at least two independent experiments where 50–200 cells were scored for each experiment. Plasma membrane localization was not determined for the C-terminally deleted Mid2p-GFP due to its ER localization; cortical ER localization was difficult to distinguish from plasma membrane localization.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
We have shown that Fus1p ₁₇₆IXTPK₁₈₀ is necessary for the localization of Fus1p to the cell surface (Table 4 and Figure 4Af). Because of the requirement for Fus1p at the cell surface for cell fusion during the yeast mating process, mutating ₁₇₆ILTPK₁₈₀ to ₁₇₆ALAAA₁₈₀ renders cells unable to fuse with their mating partner (Table 4). When the intracellular retention of Fus1p ₁₇₆ALAAA₁₈₀ is inhibited by disrupting the AP-1 complex, Fus1p ₁₇₆ALAAA₁₈₀ is redirected to the cell surface and cell fusion is restored (Table 4 and Figure 4Ah). Fus1p ₁₇₆IXTPK₁₈₀ is not only necessary for the plasma membrane localization of Fus1p but also for the physical interaction between Fus1p and the exomer component Chs5p (Figure 5). These data suggest that Fus1p ₁₇₆IXTPK₁₈₀ comprises a plasma membrane sorting signal that is recognized and bound by exomer. Once exomer binds to this IXTPK signal, Fus1p can be captured into exomer-coated vesicles and transported to the cell surface.

IXTPK Defines a Novel Plasma Membrane Sorting Signal in Yeast
To date, there has been no report of a sorting signal in yeast capable of directing cargo to the plasma membrane. Because the Fus1p IXTPK motif shares little or no similarity to other known traffic signals, it seems to represent a novel sorting signal important for the transport of a cargo molecule to the plasma membrane in yeast. In S. cerevisiae, there are no predicted plasma membrane proteins other than Fus1p that contain an IXTPK motif. However, if conservative substitutions are allowed, other plasma membrane proteins contain what may be an equivalent signal. Chs3p, which is also dependent on exomer for its transport to the plasma membrane, contains an LITPK sequence in its second predicted cytosolic loop. However, mutation of this sequence did not affect transport of Chs3p to the cell surface (data not shown). We know that multiple ChAPs (Chs6p or Bch1p and Bud7p) can direct Chs3p to the cell surface (Sanchatjate and Schekman, 2006, Trautwein et al., 2006) so it is possible that multiple sorting signals exist within Chs3p, some of which might be redundant. Future analysis of other IXTPK-like signals and the identification of additional exomer-dependent cargo proteins may help explain the nature of exomer-cargo interactions.

How Does Exomer Interact with Its Cargo?
The simplest interpretation of our data are that exomer directly recognizes and binds Fus1p at the IXTPK sequence. However, it is entirely possible that one or more additional proteins intervene. For example, we showed that exomer forms an electron dense coat on synthetic liposomes but is not sufficient to induce membrane curvature or vesicle budding (Wang et al., 2006). Therefore, exomer could function as a coat adaptor and cooperate with other coat or cytoskeletal components needed for vesicle budding from the TGN. Perhaps an unidentified component of the coat is directly responsible for interacting with Fus1p. Attempts to demonstrate a direct in vitro interaction between Fus1p and exomer have thus far not been successful.

Our two-hybrid interaction experiments were conducted using Chs5p as bait because Bch1p and Bud7p were highly self-activating. However, we suggest that the ChAPs provide cargo specificity. The two transmembrane cargo known to require exomer for transport, Chs3p and Fus1p, both require Chs5p but have different requirements for the ChAPs. In addition, Sanchatjate and Schekman found that Chs5p was required for the integrity of the exomer complex whereas mutation of the ChAPs required for Chs3p transport affected binding of exomer to Chs3p (Sanchatjate and Schekman, 2006). We show that Bch1p and Bud7p are redundantly required for the transport of Fus1p to the cell surface. Perhaps each ChAP or redundant pair recognizes distinct sorting motifs or posttranslational modifications on cargo proteins. Bch1p and Bud7p may each be able to recognize the Fus1p IXTPK signal or a modification on this sequence whereas Chs6p and Bch2p cannot.

The IXTPK Sorting Signal Is Context Dependent
We showed that Fus1p ₁₇₆IXTPK₁₈₀ is necessary for the interaction between Fus1p and exomer and for its transport to the cell surface, but when this signal is transplanted to an unrelated reporter protein, it is not sufficient for exomer-dependent transport to the cell surface (Figure 6). We suggest that the C-terminal cytosolic domain of Fus1p is sufficient for plasma membrane localization and, in the context of full-length Fus1p, the IXTPK signal is necessary. However, exomer is not required for transport when the cytosolic domain of Fus1p is transplanted to an unrelated reporter protein. Although some traffic signals can be transplanted, the nature of TGN to plasma membrane signals may not be as straightforward. For example, we know that Fus1p and Chs3p can take multiple routes to the cell surface: an exomer-dependent pathway and an alternative exomer-independent pathway that is revealed in the absence of AP-1. Perhaps, the Kex2p-Fus1p and Mid2p-Fus1p chimeric proteins are able to access yet another exomer-independent pathway to the cell surface.

	ACKNOWLEDGMENTS

 
We thank past and present members of the Schekman laboratory for valuable discussions, especially Alenka opi, Silvere Pagant, Siraprapha Sanchatjate, Danny Scott, Trevor Starr, and Chao-Wen Wang. We thank Elizabeth Roth and Trevor Starr for extremely helpful technical assistance. We thank Bob Fuller and Jennifer Lippincott-Schwartz for advice about Kex2p-Fus1p chimeric proteins and Alex Engel, Eric Grote, Max Heiman, Robin Klemm, and Chris Toret for helpful discussions about Fus1p. We are grateful to the laboratories of Georjana Barnes, David Drubin, Jasper Rine, Jeremy Thorner, and Peter Walter for reagents and use of equipment. This work was supported by National Institutes of Health National Research Service Award grant F32-GM-071164-03 (to R. B.) and grant GM-26755 (to R. S.) and funds from the Howard Hughes Medical Institute (to R. S.).

	Footnotes

 
This was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E09-04-0324) on October 7, 2009.

Address correspondence to: Randy Schekman (schekman{at}berkeley.edu).

Abbreviations used: AP, adaptor protein complex.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Audhya, A., Foti, M., and Emr, S. D. (2000). Distinct roles for the yeast phosphatidylinositol 4-kinases, Stt4p and Pik1p, in secretion, cell growth, and organelle membrane dynamics. Mol. Biol. Cell 11, 2673–2689.[Abstract/Free Full Text]

Bonifacino, J. S., and Traub, L. M. (2003). Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annu. Rev. Biochem 72, 395–447.[CrossRef][Medline]

Brachmann, C. B., Davies, A., Cost, G. J., Caputo, E., Li, J., Hieter, P., and Boeke, J. D. (1998). Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14, 115–132.[CrossRef][Medline]

Chuang, J. S., and Schekman, R. W. (1996). Differential trafficking and timed localization of two chitin synthase proteins, Chs2p and Chs3p. J. Cell Biol 135, 597–610.[Abstract/Free Full Text]

Corda, D., Hidalgo Carcedo, C., Bonazzi, M., Luini, A., and Spano, S. (2002). Molecular aspects of membrane fission in the secretary pathway. Cell Mol. Life Sci 59, 1819–1832.[CrossRef][Medline]

Folsch, H. (2008). Regulation of membrane trafficking in polarized epithelial cells. Curr. Opin. Cell Biol 20, 208–213.[CrossRef][Medline]

Hama, H., Schnieders, E. A., Thorner, J., Takemoto, J. Y., and DeWald, D. B. (1999). Direct involvement of phosphatidylinositol 4-phosphate in secretion in the yeast Saccharomyces cerevisiae. J. Biol. Chem 274, 34294–34300.[Abstract/Free Full Text]

Heiman, M. G., and Walter, P. (2000). Prm1p, a pheromone-regulated multispanning membrane protein, facilitates plasma membrane fusion during yeast mating. J. Cell Biol 151, 719–730.[Abstract/Free Full Text]

James, P., Halladay, J., and Craig, E. A. (1996). Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144, 1425–1436.[Medline]

Lee, M. C., Miller, E. A., Goldberg, J., Orci, L., and Schekman, R. (2004). Bi-directional protein transport between the ER and Golgi. Annu. Rev. Cell Dev. Biol 20, 87–123.[CrossRef][Medline]

Litvak, V., Dahan, N., Ramachandran, S., Sabanay, H., and Lev, S. (2005). Maintenance of the diacylglycerol level in the Golgi apparatus by the Nir2 protein is critical for Golgi secretory function. Nat. Cell Biol 7, 225–234.[CrossRef][Medline]

Longtine, M. S., McKenzie, A., 3rd, Demarini, D. J., Shah, N. G., Wach, A., Brachat, A., Philippsen, P., and Pringle, J. R. (1998). Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast 14, 953–961.[CrossRef][Medline]

Ma, D., Zerangue, N., Lin, Y. F., Collins, A., Yu, M., Jan, Y. N., and Jan, L. Y. (2001). Role of ER export signals in controlling surface potassium channel numbers. Science 291, 316–319.[Abstract/Free Full Text]

Maday, S., Anderson, E., Chang, H. C., Shorter, J., Satoh, A., Sfakianos, J., Folsch, H., Anderson, J. M., Walther, Z., and Mellman, I. (2008). A PDZ-binding motif controls basolateral targeting of syndecan-1 along the biosynthetic pathway in polarized epithelial cells. Traffic 9, 1915–1924.[CrossRef][Medline]

Matsuoka, K., Orci, L., Amherdt, M., Bednarek, S. Y., Hamamoto, S., Schekman, R., and Yeung, T. (1998). COPII-coated vesicle formation reconstituted with purified coat proteins and chemically defined liposomes. Cell 93, 263–275.[CrossRef][Medline]

Nelson, B., Parsons, A. B., Evangelista, M., Schaefer, K., Kennedy, K., Ritchie, S., Petryshen, T. L., and Boone, C. (2004). Fus1p interacts with components of the Hog1p mitogen-activated protein kinase and Cdc42p morphogenesis signaling pathways to control cell fusion during yeast mating. Genetics 166, 67–77.[CrossRef][Medline]

Ng, D. T., and Walter, P. (1996). ER membrane protein complex required for nuclear fusion. J. Cell Biol 132, 499–509.[Abstract/Free Full Text]

Nolan, S., Cowan, A. E., Koppel, D. E., Jin, H., and Grote, E. (2006). FUS1 regulates the opening and expansion of fusion pores between mating yeast. Mol. Biol. Cell 17, 2439–2450.[Abstract/Free Full Text]

Philip, B., and Levin, D. E. (2001). Wsc1 and Mid2 are cell surface sensors for cell wall integrity signaling that act through Rom2, a guanine nucleotide exchange factor for Rho1. Mol. Biol. Cell 21, 271–280.[CrossRef]

Proszynski, T. J., Simons, K., and Bagnat, M. (2004). O-glycosylation as a sorting determinant for cell surface delivery in yeast. Mol. Biol. Cell 15, 1533–1543.[Abstract/Free Full Text]

Redding, K., Holcomb, C., and Fuller, R. S. (1991). Immunolocalization of Kex2 protease identifies a putative late Golgi compartment in the yeast Saccharomyces cerevisiae. J. Cell Biol 113, 527–538.[Abstract/Free Full Text]

Sanchatjate, S., and Schekman, R. (2006). Chs5/6 complex: a multiprotein complex that interacts with and conveys chitin synthase III from the trans-Golgi network to the cell surface. Mol. Biol. Cell 17, 4157–4166.[Abstract/Free Full Text]

Santos, B., and Snyder, M. (1997). Targeting of chitin synthase 3 to polarized growth sites in yeast requires Chs5p and Myo2p. J. Cell Biol 136, 95–110.[Abstract/Free Full Text]

Santos, B., and Snyder, M. (2000). Sbe2p and sbe22p, two homologous Golgi proteins involved in yeast cell wall formation. Mol. Biol. Cell 11, 435–452.[Abstract/Free Full Text]

Santos, B., and Snyder, M. (2003). Specific protein targeting during cell differentiation: polarized localization of Fus1p during mating depends on Chs5p in Saccharomyces cerevisiae. Eukaryot. Cell 2, 821–825.[Abstract/Free Full Text]

Trautwein, M., Schindler, C., Gauss, R., Dengjel, J., Hartmann, E., and Spang, A. (2006). Arf1p, Chs5p and the ChAPs are required for export of specialized cargo from the Golgi. EMBO J 25, 943–954.[CrossRef][Medline]

Trilla, J. A., Duran, A., and Roncero, C. (1999). Chs7p, a new protein involved in the control of protein export from the endoplasmic reticulum that is specifically engaged in the regulation of chitin synthesis in Saccharomyces cerevisiae. J. Cell Biol 145, 1153–1163.[Abstract/Free Full Text]

Trueheart, J., Boeke, J. D., and Fink, G. R. (1987). Two genes required for cell fusion during yeast conjugation: evidence for a pheromone-induced surface protein. Mol. Cell Biol 7, 2316–2328.[Abstract/Free Full Text]

Trueheart, J., and Fink, G. R. (1989). The yeast cell fusion protein FUS1 is O-glycosylated and spans the plasma membrane. Proc. Natl. Acad. Sci. USA 86, 9916–9920.[Abstract/Free Full Text]

Valdivia, R. H., Baggott, D., Chuang, J. S., and Schekman, R. W. (2002). The yeast clathrin adaptor protein complex 1 is required for the efficient retention of a subset of late Golgi membrane proteins. Dev. Cell 2, 283–294.[CrossRef][Medline]

Walch-Solimena, C., and Novick, P. (1999). The yeast phosphatidylinositol-4-OH kinase pik1 regulates secretion at the Golgi. Nat. Cell Biol 1, 523–525.[CrossRef][Medline]

Wang, C. W., Hamamoto, S., Orci, L., and Schekman, R. (2006). Exomer: a coat complex for transport of select membrane proteins from the trans-Golgi network to the plasma membrane in yeast. J. Cell Biol 174, 973–983.[Abstract/Free Full Text]

Ziman, M., Chuang, J. S., Tsung, M., Hamamoto, S., and Schekman, R. (1998). Chs6p-dependent anterograde transport of Chs3p from the chitosome to the plasma membrane in Saccharomyces cerevisiae. Mol. Biol. Cell 9, 1565–1576.[Abstract/Free Full Text]

This Article Abstract Full Text (PDF) Supplemental Materials All Versions of this Article: E09-04-0324v1 20/23/4985    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Barfield, R. M. Articles by Schekman, R. PubMed PubMed Citation Articles by Barfield, R. M. Articles by Schekman, R.