![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 15, Issue 4, 1533-1543, April 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Max Planck Institute for Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, 01307 Dresden, Germany
Submitted July 21, 2003;
Revised December 5, 2003;
Accepted December 11, 2003
Monitoring Editor: Chris Kaiser
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
). Similar pathways exist in nonpolarized cells (Keller et al., 2001). It has been shown that short motifs present on the cytoplasmic domain of transmembrane proteins are recognized as determinants for basolateral delivery (Matter and Mellman, 1994). The principles of sorting to the apical plasma membrane are less understood. It is known that association with lipid rafts plays a role and that protein glycosylation is involved in apical exocytosis (Scheiffele et al., 1995; Simons and Ikonen, 1997; Yeaman et al., 1997). Nonglycosylated secretory proteins are sorted both apically and basolaterally, whereas glycosylated proteins are preferentially sorted to the apical membrane. Also membrane proteins have been demonstrated to use N-glycans and O-glycans for apical delivery (Yeaman et al., 1997; Gut et al., 1998; Benting et al., 1999; Spodsberg et al., 2001).
Similar to mammalian cells, the budding yeast Saccharomyces cerevisiae has been shown to have at least two different routes for biosynthetic transport from the Golgi to the cell surface (Harsay and Bretscher, 1995; David et al., 1998; Gurunathan et al., 2002; Harsay and Schekman, 2002). Little is known about the sorting determinants, specifying surface delivery of proteins. The length of the transmembrane domain has been proposed to regulate transport to the plasma membrane and to the vacuole (Rayner and Pelham, 1997; Levine et al., 2000). For Pma1p, the major plasma membrane protein, it was shown that clustering and association with lipid rafts is required for cell surface delivery (Bagnat et al., 2000, 2001, 2002b; Lee et al., 2002). Similarly, plasma membrane localization of the amino acid permease Tat2p depends on its association with lipid rafts (Umebayashi and Nakano, 2003).
In this study, we have explored whether O-glycosylation plays a role in surface delivery in the yeast S. cerevisiae. In yeast, O-glycosylation is initiated in the endoplasmatic reticulum (ER) and extended in the Golgi complex (Haselbeck and Tanner, 1983; Herscovics and Orlean, 1993). Dolicholphosphate-mannose is used as a donor for the first mannose transfer to serine or threonine residues during O-glycosylation in the ER. GDP-mannose is the mannosyl donor for the elongation process in the Golgi (Babczinski and Tanner, 1973; Sharma et al., 1974). A family of seven ER protein O-mannosyltransferases (PMTs) that initiates the assembly of O-glycosylation chains has been identified (Immervoll et al., 1995; Gentzsch and Tanner, 1996, 1997; Strahl-Bolsinger et al., 1999). The PMT family was classified into PMT1, PMT2, and PMT4 subfamilies, which differ in their protein substrate specificity (Gentzsch and Tanner, 1997; Girrbach and Strahl, 2003). The PMT family is highly redundant and only the simultaneous deletion of PMT1/PMT2 and PMT4 subfamily members is lethal (Gentzsch and Tanner, 1996; Girrbach et al., 2000). Unlike other PMT members, Pmt4p acts as a homomeric complex (Strahl-Bolsinger et al., 1999). It was shown previously that Axl2, a protein involved in bud site selection, is incompletely glycosylated, mislocalized, and rapidly degraded in cells lacking PMT4 (Sanders et al., 1999). Here, we show that Fus1p, a raft associated and O-glycosylated protein involved in cell fusion during mating (Trueheart et al., 1987; Trueheart and Fink, 1989; Bagnat and Simons, 2002a), requires O-glycosylation for its cell surface delivery. In pmt4
mutant cells, Fus1p is not glycosylated and mostly accumulated in late Golgi structures. A chimeric protein lacking an O-glycosylated sequence is missorted to the vacuole and accumulated in late Golgi in wild-type cells, but its secretion can be restored after addition of a portion of a Fus1p-glycosylated domain. Our data suggest that O-glycosylation functions as a sorting determinant for cell surface delivery of Fus1p.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Plasmid and Strain Construction
Plasmids used in this study are listed in Table 2, and expressed constructs are shown in a cartoon in Figure 1. All proteins used in this study were expressed from the centromeric plasmid p416 (Mumberg et al., 1994). Polymerase chain reaction (PCR) products were integrated into the pGEM-T vector by using the TA ligation kit (A3600; Promega, Madison, WI), cut out using restriction sites introduced on primers, and subcloned by the triple ligation method into XbaI/HindIII linearized p416. The FUS1 sequence was amplified from genomic DNA by using primers containing XbaI and BamHI restriction sites, and the TAP-tag sequence was PCR-amplified from pBS1539 (Puig et al., 2001) with primers containing BamHI and HindIII sites. Restriction digested FUS1 and TAP-tag sequences were coligated into p416 to generate the TPQ67 plasmid. pMBQ30 has been described previously (Bagnat and Simons, 2002a). MBQ35 containing Mid2-GFP construct was cloned by coligation of the PCR-amplified MID2 sequence from genomic DNA by using primers containing XbaI and BamHI sites, and the green fluorescent protein (GFP) sequence was PCR amplified with primers containing BamHI and HindIII sites. TPQ53, containing the MidFus construct (Figure 1) fused to GFP, was generated by coligation of a sequence encoding the extracellular domain of Mid2p amplified from MBQ35 by using primers containing XbaI and BglII, and the Fus1p-GFP sequence was amplified from MBQ30 with primers containing BamHI and HindIII sites. TPQ55, containing Fus1-Mid construct (Figure 1), was generated by coligation of the DNA encoding the extracellular domain of Fus1p amplified from MBQ30 with primers containing XbaI and BglII sites and part of the MID2-GFP sequence amplified from MBQ35 with primers containing BglII and HindIII sites. The sequence encoding the N-terminal part of invertase (SUC2), present in four constructs (Figure 1), was amplified from genomic DNA by using primers containing XbaI and BamHI sites. To generate TPQ34, the Inv-Fus (Figure 1) portion of the invertase sequence was coligated with a sequence encoding portion of Fus1-GFP amplified from genomic DNA from the MBY229 strain (Bagnat and Simons, 2002a) with primers containing BamHI and HindIII sites. To generate TPQ80, containing Inv-Mid, the invertase sequence was coligated with the part of MID2-GFP used to generate Fus-Mid construct described above. To generate TPQ52, containing Inv33Fus construct, part of invertase was coligated to portion of FUS1-GFP amplified from MBQ30 with primers containing BglII and HindIII sites. To generate TPQ79, containing the Inv33Mid construct, the sequence encoding the extracellular part of Inv33Fus was amplified from TPQ52 with primers containing XbaI and BglII and coligated with portion of MID2-GFP prepared as described above to generate Fus-Mid. All protein constructs were membrane associated as determined by density gradient centrifugation.
|
|
|
|
cells in the W303 background.
Western Blot
Western blot analysis was performed according to standard procedures. Cells were disrupted by beating with glass beads for 5 min at 4°C and boiled with SDS-PAGE sample buffer containing 5% mercaptoethanol. To detect protein A present in the TAP-tag, we used a peroxidase-anti-peroxidase antibody (P-2026; Sigma-Aldrich, St. Louis, MO). For Western blot analysis of GFP fusion proteins, we used a mouse monoclonal anti-GFP antibody (B-2) (sc-9996; Santa Cruz Biotechnology, Santa Cruz, CA).
Metabolic Labeling and Immunoprecipitation
Cells were grown to mid-log phase in complete synthetic medium without methionine containing 2% raffinose as carbon source and expression of Fus1-GFP was induced for 15 min by addition of 2% galactose. Then, the cells were pulse labeled with 1 mCi of [35S]methionine for 5 min and chased for various times. At the indicated times, samples were taken, and cells were killed in 0.2% sodium azide on ice. Then, the cells were lysed in lysis buffer (50 mM Tris, pH 8, 150 mM NaCl, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and CLAP protease inhibitors mix: 0.1% chymostatin, 0.1% leupeptin, 0.1% antipain, and 0.1% pepstatin) by shaking with glass beads for 5 min at 4°C. After removing cell debris, lysates were adjusted to 1% NP-40 and 0.1% SDS and warmed at 37°C for 5 min. Insoluble material was removed by centrifugation (1 min, 6000 x g), and samples were diluted twofold in IP buffer (10 mM Tris, pH 8, 150 mM NaCl, 2 mM EDTA,1% NP-40, and 0.1% SDS). Then, samples were incubated with protein A and a rabbit anti-GFP antibody (sc-9996; Santa Cruz Biotechnology) for 3 h at room temperature. Immunoprecipitates were washed four times with IP buffer, once with Tris 20 mM, pH 7.4, and subjected to SDS-PAGE. Protein bands were analyzed by autoradiography.
Tunicamycin Treatment and Mating Assay
Cells were grown overnight in raffinose-containing medium, and expression was induced by addition of galactose (2%). Tunicamycin was added (10 µg/ml), and cells were incubated for3hat24°C. Mating assay was performed as described previously (Bagnat and Simons, 2002a). Then, 0.5 ml of the overnight culture was diluted 20 times in YPD and incubated for 3 h. The density of cells in the culture was assessed by cell counting in the counting chamber, and 1 x 107 of each mating type cells were mixed together and collected on a nitrocellulose filter. Filters were incubated for 3 h at 24°C on YPD plates to allow for cell fusion. Finally, cells were resuspended in water, diluted 100 times, and an equal amount was plated on SD (selection for diploids) and YPD (nonselective) plates. Mating efficiency was calculated as a ratio of number of colonies from SD plates to number of colonies from YPD plates. MBY1102 (AAY1017) was used as a wild-type Mat alpha strain for mating with ether RH690-15D (wild-type) or MBY249 (pmt4).
Microscopy
Microscopy was performed on live cells that were washed twice in water and resuspended in water for imaging using an Olympus BX61 microscope, RT Slider SPOT camera (Diagnostic Instruments, Sterling Heights, MI) and Meta-Morph software.
Quantification of the Fluorescence Microscopy
Cells were grown until mid-log phase in YPD media, washed twice, and resuspended in YPGal media followed by incubation for 3 h at 24°C to induce protein expression. Images of cells expressing different constructs were taken using the same conditions. Stacks of the pictures were converted to the TIFF format by using ImageJ program (Wayne Rasband, National Institutes of Health, Bethesda, MD), and the fluorescence intensity was analyzed using the IpLab (Scanalytics, Fairfax, VA) software. For each image, the background fluorescence was measured and subtracted. The area representing plasma membrane of each cell was marked as a ring-like segment. The differential interference contrast images were used to localize the periphery of the cells when fluorescence from the plasma membrane was too weak to localize it. The total fluorescence of the plasma membrane was measured from the encircled area. The fluorescence corresponding to the intracellular space was measured from the area inside the ring-like segment. The bars represent the ratio of the total fluorescence measured for the plasma membrane and for the intracellular area. The data are standardized to O-glycosylated constructs (Fus1-GFP and Fus-Mid) expressed in wild-type cells.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Mutant Cells; Trueheart and Fink, 1989; Bagnat and Simons, 2002a). The function of the O-glycosylation of Fus1p is not known. To gain insight into the role of O-glycosylation, we first determined which particular member of the PMT gene family is responsible for the glycosylation of Fus1p. We expressed an epitope-tagged version of the protein in all pmt mutants under the control of the inducible GAL-S promoter. In Western blot analysis of Fus1 expressed in wild-type cells, we detected four specific bands (Figure 2A). These different forms are due to the processing of the protein (Trueheart and Fink, 1989; Bagnat and Simons, 2002a) (see below). A similar pattern was observed when the protein was expressed in pmt1, pmt2, pmt3, pmt5, and pmt6 mutants, indicating that in these mutants the protein was processed normally (Figure 2A). In contrast, in pmt4 cells Fus1 was detected as a single band, suggesting that in this mutant Fus1 was not properly processed. We observed variation in the relative amount of the different forms of the protein produced in different pmt mutants; however, similar variations were observed for the protein produced in wild-type cells in different experiments.
Next, we tested whether altered processing of Fus1 in pmt4 mutant cells affected its cellular localization. A green fluorescent protein fusion (Fus1-GFP) was expressed from the GAL-S promoter (Bagnat and Simons, 2002) in wild-type and pmt4 cells. In wild-type cells, 3 h after induction of expression, Fus1-GFP was localized to the plasma membrane of the bud and to vacuoles (Figure 2B). Fus1p has a very fast turnover (half-life <1 h), and after delivery to the cell surface it is rapidly endocytosed and transported to the vacuole for degradation. In pmt4 cells, transport of Fus1-GFP to the cell surface was inhibited (only 3.6% of cells had surface staining; N = 300); instead, the protein accumulated intracellularly in dot-like structures and in vacuoles (Figure 2B). These structures resembled Golgi or endosomal elements.
To examine which form of Fus1 is produced in pmt4 cells, we expressed the protein in thermosensitive mutants that block biosynthetic traffic along the secretory pathway at different stages. SEC53 encodes a phospho-mannomutase, necessary for the production of dolichol-P-mannose and GDP-mannose, donors of sugars for both N- and O-glycosylation (Babczinski and Tanner, 1973; Sharma et al., 1974; Ruohola and Ferro-Novick, 1987; Kepes and Schekman, 1988). When sec53 cells were incubated at 37°C (the restrictive temperature), both N- and O-glycosylation were blocked. Fus1 expressed in sec53 cells at the restrictive temperature migrated as a single band with the same mobility as the band generated in pmt4 cells (Figure 3), suggesting that Fus1 was not glycosylated in the pmt4 mutant. In sec18 cells, when the protein accumulated in the ER, Fus1 migration on SDS-PAGE was shifted compared with the sec53 form (Figure 3), due to partial glycosylation of the protein. This precursor (p) form was transported to the Golgi where elongation of the mannose chains takes place. The fully glycosylated form of Fus1, accumulating in the Golgi in the sec14 mutant cells, migrated as the slowest band, mature1 (m1). A second mature form of Fus1, mature2 (m2 that migrates as the fastest band) was also generated in the Golgi complex (Figure 3). The big shift in migration between the m1 and m2 forms suggested that the protein was proteolytically cleaved. In sec14 cells, both mature forms were generated, indicating that the cleavage occurred in the Golgi complex.
We also followed the maturation of Fus1-GFP in wild-type and pmt4 mutant cells in a pulse-chase experiment. Cells were grown in medium containing raffinose as carbon source and expression of Fus1-GFP was induced for 15 min by addition of galactose. Then, cells were pulse labeled with [35S]methionine for 5 min and chased for various times. In wild-type cells at the beginning of the chase, the unglycosylated (sec53 form) form and the ER precursor form (p) of the protein were detected (Figure 4). After 5 min of chase, the mature form (m1) was visible and after 30 min of chase the m2 form was generated. As the protein matured the amount of the precursor form was reduced. In contrast, in the pmt4 mutant Fus1-GFP migrated as the unglycosylated form throughout the whole chase period. Only faint bands that represent other forms of Fus1 were detected. Thus, in the pmt4 mutant, Fus1 was not degraded and remained unglycosylated throughout the chase period.
|
pmt4 Cells Do Not Show a General Secretory Defect
To investigate whether the secretory block, which we observed in the pmt4 mutant, is specific for Fus1-GFP or also affects other proteins, we examined the cell surface delivery of Mid2p. Mid2p, a cell wall integrity sensor, is a type I, O-glycosylated membrane protein (Rajavel et al., 1999; Philip and Levin, 2001). Several members of the PMT family are responsible for the glycosylation of Mid2p (Lommet et al., 2004). Mid2-GFP was efficiently delivered to the cell surface in wild-type and pmt4 cells (Figure 5), indicating that in the pmt4 mutant there is no general block in exocytosis.
|
O-Glycans on the Extracellular Domain of Fus1p Are Important for Surface Localization.
Because the extracellular parts of both Fus1p and Mid2p are O-glycosylated and surface delivery of Fus1p but not Mid2p is affected in pmt4, we decided to swap the extracellular domains of Fus1p and Mid2p to evaluate the role of the ectodomains in protein exocytosis. The chimeric protein containing the extracellular domain from Mid2p fused to the transmembrane domain and cytoplasmic tail from Fus1p, followed by GFP was named Mid-Fus (see cartoon in Figure 1). Similar to Mid2, Mid-Fus was efficiently delivered to the cell surface, but weak GFP fluorescence was also localized to the vacuole (Figure 5). Conversely, the GFP-tagged chimeric protein containing the extracellular part from Fus1p fused to the transmembrane domain and cytoplasmic part of Mid2p was named Fus-Mid. Fus-Mid expressed in pmt4 mutant, accumulated intracellularly in dot-like structures and in the vacuole (Figure 5). In SDS-PAGE, Fus-Mid expressed in the pmt4 mutant cells migrated with the same mobility as the protein expressed in sec53 cells at the restrictive temperature (our unpublished data), indicating that Fus-Mid was O-glycosylated in a PMT4-dependent manner.
Next, we asked in which compartments were the unglycosylated forms of Fus1 and Fus-Mid accumulated. To do so, we introduced either Fus1-GFP or Fus-Mid GFP in pmt4 cells expressing Sec7-DsRed, a late Golgi marker (Franzusoff et al., 1991; Rossanese et al., 1999; Rossanese et al., 2001). Most of the intracellular structures that accumulated GFP fusion proteins also showed intensive DsRed staining (Figure 6), indicating that the unglycosylated Fus1 and Fus-Mid accumulated in the Golgi complex. Together, our results indicate that the O-glycosylation of the extracellular part of Fus1p is required for surface delivery of Fus1p.
|
pmt4 Mutant Cells Have a Strong Unilateral Mating Defect
Because in the pmt4 cells Fus1p, a protein involved in cell fusion during yeast mating, is not delivered efficiently to the cell surface, we checked whether the pmt4 cells show a defect in mating. The pmt4 Mat a cells crossed to wild-type Mat 
cells showed an 18-fold reduced mating efficiency compared with control wild-type Mat a cells. fus1 cells show a bilateral mating phenotype that is pronounced only when FUS1 is deleted in both parental strains (Berlin et al., 1991; Gammie et al., 1998). However, the mating defect of the pmt4 cells was unilateral and strongly pronounced when only one of the mating strains was missing the PMT4 gene. Probably, in pmt4 cells not only Fus1p but also other proteins involved in mating were not glycosylated and therefore mislocalized and/or not functional.
Fus1 and Mid2 Chimeric Proteins Lacking O-Glycans Are Retained Intracellularly
Because Fus1p and Fus-Mid lacking O-glycans were blocked in their transport to the cell surface in pmt4 cells, we analyzed whether the intracellular accumulation of both markers is due to the lack of O-glycosylation or to an effect of the PMT4 deletion. To do so, we replaced the extracellular domain of both proteins with a sequence, similar in length, from the N-terminal part of invertase (Suc2p) (Figure 1). Invertase is a soluble secreted protein that is N-glycosylated, but not O-glycosylated. The protein has been used as a reporter fused to different membrane proteins to study membrane sorting (Darsow et al., 2000). Invertase contains an N-terminal signal sequence, which allows the correct translocation of the fusion constructs into the secretory pathway (Li et al., 2002). We expressed the chimeric proteins named Inv-Fus and Inv-Mid containing a GFP tag in wild-type cells. Unlike the O-glycosylated Fus1 and Fus-Mid, both Inv-Fus and Inv-Mid were inhibited in their transport to the plasma membrane and were instead missorted to the vacuole and accumulated in Golgi-like structures (Figure 7A). Out of >300 cells, we could not find any cell with surface staining.
|
|
). By Western blotting, both proteins migrated more slowly when tunicamycin was added to the cell culture compared with control cells not treated with the inhibitor (Figure 7B). In tunicamycintreated cells, both proteins migrated with the same mobility as proteins produced in sec53 cells cultured at the restrictive temperature, indicating that both Inv-Fus and Inv-Mid were N-glycosylated but not O-glycosylated. Next, we asked whether the dot-like structures that accumulated the invertase fusion construct were Golgi structures. We expressed Inv-Fus in wild-type cells producing Sec7-DsRed as before. As shown in Figure 7C, there was partial colocalization of GFP and DsRed staining. Inv-Fus was missorted to the vacuole and also localized in Golgi structures. Most likely, dots that were stained with GFP and did not colocalize with Sec7-DsRed marker were in intermediate structures on the way to the vacuole.
In summary, these results show that Fus1 and Mid2 proteins fused to the N-terminal part of the invertase sequence containing N-glycans accumulated in Golgi structures and were also missorted to the vacuole.
Surface Delivery of Invertase Chimeric Proteins Is Rescued by Addition of an O-Glycosylated Sequence from Fus1p
The invertase fusion proteins were not delivered to the cell surface. This defect could be due to the lack of O-glycosylation. To test this hypothesis, we tried to rescue surface delivery of Inv-Fus and Inv-Mid by addition of a portion (33 amino acids) of the O-glycosylated domain from Fus1p containing 15 potential O-glycosylation sites (Figure 1). We generated "rescue" constructs containing the same N-terminal part of invertase as used before fused to the 33-amino acid sequence of the juxtamembrane part of the extracellular domain of Fus1p, followed by the transmembrane domain and the cytoplasmic tail from either Fus1p or Mid2p with the GFP tag at the C terminus (Figure 1). The new constructs were named Inv33Fus and Inv33Mid. In wild-type cells, unlike the parent invertase fusion proteins, Inv33Fus and Inv33Mid were efficiently delivered to the cell surface (Figure 8A). Most of cells expressing Inv33Mid showed not only GFP fluorescence on the plasma membrane but also a weak fluorescence signal was observed in the ER-like structures. The expression level of Inv33Fus was very low, but similar to Mid2, most of the cells (64.1%; N = 256) expressing Inv33Fus showed fluorescence signal on the plasma membrane.
|
Interestingly, Inv33Fus and Inv33Mid expressed in pmt4 mostly accumulated inside cells in dot-like structures and also in vacuoles (Figure 8A), indicating that the surface delivery of "rescue" constructs was at least partially dependent on PMT4. Next, we checked whether the rescue constructs were O-glycosylated. The Inv33Mid protein expressed in pmt4 cells migrated in SDS-PAGE with a slower mobility than the protein produced in wild-type cells (Figure 8B), suggesting that PMT4 is involved in the O-glycosylation of Inv33Mid. The protein was also N-glycosylated, because we observed a shift in migration of the protein expressed in wild-type cells in the presence of tunicamycin compared with the protein expressed in control cells without tunicamycin (Figure 8B). There was an additional shift in migration of the protein between the form expressed in wild-type cells in presence of tunicamycin and the unglycosylated form produced in sec53 at 37°C (Figure 8B). These experiments show that Inv33Mid contains N-glycans and is also O-glycosylated in a PMT4-dependent manner. Similar results were obtained for Inv33Fus protein (Figure 8C).
Therefore, addition of 33 amino acids from the glycosylated domain of Fus1p allowed surface delivery of the invertase fusion constructs. When we further shortened the 33-amino acid fragment by 10 amino acids (from the N terminus) and inserted a 22-amino acid portion into Inv-Fus and Inv-Mid, these chimeric proteins were not delivered to the plasma membrane and were not O-glycosylated (our unpublished data).
Together, these experiments show that surface delivery of Inv-Fus and Inv-Mid could be rescued by addition of the 33-amino acid segment from the glycosylated domain of Fus1p. Inv33Fus and Inv33Mid were O-glycosylated, and surface delivery of both proteins depended on PMT4. Thus, we conclude that O-glycosylation is required for surface delivery of Fus1p.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
). Its O-glycosylation starts in the ER through the action of the O-mannosyl transferase Pmt4p (Figure 3). Further mannose groups are added to Fus-1p in the Golgi complex and then the processed protein is transported to the plasma membrane (Trueheart and Fink, 1989) (Figure 3). Even when the PMT4 is deleted, a small amount of glycosylated Fus1 was detected (Figures 2, 3, 4). Probably, in the absence of Pmt4p, other members of the PMT family can glycosylate Fus1p, but it occurs with low efficiency. We found that when O-glycosylation is blocked in pmt4 cells Fus1p accumulates intracellularly in punctate structures that colocalize with Sec7p, a Golgi marker (Figure 6). The surface transport of another O-glycosylated plasma membrane protein, Mid2p, was not affected in pmt4 cells. Using chimeric constructs consisting of parts of Fus1p and Mid2p, we demonstrated that the ectodomain of Fus1p was responsible for the blocked transport to the cell surface in pmt4 cells (Figure 5). To find out whether O-glycans were indeed involved in Fus1 sorting and delivery to the cell surface, we used a strategy previously used by Rose and coworkers in mammalian cells (Guan et al., 1985). They demonstrated that a nonglycosylated secretory protein, rat growth hormone, was blocked in the Golgi complex when it was anchored to the membrane by fusing the hormone to the transmembrane and cytosolic domains of vesicular stomatitis virus G protein. However, when one or two N-glycosylation sites were introduced into the growth hormone ectodomain, the N-glycosylated membrane protein was transported to the cell surface. Similarly, when we fused the N-terminal part of the invertase sequence to the transmembrane and cytosolic domains of Fus1p or of Mid2p, these chimeric proteins accumulated intracellularly (Figure 7). Next, we inserted segments from the Fus1p ectodomain between the invertase and the Fus1 transmembrane anchor (see cartoon in Figure 1) in an attempt to find out whether addition of O-glycans could influence transport to the plasma membrane. When a 33-amino acid segment containing several potential O-glycosylation sites was added to the fusion protein construct, the chimeric protein was transported to the plasma membrane (Figures 8 and 9). In pmt4 cells, the protein was blocked in the Golgi complex. Biochemical analysis demonstrated that O-glycans were present on the chimeric protein that was delivered to the cell surface in wild-type cells, whereas O-glycans were lacking in pmt4 cells (Figure 8). Our data suggest that O-glycosylation can serve as a signal for protein transport to the plasma membrane in S. cerevisae.
The first suggestive evidence for carbohydrate side chains of proteins acting as sorting determinants for cell surface delivery came from the work of Rose and coworkers in fibroblasts (Guan et al., 1985; Machamer et al., 1985). Further support for a role of glycans as signals for transport from the Golgi complex to the plasma membrane emerged from studies of epithelial cells. Scheiffele et al. (1995) demonstrated that in epithelial Madin-Darby canine kidney cells N-glycosylated rat growth hormone was secreted apically, whereas the nonglycosylated native form of the protein was secreted randomly, both apically and basolaterally. This was also the case when a glycosyl-phosphatidylinositol-anchor was added to the rat growth hormone (Benting et al., 1999). Gut and coworkers showed that a protein lacking its basolateral sorting determinants accumulated in the Golgi complex (Gut et al., 1998). Addition of N-glycans to this mutant protein promoted its delivery to the apical surface. Several reports have also demonstrated that O-glycans can serve as apical sorting determinants (Yeaman et al., 1997; Alfalah et al., 1999; Spodsberg et al., 2001). In fact, also glycosylated basolateral proteins will be delivered apically if the basolateral sorting determinants in their cytosolic protein domains are mutated or deleted (Gut et al., 1998). These data demonstrate that it is difficult to analyze sorting determinants for surface delivery in cells with two (or more) pathways from the Golgi complex to the plasma membrane. A protein can switch from one pathway to another.
Yeast also has two pathways to the cell surface (Harsay and Bretscher, 1995; David et al., 1998; Gurunathan et al., 2002; Harsay and Schekman, 2002), both using the same fusion machinery comprising the exocyst (TerBush et al., 1996), the GTPase Sec4p (Guo et al., 1999), and the soluble N-ethylmaleimide-sensitive factor attachment protein receptor complexes Snc1/2p and Sso1/2p (Aalto et al., 1993; Protopopov et al., 1993; Jahn et al., 2003). Despite numerous genetic screens for secretion mutants little is known of how sorting of surface proteins occurs in the Golgi complex. Inactivation of one pathway may be rescued by routing to the other pathway (Gurunathan et al., 2002; Harsay and Schekman, 2002). It was shown that blocking of the vacuolar pathway leads to exocytosis of vacuolar proteins (Marcusson et al., 1994; Nothwehr et al., 1995).
Another problem complicates the analysis of the role of glycosylation in secretion. Inhibition of glycosylation often leads to misfolding of the protein, and accumulation and degradation in the ER (Ng et al., 2000; Hampton, 2002; Ellgaard and Helenius, 2003). In S. cerevisae, the Golgi complex also seems to be a site for quality control (Hong et al., 1996; Jorgensen et al., 1999). Proteins that are not folded correctly or are incompletely oligomerized are singled out for the vacuolar delivery in the Golgi complex. Obviously, therefore, one has to differentiate between effects due to quality control mechanisms or to sorting for surface delivery in the Golgi complex. It was reported that Axl2p, involved in bud site selection has altered glycosylation in pmt4 cells and is rapidly degraded before reaching cell surface, most probably in the Golgi, (Sanders et al., 1999). The kinetics of the degradation process of misfolded proteins is usually fast. The chimeric proteins composed of Fus1p ectodomain and of the Mid2p transmembrane and cytosolic domains in pmt4 cells were in fact more long lived than the wild-type Fus1p. Thus, we had no indication of misfolding or increased vacuolar delivery. Instead, the chimeric protein mostly accumulated in the Golgi complex. We did not observe accumulation of our markers in ER structures (except for Inv33Mid, which showed a faint staining of the ER but most of the fluorescence was localized to the plasma membrane), and we did not observe enhanced degradation as was the case for Axl2p (Sanders et al., 1999).
If O-glycans of Fus1p function as sorting determinants for delivery to the cell surface, the question arises how this signal functions mechanistically. Two models have been proposed for how carbohydrate-sorting determinants function in apical transport in epithelial cells. Rodriguez-Boulan and Gonzalez (1999) have suggested that glycans change the biophysical properties of an apical protein such that the presentation of a proteinaceous sorting signal to a hypothetical sorting receptor is facilitated. Alternatively, the glycans contribute to a transport-permissive conformation of the apical protein that facilitate its incorporation into lipid rafts and thus into the apical targeting pathway. The latter possibility seems unlikely because it has been shown that raft association by itself is not sufficient for apical delivery (Simons and Ikonen, 1997; Rietveld and Simons, 1998). Rafts are also routed basolaterally. The second model postulates the existence of apical lectins that bind to the glycans and sort the apical proteins into transport carriers in the trans-Golgi network (Scheiffele et al., 1995). However, such a lectin has not yet been identified. According to this model O-glycosylated Fus1p would be bound by a lectin, which facilitates its surface delivery. Our findings that O-glycans promote plasma membrane transport of proteins in yeast provide all the tools that this model organism supplies for identification of the underlying sorting mechanisms.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Footnotes |
|---|
Article published online ahead of print. Mol. Biol. Cell 10.1091/mbc.E03–07–0511. Article and publication date are available at www.molbiolcell.org/cgi/doi/10.1091/mbc.E03–07–0511.
Present address: Department of Biochemistry and Biophysics, 513 Parnassus Ave., University of California, San Francisco, San Francisco, CA 94143-0448. 
* Corresponding author. E-mail address: simons{at}mpi-cbg.de.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Alfalah, M., Jacob, R., Preuss, U., Zimmer, K.P., Naim, H., and Naim, H.Y. (1999). O-Linked glycans mediate apical sorting of human intestinal sucraseisomaltase through association with lipid rafts. Curr. Biol. 9, 593-596.[CrossRef][Medline]
Babczinski, P., and Tanner, W. (1973). Involvement of dolicholmonophosphate in the formation of specific mannosyl-linkages in yeast glycoproteins. Biochem. Biophys. Res. Commun. 54, 1119-1124.[CrossRef][Medline]
Bagnat, M., Chang, A., and Simons, K. (2001). Plasma membrane proton ATPase Pma1p requires raft association for surface delivery in yeast. Mol. Biol. Cell 12, 4129-4138.
Bagnat, M., Keranen, S., Shevchenko, A., and Simons, K. (2000). Lipid rafts function in biosynthetic delivery of proteins to the cell surface in yeast. Proc. Natl. Acad. Sci. USA 97, 3254-3259.
Bagnat, M., and Simons, K. (2002a). Cell surface polarization during yeast mating. Proc. Natl. Acad. Sci. USA 99, 14183-14188.
Bagnat, M., and Simons, K. (2002b). Lipid rafts in protein sorting and cell polarity in budding yeast Saccharomyces cerevisiae. Biol. Chem. 383, 1475-1480.[CrossRef][Medline]
Benting, J.H., Rietveld, A.G., and Simons, K. (1999). N-Glycans mediate the apical sorting of a GPI-anchored, raft-associated protein in Madin-Darby canine kidney cells. J. Cell Biol. 146, 313-320.
Berlin, V., Brill, J.A., Trueheart, J., Boeke, J.D., and Fink, G.R. (1991). Genetic screens and selections for cell and nuclear fusion mutants. Methods Enzymol. 194, 774-792.[Medline]
Darsow, T., Odorizzi, G., and Emr, S.D. (2000). Invertase fusion proteins for analysis of protein trafficking in yeast. Methods Enzymol. 327, 95-106.[CrossRef][Medline]
David, D., Sundarababu, S., and Gerst, J.E. (1998). Involvement of long chain fatty acid elongation in the trafficking of secretory vesicles in yeast. J. Cell Biol. 143, 1167-1182.
Ellgaard, L., and Helenius, A. (2003). Quality control in the endoplasmic reticulum. Nat. Rev. Mol. Cell. Biol. 4, 181-191.[CrossRef][Medline]
Franzusoff, A., Redding, K., Crosby, J., Fuller, R.S., and Schekman, R. (1991). Localization of components involved in protein transport and processing through the yeast Golgi apparatus. J. Cell Biol. 112, 27-37.
Gammie, A.E., Brizzio, V., and Rose, M.D. (1998). Distinct morphological phenotypes of cell fusion mutants. Mol. Biol. Cell 9, 1395-1410.
Gentzsch, M., and Tanner, W. (1996). The PMT gene family: protein Oglycosylation in Saccharomyces cerevisiae is vital. EMBO J. 15, 5752-5759.[Medline]
Gentzsch, M., and Tanner, W. (1997). Protein-O-glycosylation in yeast: protein-specific mannosyltransferases. Glycobiology 7, 481-486.
Girrbach, V., and Strahl, S. (2003). Members of the evolutionarily conserved PMT family of protein O-mannosyltransferases form distinct protein complexes among themselves. J. Biol. Chem. 278, 12554-12562.
Girrbach, V., Zeller, T., Priesmeier, M., and Strahl-Bolsinger, S. (2000). Structure-function analysis of the dolichyl phosphate-mannose: protein O-mannosyltransferase ScPmt1p. J. Biol. Chem. 275, 19288-19296.
Guan, J.L., Machamer, C.E., and Rose, J.K. (1985). Glycosylation allows cell-surface transport of an anchored secretory protein. Cell 42, 489-496.[CrossRef][Medline]
Guo, W., Roth, D., Walch-Solimena, C., and Novick, P. (1999). The exocyst is an effector for Sec4p, targeting secretory vesicles to sites of exocytosis. EMBO J. 18, 1071-1080.[CrossRef][Medline]
Gurunathan, S., David, D., and Gerst, J.E. (2002). Dynamin and clathrin are required for the biogenesis of a distinct class of secretory vesicles in yeast. EMBO J. 21, 602-614.[CrossRef][Medline]
Gut, A., Kappeler, F., Hyka, N., Balda, M.S., Hauri, H.P., and Matter, K. (1998). Carbohydrate-mediated Golgi to cell surface transport and apical targeting of membrane proteins. EMBO J. 17, 1919-1929.[CrossRef][Medline]
Hampton, R.Y. (2002). ER-associated degradation in protein quality control and cellular regulation. Curr. Opin. Cell Biol. 14, 476-482.[CrossRef][Medline]
Harsay, E., and Bretscher, A. (1995). Parallel secretory pathways to the cell surface in yeast. J. Cell Biol. 131, 297-310.
Harsay, E., and Schekman, R. (2002). A subset of yeast vacuolar protein sorting mutants is blocked in one branch of the exocytic pathway. J. Cell Biol. 156, 271-285.
Haselbeck, A., and Tanner, W. (1983). O-glycosylation in Saccharomyces cerevisiae is initiated at the endoplasmic reticulum. FEBS Lett. 158, 335-338.[CrossRef][Medline]
Herscovics, A., and Orlean, P. (1993). Glycoprotein biosynthesis in yeast. FASEB J. 7, 540-550.[Abstract]
Hong, E., Davidson, A.R., and Kaiser, C.A. (1996). A pathway for targeting soluble misfolded proteins to the yeast vacuole. J. Cell Biol. 135, 623-633.
Immervoll, T., Gentzsch, M., and Tanner, W. (1995). PMT3 and PMT4, two new members of the protein-O-mannosyltransferase gene family of Saccharomyces cerevisiae. Yeast 11, 1345-1351.[CrossRef][Medline]
Jahn, R., Lang, T., and Sudhof, T.C. (2003). Membrane fusion. Cell 112, 519-533.[CrossRef][Medline]
Jorgensen, M.U., Emr, S.D., and Winther, J.R. (1999). Ligand recognition and domain structure of Vps10p, a vacuolar protein sorting receptor in Saccharomyces cerevisiae. Eur. J. Biochem. 260, 461-469.[Medline]
Keller, P., Toomre, D., Diaz, E., White, J., and Simons, K. (2001). Multicolour imaging of post-Golgi sorting and trafficking in live cells. Nat. Cell Biol. 3, 140-149.[CrossRef][Medline]
Kepes, F., and Schekman, R. (1988). The yeast SEC53 gene encodes phosphomannomutase. J. Biol. Chem. 263, 9155-9161.
Lee, M.C., Hamamoto, S., and Schekman, R. (2002). Ceramide biosynthesis is required for the formation of the oligomeric H+-ATPase Pma1p in the yeast endoplasmic reticulum. J. Biol. Chem. 277, 22395-22401.
Lehle, L., and Tanner, W. (1976). The specific site of tunicamycin inhibition in the formation of dolichol-bound N-acetylglucosamine derivatives. FEBS Lett. 72, 167-170.[Medline]
Levine, T.P., Wiggins, C.A., and Munro, S. (2000). Inositol phosphorylceramide synthase is located in the Golgi apparatus of Saccharomyces cerevisiae. Mol. Biol. Cell 11, 2267-2281.
Lommel, M., Bagnat, M., and Strahl, S. (2004). Aberrant processing of the WSC family and Mid2p cell surface sensors results in cell death of Saccharomyces cerevisiae O-Mannosylation mutants. Mol. Cell. Biol. 24, 46-57.
Li, J., Xu, H., Bentley, W.E., and Rao, G. (2002). Impediments to secretion of green fluorescent protein and its fusion from Saccharomyces cerevisiae. Biotechnol. Prog. 18, 831-838.[CrossRef][Medline]
Machamer, C.E., Florkiewicz, R.Z., and Rose, J.K. (1985). A single N-linked oligosaccharide at either of the two normal sites is sufficient for transport of vesicular stomatitis virus G protein to the cell surface. Mol. Cell. Biol. 5, 3074-3083.
Marcusson, E.G., Horazdovsky, B.F., Cereghino, J.L., Gharakhanian, E., and Emr, S.D. (1994). The sorting receptor for yeast vacuolar carboxypeptidase Y is encoded by the VPS10 gene. Cell 77, 579-586.[CrossRef][Medline]
Matter, K., and Mellman, I. (1994). Mechanisms of cell polarity: sorting and transport in epithelial cells. Curr. Opin. Cell Biol. 6, 545-554.[CrossRef][Medline]
Mumberg, D., Muller, R., and Funk, M. (1994). Regulatable promoters of Saccharomyces cerevisiae: comparison of transcriptional activity and their use for heterologous expression. Nucleic Acids Res. 22, 5767-5768.
Ng, D.T., Spear, E.D., and Walter, P. (2000). The unfolded protein response regulates multiple aspects of secretory and membrane protein biogenesis and endoplasmic reticulum quality control. J. Cell Biol. 150, 77-88.
Nothwehr, S.F., Conibear, E., and Stevens, T.H. (1995). Golgi and vacuolar membrane proteins reach the vacuole in vps1 mutant yeast cells via the plasma membrane. J. Cell Biol. 129, 35-46.
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. Cell. Biol. 21, 271-280.
Protopopov, V., Govindan, B., Novick, P., and Gerst, J.E. (1993). Homologs of the synaptobrevin/VAMP family of synaptic vesicle proteins function on the late secretory pathway in S. cerevisiae. Cell 74, 855-861.[CrossRef][Medline]
Puig, O., Caspary, F., Rigaut, G., Rutz, B., Bouveret, E., Bragado-Nilsson, E., Wilm, M., and Seraphin, B. (2001). The tandem affinity purification (TAP) method: a general procedure of protein complex purification. Methods 24, 218-229.[CrossRef][Medline]
Rajavel, M., Philip, B., Buehrer, B.M., Errede, B., and Levin, D.E. (1999). Mid2 is a putative sensor for cell integrity signaling in Saccharomyces cerevisiae. Mol. Cell. Biol. 19, 3969-3976.
Rayner, J.C., and Pelham, H.R. (1997). Transmembrane domain-dependent sorting of proteins to the ER and plasma membrane in yeast. EMBO J. 16, 1832-1841.[CrossRef][Medline]
Rietveld, A., and Simons, K. (1998). The differential miscibility of lipids as the basis for the formation of functional membrane rafts. Biochim. Biophys. Acta 1376, 467-479.[Medline]
Rodriguez-Boulan, E., and Gonzalez, A. (1999). Glycans in post-Golgi apical targeting: sorting signals or structural props? Trends Cell Biol. 9, 291-294.[CrossRef][Medline]
Rossanese, O.W., Reinke, C.A., Bevis, B.J., Hammond, A.T., Sears, I.B., O'Connor, J., and Glick, B.S. (2001). A role for actin, Cdc1p, and Myo2p in the inheritance of late Golgi elements in Saccharomyces cerevisiae. J. Cell Biol. 153, 47-62.
Rossanese, O.W., Soderholm, J., Bevis, B.J., Sears, I.B., O'Connor, J., Williamson, E.K., and Glick, B.S. (1999). Golgi structure correlates with transitional endoplasmic reticulum organization in Pichia pastoris and Saccharomyces cerevisiae. J. Cell Biol. 145, 69-81.
Ruohola, H., and Ferro-Novick, S. (1987). Sec53, a protein required for an early step in secretory protein processing and transport in yeast, interacts with the cytoplasmic surface of the endoplasmic reticulum. Proc. Natl. Acad. Sci. USA 84, 8468-8472.
Sanders, S.L., Gentzsch, M., Tanner, W., and Herskowitz, I. (1999). O-Glycosylation of Axl2/Bud10p by Pmt4p is required for its stability, localization, and function in daughter cells. J. Cell Biol. 145, 1177-1188.
