Portal de Periódicos da CAPES

The VRG4 Gene Is Required for GDP-mannose Transport into the Lumen of the Golgi in the Yeast, Saccharomyces cerevisiae

1997; Elsevier BV; Volume: 272; Issue: 50 Linguagem: Inglês

10.1074/jbc.272.50.31908

1083-351X

Neta Dean, Yian B. Zhang, Jay B. Poster,

Endoplasmic Reticulum Stress and Disease

In the yeast Saccharomyces cerevisiae, glycoproteins and sphingolipids are modified in the Golgi by the addition of mannose residues. The critical mannosyl donor for these reactions is the nucleotide sugar, GDP-mannose, whose transport into the Golgi from the cytoplasm is required for mannosylation. This transport reaction has been well characterized, but the nucleotide sugar transporter has yet to be identified in yeast.VRG4 is an essential gene whose product is required for a number of Golgi-specific functions, including glycosylation and the organization of the endomembrane system. Here, data are presented that demonstrate that the primary role of Vrg4p is in the transport of GDP-mannose into the Golgi. The vrg4 mutation causes a general impairment in mannosylation, affecting N-linked andO-linked glycoprotein modifications as well as the mannosylation of sphingolipids. By using an in vitro assay,vrg4 mutants were shown to be specifically defective in the transport of GDP-mannose into Golgi vesicles. The Vrg4 protein localizes to the Golgi complex in a pattern that suggests a wide distribution throughout the Golgi. Vrg4p displays homology to other putative nucleotide sugar transporters, suggesting that theVRG4 gene encodes a Golgi GDP-mannose transporter. As Vrg4p is essential, these results suggest that a complete lack of mannosylation of glycoproteins in the Golgi leads to inviability. Alternatively, the essential function of Vrg4p in yeast involves its effect on sphingolipids, which would imply a critical role for mannosylinositol phosphorylceramides or mannosyl diphosphoinositol ceramides on growth and viability. In the yeast Saccharomyces cerevisiae, glycoproteins and sphingolipids are modified in the Golgi by the addition of mannose residues. The critical mannosyl donor for these reactions is the nucleotide sugar, GDP-mannose, whose transport into the Golgi from the cytoplasm is required for mannosylation. This transport reaction has been well characterized, but the nucleotide sugar transporter has yet to be identified in yeast.VRG4 is an essential gene whose product is required for a number of Golgi-specific functions, including glycosylation and the organization of the endomembrane system. Here, data are presented that demonstrate that the primary role of Vrg4p is in the transport of GDP-mannose into the Golgi. The vrg4 mutation causes a general impairment in mannosylation, affecting N-linked andO-linked glycoprotein modifications as well as the mannosylation of sphingolipids. By using an in vitro assay,vrg4 mutants were shown to be specifically defective in the transport of GDP-mannose into Golgi vesicles. The Vrg4 protein localizes to the Golgi complex in a pattern that suggests a wide distribution throughout the Golgi. Vrg4p displays homology to other putative nucleotide sugar transporters, suggesting that theVRG4 gene encodes a Golgi GDP-mannose transporter. As Vrg4p is essential, these results suggest that a complete lack of mannosylation of glycoproteins in the Golgi leads to inviability. Alternatively, the essential function of Vrg4p in yeast involves its effect on sphingolipids, which would imply a critical role for mannosylinositol phosphorylceramides or mannosyl diphosphoinositol ceramides on growth and viability. The Golgi complex is the site where the terminal glycosylation of both proteins and lipids occurs. Unlike mammalian cells, in the yeastSaccharomyces cerevisiae, glycoproteins and sphingolipids are exclusively modified by the addition of mannose residues in the Golgi. Glycoproteins can undergo two types of modifications in which oligosaccharides are linked to either asparagine residues (N-linked) or serine/threonine residues (O-linked) (for review see Refs. 1Herscovics A. Orlean P. FASEB J. 1993; 7: 540-550Crossref PubMed Scopus (437) Google Scholar and 2Ballou C.E. Methods Enzymol. 1990; 185: 440-470Crossref PubMed Scopus (271) Google Scholar). Both of these glycosylation pathways initiate in the endoplasmic reticulum (ER) 1The abbreviations used are: ER, endoplasmic reticulum; PCR, polymerase chain reaction; IPC, inositol phosphorylceramides; MIPC, mannosylinositol phosphorylceramides; m(IP)2C, mannose(inositol phosphate)2 ceramide; HA, hemagglutinin; ORF, open reading frame; PYC, permeabilized yeast cells.1The abbreviations used are: ER, endoplasmic reticulum; PCR, polymerase chain reaction; IPC, inositol phosphorylceramides; MIPC, mannosylinositol phosphorylceramides; m(IP)2C, mannose(inositol phosphate)2 ceramide; HA, hemagglutinin; ORF, open reading frame; PYC, permeabilized yeast cells. and terminate in the Golgi. After transport of the protein to the Golgi, mostN-linked oligosaccharides are elongated by a series of different mannosyltransferases to form glycoproteins that contain outer chains of 50 or more mannose residues. The α1,6-linked outer chain is highly branched with α1,2- and α1,3-linked mannoses. As in higher eukaryotes, it appears that the various mannosyltransferases that catalyze these sequential reactions are compartmentalized from one another within the individual Golgi cisternae. In the case ofO-linked sugars, up to five mannoses are added after the addition of the first mannose in the ER (3Gentzsch M. Tanner W. EMBO J. 1996; 15: 5752-5759Crossref PubMed Scopus (217) Google Scholar, 4Gentzsch M. Tanner W. Glycobiology. 1997; 7: 481-486Crossref PubMed Scopus (140) Google Scholar). The phosphoinositol-containing sphingolipids in yeast also undergo mannosylation in the Golgi. In S. cerevisiae, there are three major classes of sphingolipids. These include the inositol phosphorylceramides (IPCs) and the mannosylinositol phosphorylceramides (MIPC and M(IP)2C) (for review, see Ref. 5Lester R.L. Dickson R.C. Adv. Lipid Res. 1993; 26: 253-274PubMed Google Scholar). MIPC and M(IP)2C contain a single mannose attached to the inositol (6Smith S.W. Lester R.L. J. Biol. Chem. 1974; 249: 3395-3405Abstract Full Text PDF PubMed Google Scholar), although little is known about the mannosyltransferase(s) that catalyzes this reaction.The mannosyl donor for all of these Golgi-localized reactions is the nucleotide sugar GDP-mannose, whose site of synthesis is the cytoplasm. Before it can be utilized by the different lumenal mannosyltransferases, GDP-mannose must be transported into the Golgi by a specific nucleotide sugar transporter (7Abeijon C. Orlean P. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 6935-6939Crossref PubMed Scopus (131) Google Scholar). Once the sugar is donated to lumenal mannosyltransferase acceptors, the nucleoside diphosphate GDP is converted to a monophosphate by a nucleoside diphosphatase (7Abeijon C. Orlean P. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 6935-6939Crossref PubMed Scopus (131) Google Scholar). As in the mammalian Golgi, the transport of the nucleotide sugar into the lumen is coupled to the outward exit of the monophosphate in yeast. The yeast GDPase-encoding gene, GDA1, has been isolated (8Abeijon C. Yanagisawa K. Mandon E.C. Hausler A. Moremen K. Hirschberg C.B. Robbins P.W. J. Cell Biol. 1993; 122: 307-323Crossref PubMed Scopus (158) Google Scholar). As predicted, a deletion of GDA1 results in the underglycosylation of proteins and lipids, although the null allele has no effect on growth.Many nucleotide sugar transport activities have been reported, which differ from one another in their substrate specificity and subcelluar localization (9Hirschberg C.B. Snider M.D. Annu. Rev. Biochem. 1987; 56: 63-87Crossref PubMed Scopus (439) Google Scholar). Since the cytoplasm is the sole site at which nucleotide sugars are synthesized, they must be transported into the various organelles in which glycosylation occurs. Mammalian cells require the transport of many different nucleotide sugars due to the diversity of carbohydrate processing in the Golgi. Carbohydrate chains may contain galactose, sialic acid, fucose, xylose,N-acetylglucosamine, and N-acetylgalactosamine. In contrast, in S. cerevisiae, glycosylation in the Golgi is largely restricted to mannosylation which in principle requires only a single transporter.The VRG4 gene is an essential gene that is required for a number of different Golgi-specific functions, includingN-linked glycosylation (10Ballou L. Hitzeman R.A. Lewis M.S. Ballou C.E. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 3209-3212Crossref PubMed Scopus (114) Google Scholar, 11Kanik-Enulat C. Montalvo E. Neff N. Genetics. 1995; 140: 933-943Crossref PubMed Google Scholar, 12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar), secretion, protein sorting, and the maintenance of a normal endomembrane system (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar). The data in this report demonstrate that the transport of GDP-mannose into the Golgi is the principal function of the VRG4 gene product. A simple system to assay GDP-mannose transport is described, using permeabilized yeast cells. With this system, we demonstrate thatvrg4 mutants are specifically defective in lumenal GDP-mannose transport in vitro. The protein sorting and membrane defects in vrg4 mutants may be explained by an indirect effect on sphingolipid mannosylation that normally occurs in the Golgi.DISCUSSIONThe transport of GDP-mannose into the lumen of the Golgi is requisite for glycosylation of both proteins and lipids in S. cerevisiae. The work presented here shows that the VRG4gene is required for this transport event and suggests thatVRG4 encodes the nucleotide sugar transporter. Our major observations and conclusions are as follows. (i) vrg4mutants accumulate under-mannosylated proteins and sphingolipidsin vivo. (ii) Membranes from the vrg4 mutant are specifically impaired in the lumenal uptake of GDP-[3H]mannose in vitro. Although these mutant membranes are drastically reduced in the ability to transport GDP-mannose, the activity of another Golgi enzyme, GDPase, is unaffected. (iii) VRG4 encodes a protein that is part of a large family of related proteins, some of which are known to function in nucleotide sugar transport. (iv) As predicted, Vrg4p is a resident of the Golgi complex. Although we cannot formally rule out the possibility that VRG4 encodes a regulator of the GDP-mannose transporter, the data demonstrate a pivotal role for Vrg4p in nucleotide sugar transport.Several putative nucleotide sugar transporters exhibit a high degree of similarity to Vrg4p (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar, 29Ma D. Russell D.G. Beverley S.M. Turco S.J. J. Biol. Chem. 1997; 272: 3799-3805Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar). These include the LeishmaniaLpg2p which also regulates Golgi GDP-mannose transport (29Ma D. Russell D.G. Beverley S.M. Turco S.J. J. Biol. Chem. 1997; 272: 3799-3805Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar) and theK. lactis Mnn2p which regulates UDP-GlcNAc transport (30Abeijon C. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5963-5968Crossref PubMed Scopus (110) Google Scholar,31Abeijon C. Mandon E.C. Robbins P.W. Hirschberg C.B. J. Biol. Chem. 1996; 271: 8851-8854Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Lpg2 is most similar to Vrg4, whereas Mnn2p is more related to another S. cerevisiae ORF, Yel004p (30Abeijon C. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5963-5968Crossref PubMed Scopus (110) Google Scholar). The similarity between Mnn2p and Vrg4p is somewhat surprising based upon their different substrate specificities. In K. lactis, glycoproteins are terminally modified by the addition ofN-acetylglucosamine and require the lumenal transport of UDP-GlcNAc in the Golgi. Glycoproteins in S. cerevisiae do not undergo this modification in the Golgi. Instead, they are modified entirely by mannose (see Ref. 2Ballou C.E. Methods Enzymol. 1990; 185: 440-470Crossref PubMed Scopus (271) Google Scholar). Therefore, the homology between Mnn2p and Vrg4p must include domains that are not involved in substrate specificity.The vrg4 mutation specifically affected MIPC and M(IP)2C biosynthesis. Coupled with the demonstration thatVRG4 is a resident Golgi protein, these data are consistent with the hypothesis that mannosylation of IPC to form MIPC and M(IP)2C is catalyzed by a glycosyltransferase that resides in the Golgi and utilizes GDP-mannose (24Puoti A. Desponds C. Conzelmann A. J. Cell Biol. 1991; 113: 515-525Crossref PubMed Scopus (77) Google Scholar). These results support the conclusions of Conzelmann and co-workers (24Puoti A. Desponds C. Conzelmann A. J. Cell Biol. 1991; 113: 515-525Crossref PubMed Scopus (77) Google Scholar), showing that the biosynthesis of MIPC and M(IP)2C is dependent on vesicular transport from the ER to Golgi, suggesting that these molecules are made in the Golgi.VRG4 is an essential gene that is pleiotropically required for a number of different Golgi functions, including secretion and the maintenance of normal membrane morphology. These observations led to the proposal that Vrg4p plays an important role in establishing or maintaining the organization of the Golgi (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar). The effect on sphingolipid biosynthesis may explain the pleiotropy associated with the vrg4 phenotype. Sphingolipids are essential membrane components. Although their biological role in yeast is still unclear, there is evidence that sphingolipids modulate the activity of the plasma membrane ATPase (32Patton J.L. Lester R.L. Arch. Biochem. Biophys. 1992; 292: 70-76Crossref PubMed Scopus (28) Google Scholar) and are involved in phospholipid biosynthesis (33Wu W.I. Lin Y.P. Wang E. Merrill Jr., A.H. Carman G.M. J. Biol. Chem. 1993; 268: 13830-13837Abstract Full Text PDF PubMed Google Scholar) and anchoring of cell-surface glycosylphosphatidylinositol-linked proteins (34Conzelmann A. Puoti A. Lester R.L. Desponds C. EMBO J. 1992; 11: 457-466Crossref PubMed Scopus (106) Google Scholar). Furthermore, the enrichment of sphingolipids in organelles of the secretory pathway (35Hechtberger P. Zinser E. Saf R. Hummel K. Paltauf F. Daum G. Eur. J. Biochem. 1994; 225: 641-649Crossref PubMed Scopus (103) Google Scholar) supports the notion that the processes of protein secretion and the intracellular trafficking of sphingolipids are linked processes.VRG4 may indirectly affect these processes by inhibiting the synthesis of MIPC and M(IP)2C in the Golgi. The lack of these mature forms may alter the structure and functional properties of membranes which in turn affects a range of biological functions, including secretion.What is the essential role of VRG4? Both N- andO-linked glycosylation are essential, but it appears that only modifications that occur in the ER are vital. For instance, mutations that block the transfer of the preassembled lipid-linked oligosaccharide onto the asparagine residues of proteins are lethal (Ref. 36te Heesen S. Janetsky B. Lehle L. Aebi M. EMBO J. 1992; 11: 2071-2075Crossref PubMed Scopus (118) Google Scholar and see Ref. 37Silberstein S. Gilmore R. FASEB J. 1996; 10: 849-858Crossref PubMed Scopus (206) Google Scholar for review). Similarly, mutants that completely lack the enzymes which catalyze the addition of the firstO-linked mannose in the ER are not viable (3Gentzsch M. Tanner W. EMBO J. 1996; 15: 5752-5759Crossref PubMed Scopus (217) Google Scholar). However, all of the available evidence suggests that the later modifications that occur in the Golgi are not essential. For example, no single mannosyltransferase that affects N- or O-linked sugar additions in the yeast Golgi is essential. In the case ofN-linked mannoses, a deletion of the OCH1 gene, which encodes the initiating mannosyltransferase for the elongation ofN-linked oligosaccharides, results in the accumulation of glycoproteins that completely lack an outer chain (38Nakanishi-Shindo Y. Nakayama K.I. Tanaka A. Toda Y. Jigami Y. J. Biol. Chem. 1993; 268: 26338-26345Abstract Full Text PDF PubMed Google Scholar). Despite this drastic effect on glycosylation, the och1Δ mutant is viable (39Nakayama K. Nagasu T. Shimma Y. Kuromitsu J. Jigami Y. EMBO J. 1992; 7: 2511-2519Crossref Scopus (242) Google Scholar). Similarly, proteins that affect the elongation ofO-linked mannose residues in the Golgi appear to be dispensable for viability (40Lussier M. Sdicu A.-M. Bussereau F. Jacquet M. Bussey H. J. Biol. Chem. 1997; 272: 15527-15531Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). Although these data suggest that mannosylation of proteins in the yeast Golgi does not appear to be essential, the strict requirement of VRG4 may indicate that an absolute loss of glycoprotein mannosylation in the Golgi, due to the complete loss of the substrate required for these modifications, leads to inviability. An alternative model is that the essential function of Vrg4p in yeast may simply involve its effect on sphingolipids. If this is so, MIPC and/or M(IP)2C, whose biosynthesis is blocked in the absence of Vrg4p are implicated as playing a critical role in growth and viability. The isolation of the transferase(s) that catalyzes the mannosylation of MIPC and M(IP)2C will allow this idea to be tested. The Golgi complex is the site where the terminal glycosylation of both proteins and lipids occurs. Unlike mammalian cells, in the yeastSaccharomyces cerevisiae, glycoproteins and sphingolipids are exclusively modified by the addition of mannose residues in the Golgi. Glycoproteins can undergo two types of modifications in which oligosaccharides are linked to either asparagine residues (N-linked) or serine/threonine residues (O-linked) (for review see Refs. 1Herscovics A. Orlean P. FASEB J. 1993; 7: 540-550Crossref PubMed Scopus (437) Google Scholar and 2Ballou C.E. Methods Enzymol. 1990; 185: 440-470Crossref PubMed Scopus (271) Google Scholar). Both of these glycosylation pathways initiate in the endoplasmic reticulum (ER) 1The abbreviations used are: ER, endoplasmic reticulum; PCR, polymerase chain reaction; IPC, inositol phosphorylceramides; MIPC, mannosylinositol phosphorylceramides; m(IP)2C, mannose(inositol phosphate)2 ceramide; HA, hemagglutinin; ORF, open reading frame; PYC, permeabilized yeast cells.1The abbreviations used are: ER, endoplasmic reticulum; PCR, polymerase chain reaction; IPC, inositol phosphorylceramides; MIPC, mannosylinositol phosphorylceramides; m(IP)2C, mannose(inositol phosphate)2 ceramide; HA, hemagglutinin; ORF, open reading frame; PYC, permeabilized yeast cells. and terminate in the Golgi. After transport of the protein to the Golgi, mostN-linked oligosaccharides are elongated by a series of different mannosyltransferases to form glycoproteins that contain outer chains of 50 or more mannose residues. The α1,6-linked outer chain is highly branched with α1,2- and α1,3-linked mannoses. As in higher eukaryotes, it appears that the various mannosyltransferases that catalyze these sequential reactions are compartmentalized from one another within the individual Golgi cisternae. In the case ofO-linked sugars, up to five mannoses are added after the addition of the first mannose in the ER (3Gentzsch M. Tanner W. EMBO J. 1996; 15: 5752-5759Crossref PubMed Scopus (217) Google Scholar, 4Gentzsch M. Tanner W. Glycobiology. 1997; 7: 481-486Crossref PubMed Scopus (140) Google Scholar). The phosphoinositol-containing sphingolipids in yeast also undergo mannosylation in the Golgi. In S. cerevisiae, there are three major classes of sphingolipids. These include the inositol phosphorylceramides (IPCs) and the mannosylinositol phosphorylceramides (MIPC and M(IP)2C) (for review, see Ref. 5Lester R.L. Dickson R.C. Adv. Lipid Res. 1993; 26: 253-274PubMed Google Scholar). MIPC and M(IP)2C contain a single mannose attached to the inositol (6Smith S.W. Lester R.L. J. Biol. Chem. 1974; 249: 3395-3405Abstract Full Text PDF PubMed Google Scholar), although little is known about the mannosyltransferase(s) that catalyzes this reaction. The mannosyl donor for all of these Golgi-localized reactions is the nucleotide sugar GDP-mannose, whose site of synthesis is the cytoplasm. Before it can be utilized by the different lumenal mannosyltransferases, GDP-mannose must be transported into the Golgi by a specific nucleotide sugar transporter (7Abeijon C. Orlean P. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 6935-6939Crossref PubMed Scopus (131) Google Scholar). Once the sugar is donated to lumenal mannosyltransferase acceptors, the nucleoside diphosphate GDP is converted to a monophosphate by a nucleoside diphosphatase (7Abeijon C. Orlean P. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 6935-6939Crossref PubMed Scopus (131) Google Scholar). As in the mammalian Golgi, the transport of the nucleotide sugar into the lumen is coupled to the outward exit of the monophosphate in yeast. The yeast GDPase-encoding gene, GDA1, has been isolated (8Abeijon C. Yanagisawa K. Mandon E.C. Hausler A. Moremen K. Hirschberg C.B. Robbins P.W. J. Cell Biol. 1993; 122: 307-323Crossref PubMed Scopus (158) Google Scholar). As predicted, a deletion of GDA1 results in the underglycosylation of proteins and lipids, although the null allele has no effect on growth. Many nucleotide sugar transport activities have been reported, which differ from one another in their substrate specificity and subcelluar localization (9Hirschberg C.B. Snider M.D. Annu. Rev. Biochem. 1987; 56: 63-87Crossref PubMed Scopus (439) Google Scholar). Since the cytoplasm is the sole site at which nucleotide sugars are synthesized, they must be transported into the various organelles in which glycosylation occurs. Mammalian cells require the transport of many different nucleotide sugars due to the diversity of carbohydrate processing in the Golgi. Carbohydrate chains may contain galactose, sialic acid, fucose, xylose,N-acetylglucosamine, and N-acetylgalactosamine. In contrast, in S. cerevisiae, glycosylation in the Golgi is largely restricted to mannosylation which in principle requires only a single transporter. The VRG4 gene is an essential gene that is required for a number of different Golgi-specific functions, includingN-linked glycosylation (10Ballou L. Hitzeman R.A. Lewis M.S. Ballou C.E. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 3209-3212Crossref PubMed Scopus (114) Google Scholar, 11Kanik-Enulat C. Montalvo E. Neff N. Genetics. 1995; 140: 933-943Crossref PubMed Google Scholar, 12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar), secretion, protein sorting, and the maintenance of a normal endomembrane system (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar). The data in this report demonstrate that the transport of GDP-mannose into the Golgi is the principal function of the VRG4 gene product. A simple system to assay GDP-mannose transport is described, using permeabilized yeast cells. With this system, we demonstrate thatvrg4 mutants are specifically defective in lumenal GDP-mannose transport in vitro. The protein sorting and membrane defects in vrg4 mutants may be explained by an indirect effect on sphingolipid mannosylation that normally occurs in the Golgi. DISCUSSIONThe transport of GDP-mannose into the lumen of the Golgi is requisite for glycosylation of both proteins and lipids in S. cerevisiae. The work presented here shows that the VRG4gene is required for this transport event and suggests thatVRG4 encodes the nucleotide sugar transporter. Our major observations and conclusions are as follows. (i) vrg4mutants accumulate under-mannosylated proteins and sphingolipidsin vivo. (ii) Membranes from the vrg4 mutant are specifically impaired in the lumenal uptake of GDP-[3H]mannose in vitro. Although these mutant membranes are drastically reduced in the ability to transport GDP-mannose, the activity of another Golgi enzyme, GDPase, is unaffected. (iii) VRG4 encodes a protein that is part of a large family of related proteins, some of which are known to function in nucleotide sugar transport. (iv) As predicted, Vrg4p is a resident of the Golgi complex. Although we cannot formally rule out the possibility that VRG4 encodes a regulator of the GDP-mannose transporter, the data demonstrate a pivotal role for Vrg4p in nucleotide sugar transport.Several putative nucleotide sugar transporters exhibit a high degree of similarity to Vrg4p (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar, 29Ma D. Russell D.G. Beverley S.M. Turco S.J. J. Biol. Chem. 1997; 272: 3799-3805Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar). These include the LeishmaniaLpg2p which also regulates Golgi GDP-mannose transport (29Ma D. Russell D.G. Beverley S.M. Turco S.J. J. Biol. Chem. 1997; 272: 3799-3805Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar) and theK. lactis Mnn2p which regulates UDP-GlcNAc transport (30Abeijon C. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5963-5968Crossref PubMed Scopus (110) Google Scholar,31Abeijon C. Mandon E.C. Robbins P.W. Hirschberg C.B. J. Biol. Chem. 1996; 271: 8851-8854Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Lpg2 is most similar to Vrg4, whereas Mnn2p is more related to another S. cerevisiae ORF, Yel004p (30Abeijon C. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5963-5968Crossref PubMed Scopus (110) Google Scholar). The similarity between Mnn2p and Vrg4p is somewhat surprising based upon their different substrate specificities. In K. lactis, glycoproteins are terminally modified by the addition ofN-acetylglucosamine and require the lumenal transport of UDP-GlcNAc in the Golgi. Glycoproteins in S. cerevisiae do not undergo this modification in the Golgi. Instead, they are modified entirely by mannose (see Ref. 2Ballou C.E. Methods Enzymol. 1990; 185: 440-470Crossref PubMed Scopus (271) Google Scholar). Therefore, the homology between Mnn2p and Vrg4p must include domains that are not involved in substrate specificity.The vrg4 mutation specifically affected MIPC and M(IP)2C biosynthesis. Coupled with the demonstration thatVRG4 is a resident Golgi protein, these data are consistent with the hypothesis that mannosylation of IPC to form MIPC and M(IP)2C is catalyzed by a glycosyltransferase that resides in the Golgi and utilizes GDP-mannose (24Puoti A. Desponds C. Conzelmann A. J. Cell Biol. 1991; 113: 515-525Crossref PubMed Scopus (77) Google Scholar). These results support the conclusions of Conzelmann and co-workers (24Puoti A. Desponds C. Conzelmann A. J. Cell Biol. 1991; 113: 515-525Crossref PubMed Scopus (77) Google Scholar), showing that the biosynthesis of MIPC and M(IP)2C is dependent on vesicular transport from the ER to Golgi, suggesting that these molecules are made in the Golgi.VRG4 is an essential gene that is pleiotropically required for a number of different Golgi functions, including secretion and the maintenance of normal membrane morphology. These observations led to the proposal that Vrg4p plays an important role in establishing or maintaining the organization of the Golgi (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar). The effect on sphingolipid biosynthesis may explain the pleiotropy associated with the vrg4 phenotype. Sphingolipids are essential membrane components. Although their biological role in yeast is still unclear, there is evidence that sphingolipids modulate the activity of the plasma membrane ATPase (32Patton J.L. Lester R.L. Arch. Biochem. Biophys. 1992; 292: 70-76Crossref PubMed Scopus (28) Google Scholar) and are involved in phospholipid biosynthesis (33Wu W.I. Lin Y.P. Wang E. Merrill Jr., A.H. Carman G.M. J. Biol. Chem. 1993; 268: 13830-13837Abstract Full Text PDF PubMed Google Scholar) and anchoring of cell-surface glycosylphosphatidylinositol-linked proteins (34Conzelmann A. Puoti A. Lester R.L. Desponds C. EMBO J. 1992; 11: 457-466Crossref PubMed Scopus (106) Google Scholar). Furthermore, the enrichment of sphingolipids in organelles of the secretory pathway (35Hechtberger P. Zinser E. Saf R. Hummel K. Paltauf F. Daum G. Eur. J. Biochem. 1994; 225: 641-649Crossref PubMed Scopus (103) Google Scholar) supports the notion that the processes of protein secretion and the intracellular trafficking of sphingolipids are linked processes.VRG4 may indirectly affect these processes by inhibiting the synthesis of MIPC and M(IP)2C in the Golgi. The lack of these mature forms may alter the structure and functional properties of membranes which in turn affects a range of biological functions, including secretion.What is the essential role of VRG4? Both N- andO-linked glycosylation are essential, but it appears that only modifications that occur in the ER are vital. For instance, mutations that block the transfer of the preassembled lipid-linked oligosaccharide onto the asparagine residues of proteins are lethal (Ref. 36te Heesen S. Janetsky B. Lehle L. Aebi M. EMBO J. 1992; 11: 2071-2075Crossref PubMed Scopus (118) Google Scholar and see Ref. 37Silberstein S. Gilmore R. FASEB J. 1996; 10: 849-858Crossref PubMed Scopus (206) Google Scholar for review). Similarly, mutants that completely lack the enzymes which catalyze the addition of the firstO-linked mannose in the ER are not viable (3Gentzsch M. Tanner W. EMBO J. 1996; 15: 5752-5759Crossref PubMed Scopus (217) Google Scholar). However, all of the available evidence suggests that the later modifications that occur in the Golgi are not essential. For example, no single mannosyltransferase that affects N- or O-linked sugar additions in the yeast Golgi is essential. In the case ofN-linked mannoses, a deletion of the OCH1 gene, which encodes the initiating mannosyltransferase for the elongation ofN-linked oligosaccharides, results in the accumulation of glycoproteins that completely lack an outer chain (38Nakanishi-Shindo Y. Nakayama K.I. Tanaka A. Toda Y. Jigami Y. J. Biol. Chem. 1993; 268: 26338-26345Abstract Full Text PDF PubMed Google Scholar). Despite this drastic effect on glycosylation, the och1Δ mutant is viable (39Nakayama K. Nagasu T. Shimma Y. Kuromitsu J. Jigami Y. EMBO J. 1992; 7: 2511-2519Crossref Scopus (242) Google Scholar). Similarly, proteins that affect the elongation ofO-linked mannose residues in the Golgi appear to be dispensable for viability (40Lussier M. Sdicu A.-M. Bussereau F. Jacquet M. Bussey H. J. Biol. Chem. 1997; 272: 15527-15531Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). Although these data suggest that mannosylation of proteins in the yeast Golgi does not appear to be essential, the strict requirement of VRG4 may indicate that an absolute loss of glycoprotein mannosylation in the Golgi, due to the complete loss of the substrate required for these modifications, leads to inviability. An alternative model is that the essential function of Vrg4p in yeast may simply involve its effect on sphingolipids. If this is so, MIPC and/or M(IP)2C, whose biosynthesis is blocked in the absence of Vrg4p are implicated as playing a critical role in growth and viability. The isolation of the transferase(s) that catalyzes the mannosylation of MIPC and M(IP)2C will allow this idea to be tested. The transport of GDP-mannose into the lumen of the Golgi is requisite for glycosylation of both proteins and lipids in S. cerevisiae. The work presented here shows that the VRG4gene is required for this transport event and suggests thatVRG4 encodes the nucleotide sugar transporter. Our major observations and conclusions are as follows. (i) vrg4mutants accumulate under-mannosylated proteins and sphingolipidsin vivo. (ii) Membranes from the vrg4 mutant are specifically impaired in the lumenal uptake of GDP-[3H]mannose in vitro. Although these mutant membranes are drastically reduced in the ability to transport GDP-mannose, the activity of another Golgi enzyme, GDPase, is unaffected. (iii) VRG4 encodes a protein that is part of a large family of related proteins, some of which are known to function in nucleotide sugar transport. (iv) As predicted, Vrg4p is a resident of the Golgi complex. Although we cannot formally rule out the possibility that VRG4 encodes a regulator of the GDP-mannose transporter, the data demonstrate a pivotal role for Vrg4p in nucleotide sugar transport. Several putative nucleotide sugar transporters exhibit a high degree of similarity to Vrg4p (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar, 29Ma D. Russell D.G. Beverley S.M. Turco S.J. J. Biol. Chem. 1997; 272: 3799-3805Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar). These include the LeishmaniaLpg2p which also regulates Golgi GDP-mannose transport (29Ma D. Russell D.G. Beverley S.M. Turco S.J. J. Biol. Chem. 1997; 272: 3799-3805Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar) and theK. lactis Mnn2p which regulates UDP-GlcNAc transport (30Abeijon C. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5963-5968Crossref PubMed Scopus (110) Google Scholar,31Abeijon C. Mandon E.C. Robbins P.W. Hirschberg C.B. J. Biol. Chem. 1996; 271: 8851-8854Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Lpg2 is most similar to Vrg4, whereas Mnn2p is more related to another S. cerevisiae ORF, Yel004p (30Abeijon C. Robbins P.W. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5963-5968Crossref PubMed Scopus (110) Google Scholar). The similarity between Mnn2p and Vrg4p is somewhat surprising based upon their different substrate specificities. In K. lactis, glycoproteins are terminally modified by the addition ofN-acetylglucosamine and require the lumenal transport of UDP-GlcNAc in the Golgi. Glycoproteins in S. cerevisiae do not undergo this modification in the Golgi. Instead, they are modified entirely by mannose (see Ref. 2Ballou C.E. Methods Enzymol. 1990; 185: 440-470Crossref PubMed Scopus (271) Google Scholar). Therefore, the homology between Mnn2p and Vrg4p must include domains that are not involved in substrate specificity. The vrg4 mutation specifically affected MIPC and M(IP)2C biosynthesis. Coupled with the demonstration thatVRG4 is a resident Golgi protein, these data are consistent with the hypothesis that mannosylation of IPC to form MIPC and M(IP)2C is catalyzed by a glycosyltransferase that resides in the Golgi and utilizes GDP-mannose (24Puoti A. Desponds C. Conzelmann A. J. Cell Biol. 1991; 113: 515-525Crossref PubMed Scopus (77) Google Scholar). These results support the conclusions of Conzelmann and co-workers (24Puoti A. Desponds C. Conzelmann A. J. Cell Biol. 1991; 113: 515-525Crossref PubMed Scopus (77) Google Scholar), showing that the biosynthesis of MIPC and M(IP)2C is dependent on vesicular transport from the ER to Golgi, suggesting that these molecules are made in the Golgi. VRG4 is an essential gene that is pleiotropically required for a number of different Golgi functions, including secretion and the maintenance of normal membrane morphology. These observations led to the proposal that Vrg4p plays an important role in establishing or maintaining the organization of the Golgi (12Poster J.B. Dean N. J. Biol. Chem. 1996; 271: 3837-3845Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar). The effect on sphingolipid biosynthesis may explain the pleiotropy associated with the vrg4 phenotype. Sphingolipids are essential membrane components. Although their biological role in yeast is still unclear, there is evidence that sphingolipids modulate the activity of the plasma membrane ATPase (32Patton J.L. Lester R.L. Arch. Biochem. Biophys. 1992; 292: 70-76Crossref PubMed Scopus (28) Google Scholar) and are involved in phospholipid biosynthesis (33Wu W.I. Lin Y.P. Wang E. Merrill Jr., A.H. Carman G.M. J. Biol. Chem. 1993; 268: 13830-13837Abstract Full Text PDF PubMed Google Scholar) and anchoring of cell-surface glycosylphosphatidylinositol-linked proteins (34Conzelmann A. Puoti A. Lester R.L. Desponds C. EMBO J. 1992; 11: 457-466Crossref PubMed Scopus (106) Google Scholar). Furthermore, the enrichment of sphingolipids in organelles of the secretory pathway (35Hechtberger P. Zinser E. Saf R. Hummel K. Paltauf F. Daum G. Eur. J. Biochem. 1994; 225: 641-649Crossref PubMed Scopus (103) Google Scholar) supports the notion that the processes of protein secretion and the intracellular trafficking of sphingolipids are linked processes.VRG4 may indirectly affect these processes by inhibiting the synthesis of MIPC and M(IP)2C in the Golgi. The lack of these mature forms may alter the structure and functional properties of membranes which in turn affects a range of biological functions, including secretion. What is the essential role of VRG4? Both N- andO-linked glycosylation are essential, but it appears that only modifications that occur in the ER are vital. For instance, mutations that block the transfer of the preassembled lipid-linked oligosaccharide onto the asparagine residues of proteins are lethal (Ref. 36te Heesen S. Janetsky B. Lehle L. Aebi M. EMBO J. 1992; 11: 2071-2075Crossref PubMed Scopus (118) Google Scholar and see Ref. 37Silberstein S. Gilmore R. FASEB J. 1996; 10: 849-858Crossref PubMed Scopus (206) Google Scholar for review). Similarly, mutants that completely lack the enzymes which catalyze the addition of the firstO-linked mannose in the ER are not viable (3Gentzsch M. Tanner W. EMBO J. 1996; 15: 5752-5759Crossref PubMed Scopus (217) Google Scholar). However, all of the available evidence suggests that the later modifications that occur in the Golgi are not essential. For example, no single mannosyltransferase that affects N- or O-linked sugar additions in the yeast Golgi is essential. In the case ofN-linked mannoses, a deletion of the OCH1 gene, which encodes the initiating mannosyltransferase for the elongation ofN-linked oligosaccharides, results in the accumulation of glycoproteins that completely lack an outer chain (38Nakanishi-Shindo Y. Nakayama K.I. Tanaka A. Toda Y. Jigami Y. J. Biol. Chem. 1993; 268: 26338-26345Abstract Full Text PDF PubMed Google Scholar). Despite this drastic effect on glycosylation, the och1Δ mutant is viable (39Nakayama K. Nagasu T. Shimma Y. Kuromitsu J. Jigami Y. EMBO J. 1992; 7: 2511-2519Crossref Scopus (242) Google Scholar). Similarly, proteins that affect the elongation ofO-linked mannose residues in the Golgi appear to be dispensable for viability (40Lussier M. Sdicu A.-M. Bussereau F. Jacquet M. Bussey H. J. Biol. Chem. 1997; 272: 15527-15531Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). Although these data suggest that mannosylation of proteins in the yeast Golgi does not appear to be essential, the strict requirement of VRG4 may indicate that an absolute loss of glycoprotein mannosylation in the Golgi, due to the complete loss of the substrate required for these modifications, leads to inviability. An alternative model is that the essential function of Vrg4p in yeast may simply involve its effect on sphingolipids. If this is so, MIPC and/or M(IP)2C, whose biosynthesis is blocked in the absence of Vrg4p are implicated as playing a critical role in growth and viability. The isolation of the transferase(s) that catalyzes the mannosylation of MIPC and M(IP)2C will allow this idea to be tested. We thank Debbie Brown for help and advice on lipid analyses and Nancy Hollingsworth and members of the Dean lab for critical reading of the manuscript.