The Yptl GTPase Is Essential for the First Two Steps of the Yeast Secretory Pathway

Small GTPases of the rab family are involved in the regulation of vesicular transport. The restricted distribution of each of these proteins in mammalian cells has led to the suggestion that different rab proteins act at different steps of t ransport (Pryer, N. K., L. J. Wuestehube, and R. Scheckman. 1992. Annu Rev. Biochem. 61:471-516; Zerial, M., and H. Stenmark. 1993. Curr. Opin. Cell Biol. 5:613-620). However, in this report we show that the Ypt l -GTPase , a member of the rab family, is essential for more than one step of the yeast secretory pathway. We determined the secretory defect conferred by a novel yptl mutation by comparing the processing of several transported glycoproteins in wild-type and mutant cells. The ypt1-A136D mutant has a change in an amino acid that is conserved among rab GTPases. This mutation leads to a rapid and tight secretory block upon a shift to the restrictive temperature, and allows for the identification of the specific steps in the secretory pathway that directly require Ypt l protein (Yptlp) . The ypt1-A136D mutant exhibits tight blocks in two secretory steps, E R to cis-Golgi and cisto medial-Golgi, but later steps are unaffected. Thus, it is unlikely that Yp t lp functions as the sole determinant of fusion specificity. Our results are more consistent with a role for Ypt l / rab proteins in determining the directionality or fidelity of protein sorting. T RANSPORX of proteins in eukaryotic cells involves their orderly progression through a series of membranous compartments (Palade, 1975). The different steps of this transport pathway appear to be mediated by vesicles that bud from one compartment and fuse with the next (Pfeffer and Rothman, 1987). The mechanisms that control directionality and specificity of the different steps are not known. A large number of GTPases belonging to the rab/Yptl/Sec4 family have been shown to play a role in vesicular trafficking in yeast and mammalian ceils. Because different rab proteins have distinct subcellular distributions, it has been suggested that each rab protein acts at a specific step in transport through the exocytic, endocytic, or transcytotic pathway (Pfeffer, 1992; Ferro-Novick and Novick, 1993; Zerial and Stenmark, 1993). Current models for the determination of vesicle targeting specificity propose that a vesicle-associated rab protein interacts with a vesicle-associated SNAP receptor (v-SNARE) specificity component (Novick and Brennwald, 1993; Rothman, 1994). Such a mechanism would imply that a given tab protein acts at only one step rather than at multiple Address correspondence to Nava Segev, Dept. of Pharmacological and Physiological Sciences, The University of Chicago, 947 East 58th Street, Box 271, Chicago, IL 60637. Tel.: (312) 702-3526. Fax: (312) 702-3774. 1. Abbreviat ions used in this paper: EF-Tu, elongation factor Tu; GAP, GTPase activating protein; SD, synthetic dextrose; v-SNARE, vesicleassociated SNAP receptor; YPD, yeast extract, peptone, and dextrose. steps of transport, a question that has not been tested experimentally. Resolution of this issue will help elucidate the functions of tab proteins. Two alternative models have been proposed for the mechanism of action of rab GTPases in vesicular transport. First, they might serve as specificity determinants for each step, either alone or in conjunction with the SNAREs (Novick and Brennwald, 1993; Zerial and Stenmark, 1993; Rothman, 1994). Alternatively, tab proteins might regulate the assembly of docking and/or fusion complexes that link vesicles with their target membranes (Sogaard et al., 1994). An argument against the first model was the finding that an artificial tab protein can function in more than one step of the yeast secretory pathway. Thus, a Yptlp-Sec4p chimera was shown to function as both Sec4p and a minimal Yptlp (it can complement a yptl-deletion, but the cells are heat and cold sensitive; Brennwald and Novick, 1992; Dunn et al., 1993). The bifunctionality of this artificial GTPase might mean that tab proteins do not, by themselves, function as specificity determinants (Ferro-Novick and Jahn, 1994). However, it is conceivable that Yptlp and Sec4p do function as specificity factors, but the specificitybearing domains are in different portions of the two proteins, so that the artificial fusion protein contains both domains. We therefore did a complementary experiment by testing whether a single rab GTPase functions in vivo at more than one step in the secretory pathway. The Yptl GTPase has been demonstrated to function © The Rockefeller University Press, 0021-9525/95/11/583/8 $2.00 The Journal of Cell Biology, Volume 131, Number 3, November 1995 583-590 583 on N ovem er 9, 2017 jcb.rress.org D ow nladed fom early in the yeast secretory pathway, and this protein has been localized to the yeast ER and Golgi complex (Segev et al., 1988; Preuss et al., 1992). Although Yptlp appears to function after vesicle formation in the targeting of vesicles to an acceptor compartment (Rexach et al., 1991; Segev, 1991), it is not clear whether Yptlp functions at more than one transport step. The role of Yptlp in the first step of the pathway, ER to Golgi, is well established since antibodies against Yptlp completely inhibit this step in a cellfree system (Baker et al., 1990). However, in vivo and in vitro studies of yptl mutants have been inconclusive, suggesting that Yptlp might act in ER to Golgi transport or in transport between early Golgi compartments (Schmitt et al., 1988; Segev et al., 1988; Bacon et al., 1989). Although the mutant studies implicated Yptlp in each of these steps, Yptlp did not appear to be required because the secretory blocks were not tight. These results were inconclusive for two reasons. First, the three yptl mutants used in previous studies (yptl-1, yptl-2, and ypfS) exhibit severe secretory defects even under permissive conditions, and/or the blocks are either leaky or slow to take effect. These factors complicate the analysis due to possible indirect effects that are unrelated to the specific processes controlled by Yptlp. Second, previous studies of the role of Yptlp in glycoprotein transport did not examine the nature of the carbohydrate modifications, which are indicative of passage through particular secretory compartments. Rather, the analyses were based on the electrophoretic mobility of the marker proteins, and, as we discuss in this report, such analyses can be misleading. We wished to address two questions concerning the role of Yptlp in the yeast secretory pathway: first, is there a primary intra-Golgi defect in yptl mutant cells? Second, if there is such a defect, which of the intra-Golgi steps are affected? In this study, we determined the steps in which Yptlp functions in vivo using a novel yptl mutant whose phenotype does not exhibit the complications discussed above. This mutant, yptl-A136D, exhibits a rapid and tight block of secretion upon shifting to the nonpermissive temperature. We determined the nature of the secretory block by monitoring the specific modifications acquired by transported glycoproteins in these mutant cells. Our results demonstrated that the yptl-A136D mutation confers a block in the first two steps of the secretory pathway (i.e., ER to cis-Golgi and cisto medial-Golgi). Thus, it appears that rab GTPases do not define the specificity of membrane trafficking, but instead perform an essential regulatory function in vesicular transport. Materials and Methods Cells, Mutants, and Plasmids The yptl-1 allele was amplified by PCR from genomic DNA using the upstream primer 5 ' -GGGCC CGCAT GCGCA CCAGT TI 'TGA GGAGG-3' and the downstream primer 5 ' -GGGCC CGGAT CCGAT AAG A A A G A A T G-3'. PCR products were inserted into pGEM3Zf ( ) (Promega Corp., Madison, WI), and 20 insert-containing clones were pooled and sequenced. The yptl-A136D allele was made by site-directed mutagenesis using the method of Kunkel (Kunkel, 1987) in the Escherichia coli phagemid vector pRS300 that contained the 767-bp fragment from EcoRI to BamHI of YPTI inserted into pIBI31 (International Biotechnologies, Inc., New Haven, CT). The mutagenic oligonucleotide was: 5 ' -GTG G A A TAT GAC GTC GAC A A G G A A TTT GCG GAC-3'. DNA modifying enzymes were from Boehringer Mannheim Corp. (Indianapolis, IN). The mutation was confirmed by sequencing with the Sequenase sequencing kit (United States Biochemical Corp., Cleveland, OH). The ypt1-A136D mutation was targeted to the chromosome as previously described (Segev and Botstein, 1987). Two pairs of Saccharomyces cerevisiae strains were used in this study: (a) wild-type, DBY1034; and yptlT4OK, DBY1803 (Segev and Botstein, 1987); (b) wild-type, NSY160: MATa his4-539(am) lys2-8Ol(am) ura3-52; and yptl-A136D, NSY161: MATa his4-539(am) ura3-52 (this study). Yeast strains were grown on yeast extract, peptone and dextrose (YPD) or synthetic dextrose (SD) medium containing the appropriate nutritional supplements (Sherman et al.,