rab gtpases in vesicular transport

Rab GTPases in vesicular transport

Introduction

Marino Zerial and Harald Stenmark

European Molecular Biology Laboratory, Heidelberg, Germany

Specificity and directionality are two features shared by the numerous steps of membrane transport that connect cellular organelles. By shuttling between specific membrane compartments and the cytoplasm, small GTPases of the Rab family appear to regulate membrane traffic in a cyclical manner. The restriction of certain Rab proteins to differentiated cell types supports a role for these GTPases in defining the specificity of

membrane trafficking.

Current Opinion in Cell Biology 1993, 5~613-620

Eukaryotic cells are divided into distinct membrane- bound organelles. While this functional compartmen- talization has the advantage of both efficiency and regulation, it also requires controlled interactions between the various organelles. Exchange of constituents between organefles is thought to occur through vesicular or tubular intermediates that selectively deliver proteins and lipids from one compartment to another [ 10-3.1. Specific molecules control this complex trafficking network. Over the past few years, a large number of small GTPases of the Rab family have been described and localized to various intracellular compartments. By alternating between two distinct conformations in a nucleotide-dependent fashion, these proteins function as molecular switches. Through this mechanism they have been proposed to ensure fidelity in the process of docking and/or fusion of vesicles with their correct acceptor compartment [4]. The cycle of GTP binding and hydrolysis would allow multiple rounds of vesicular transport and confer vec- torial@. This raiew focuses on the involvement of Rab proteins in individual trafficking routes between intracellular compartments.

Function of Rab proteins in membrane traffic

The role of Rab GTRases in membrane traffic was first demonstrated by studies on Sec4p and Yptlp, which function in the exocytic pathway of Saccbaromyces cere ~r&ae [2*]. Rab proteins, the mammalian counterparts of yeast Sec4p and Yptlp [ 51, are localized to distinct intracellular compartments (Table 1) and a combination of in vitro and in vivo studies has emphasized their role in the regulation of different exocytic and endocytic transport processes.

In the exocytic pathway, in vitro studies have shown that the Rablb protein is involved in endoplasmic reticulum (ER) to Golgi transport [6] and work in intact cells has indicated that this transport process requires also Rabla and Rab2 proteins [ 71. The participation of three different Rab proteins in ERGolgi transport might reflect se- quential steps of vesicular transport or, alternatively, the concerted function of several Rab GTPases at the same step. Rab proteins also function in regulated secretion. Synthetic peptides corresponding to the effector region of Rab3a stimulate regulated exocytosis [S-lo].

In the endocytic pathway, both Rab4 and Rab5 proteins are associated with early endosomes (Rab5 is also on the plasma membrane) [ 11,121. However, they have distinct activities. An increase in the level of Rab5 stimulates endocytosis and expands the size of the early endosomes, whereas expression of a Rab5 GTP-binding defective mutant has the opposite effect [13*]. These morphological alterations are consistent with a role of Rab5 in the fusion between early endosomes in vitro [ 141. In contrast, overexpression of Rab4 increases the re&x of endocytic markers to the cell surface and induces the accumulation of endocytic tubules and tubule clusters that could correspond either to structurally altered early endosomes or even to a distinct recycling compartment [IS*].

Another example of distinct function of Rab proteins in the same organelle is provided by Rab7 and Rab9. Both proteins have been local@ed to late endosomes [12,16], but only Rab9 spectically functions in transport from this compartment to the tram&olgi network (TGN) in vitro [16]. While the role of mammalian Rab7 remains unknown, functional studies on a structural homologue in S. cerevzs&e, Ypt7p, suggest that this protein may function at a late step of the endocytic pathway [17’]. Ypt7 null mutants are viable and exhibit normal exocytosis and u-factor internalization, but a-factor degradation is inhibited and the vacuole is fragmented. By analogy

Abbreviations CHM-choroideremia; DC&-dominant suppressor of Set+ ER-endoplasmic reticulum; CAP-GTPase activating protein; CDL-GDP dissociation inhibitor; W-mammalian suppressor of Sec4; NSF-N-ethylmaleimide-sensitive fusion protein;

SNAP-soluble NSF attachment protein; SNARE-SNAP receptor; TCN-trans-Colgi network.

@ Current Biology Ltd ISSN 0955+674 613

614 Membranes .

Table 1. Summary of the localization, functional properties and interacting components of Yptl, Sec4 and Rab proteins in yeast and mammalian cells.

Protein Localization Function Interacting components

Yptl (yeast) Sec4

dabla

Rabl b

Rab2

Rab3a

RaMa

Rab5a

Rab6 Ypt7

ER-Golgi

Secretory vesicles- plasma membrane

ER-Golgi

ER-Golgi intermediate compartment

Synaptic vesicles Chromaffin granules

Early endosomes

Plasma membrane, clathrin- coated vesicles, early endosomes

Middle Colgi-TGN Not determined

Fusion of post-ER vesicles with plasma membrane

Fusion of post-Golgi vesicles with plasma membrane

ER-Golgi transport in yeast and mammalian cells ER-cis-Colgi and c&medial

Colgi transport ER-Goigi transport

DSSC, MSSC

Regulated exocytosis in pancreatic acinar cells, adrenal chromaffin cells

and mast cells Early endosomes-plasma membrane

recycling pathway Plasma membrane-early endosome

transport and homotypic fusion between early endosomes

Not determined Transport in endocytic pathway

Rab7 Rab8

Rab9

Rab17

Late endosomes Post-Golgi basolateral

secretory vesicles Late endosomes, TGN

Apical dense tubules, basolateral plasma membrane

Not determined Golgi-plasma membrane

transport Transport from late endosomes

to TGN Not determined

Rabphilin-3A

‘Regulatory factors active on different Rab proteins. Rab GDI is not listed among the interacting components, as it interacts with

most Rab proteins. See text for references.

with mammalian endocytic pathway organization and localization of Rab7 to late endosomes, Ypt7p is thought to control transport between early and late endosomal compartments in yeast [ 17’1. This interpretation is consistent with the existence of kinetically and biochemically distinct endqtic organelles in yeast [ 181. Certainly, these observations suggest that the yeast endocytic machinery shares features with the mammalian endocytic apparatus. A vesicle transport process can be typically divided into several steps: budding, uncoating, docking and fusion [1~3*]. While Rab proteins are implicated in the regulation of distinct membrane traffic events, the exact step controlled by each protein is not clear. Both Sedp and Rab5 appear to control vesicle docking/fusion. However, while yeast Yptlp is thought to function in the same step [2*], its mammalian-related protein, Rablb, is thought to be essential for vesicle formation [ 61. The interpretation of in vivo studies is complicated by the fact that transport between compartments may occur in both di- rections, forwards and backwards, and affecting one di- rection is likely to influence the other. Overexpression

of Rab5 stimulates endocytosis but also accelerates recycling from the early endosomes to the plasma membrane [WI. Likewise, inhibition of retrograde transport could, in principle, indirectly affect anterograde move- ment between the ER and the Golgi, and vice versa Clearly, these studies have shown that changes in the balance between vesicles coming from and returning to a compartment have profound effects on organelle organization. Similar effects have also been observed with other components of the traf3cking machinery. For example, the sbi6ire mutation in a dynamin-related protein of Drosophila mehnogaster is thought to block coated vesicle budding from the plasma membrane [W-21 1. This causes an increase in its surface area, an accumulation of elongated coated pits and a decrease in the number of endocytic tubular elements [ 191. Thus, similar to the overexpression of Rab4 and Rab5, the mutation induces drastic morphological changes. Altogether, these biochemical and morphological studies only represent a first step towards a more detailed dissection of Rab protein function.

Rab CTPases in vesicular transport Zerial and Stenmark 615

Complexity of the Rab protein family

The large number of Rab GTPases identiiied in mammalian cells (30 different members) 122-241 led to the hypothesis that each step of vesicular traffic is regulated by at least one Rab protein [4]. Nevertheless, the implica- tions of this complexity are not clear. First, the picture is complicated by the existence of subgroups of Rab proteins sharing high sequence identity (e.g. Rabla, Rablb, Rab3a, Rab3b, Rab3c, Rab3d). The high sequence conservation suggests that proteins belonging to one subgroup may have a similar function, as supported by the finding that both Rabla and Rablb regulate ER-Golgi transport [7]. However, this simple interpretation is countered by the fact that members of a subgroup may have different biochemical properties: both the exocytlc Rabla and the endocytlc RaMa proteins contain a phosphotylation site for p34u kinase [25,26*] whereas Rablb and Rab4b do not [27]. Anti-Rab5a specific antibodies block early endosome fusion in vitro, despite the presence of the Rab5b and Rab5c proteins [28]. Different members of the Rab3 subgroup carry out distinct functions (see be- low). Altogether, these differences question the idea that Rab isofonns are functionally redundant. Instead, the di- versity might be responsible for hne tuning intracellular transport by the coordinated activity of several Rab iso- forms. Second, recent morphological data indicate that several Rab proteins localize to the same organelle. Early endosomes provide a striking example: they contain at least three additional Rab proteins besides the Rab4 and Rab5 subgroups (VM Olkkonen, P Dupree, 1 Killisch, A Lutcke, M Zerial, et al. and A Liitcke, A Valencia, VM Olkkonen, G Griffiths, RG Patton, et al., unpublished data). What are the roles played by these different Rab proteins? One possibility is that they function in either of the transport routes that meet in this compartment: transport from and to the cell surface, and from the TGN and to late endosomes [28]. Alternatively, several distinct Rab proteins might be required in a given transport process. These proteins might carry out either the same function (e.g. vesicle docking) or different functions (e.g. budding, uncoating, docking or fusion).

Cell type specific Rab proteins

The notion that Rab proteins control distinct intracellular transport processes suggests that cell type speciiic steps in membrane traific also require specific Rab proteins. Several Rab proteins that are cell type or tissue-specific have been identiIied. For instance, neurons, exocrine and endocrine cells, which display regulated secretion, speciIically express the Rab3a protein [ 29-311. The expression of cell type specific Rab proteins is regulated during differentiation. Expression of Rab3d increases during ditferentiation of 3T3-Ll cells into adipocytes. This protein might control the insulin- dependent exocytosis of vesicles containing glucose transporter in these specialized cells [32]. Recent data

indicate that epithelial cells, which have distinct apical and basolateral transport pathways, and tmnscymtic routes &vkewed in [33,34]), also express specific Rab proteins once they become polarized. In the developing kidney, Rab17 is absent from the mesenchymal precur- sors but is induced upon their differentiation into epithelial cells. In the adult organ, the protein is associated with the basolateral plasma membrane and with apical tubules [35*]. Both expression and localization suggest that Rabl7 may function in transcytosis, an epithelial cell specilic transport process [ 331.

Rab8, a ubiquitously expressed protein sharing high sequence homology with Sec4 [27], functions in tratfic from the TGN to the basolateral surface in polarized MDCK cells and to the somatodendrltic plasma membrane in hippocampal neurons [36,37]. These two domains are thought to correspond to tile plasma membrane of non-polarized cells [ 381. This suggests that Rab8 does not function in an epithelial cell speciIic pathway. In contrast, another as yet unidenti6ed GTP-binding protein is found on immuno-isolated apical exocytic vesicles and is implicated in apical exocytosis, the epithelial cell specific transport step (361.

Collectively, these data suggest that specialized Rab proteins are part of the machinery regulating cell type specilic transport processes. The localization and function of the Rab proteins must depend on other cell type specific regulatory components, for example, factors that regulate their GTP-GDP cycle.

Regulatory factors for Rab proteins

The cycle of GTP binding and hydrolysis of Rab proteins has been postulated to ensure directionality in the vesicle docking/fusion process [ 41. According to a common view [4,39], a speciiic Rab protein is recruited on the donor membrane or on nascent transport vesicles in its GTP-bound form. Hydrolysis of GTP by the Rab protein is thought to trigger fusion of the vesicle with the acceptor membrane [4,40*]. Subsequently, the GDP-bound Rab protein is released into the cytosol and returned to the donor membrane. There, a GDP-GTP exchange protein catalyzes GDP release and GTP binding and allows the cycle to continue.

Proteins that promote GDP-GTP exchange by members of the Rab protein family have been characterized in yeast and mammalian cells [41,42*,43m]. Do these factors regulate the activity of speciiic Rab proteins? In vitro, yeast Dss4 stimulates GDP dissociation of both Sec4p and Yptlp [42*]. Mammalian suppressor of Sec4, Mss4 protein, that acts on Sec4p, also stimulates GDP release from Yptlp and mammalian Rab3a [43’]. Thus, exchange proteins appear to be active on multiple Rab proteins. If Rab proteins associate with their respective compartments in the GTP-bound form [ 391, then exchange factors might function in association with organelle-specific components. In contrast, GTPase-activating proteins (GAPS) appear to be more selective as the recently identified GAP

616 Membranes,

of Ypt6p (GYP6) does not act efficiently 0; other Ypt proteins [44*].

Association of Rab proteins with membranes requires gemnylgeranyl groups on the carboxyl-terminal cysteine motif [45,46]. This process is reversed by a cytosolic factor, Rab GDI, originally puriiied as a GDP dissociation inhllitor for Rab3a 1471. Rab GDI is able to recognize several geranylgeranylated Rab proteins and extract them from membranes in vitro [48-521. These combined properties suggest that Rab GDI could function to remove Rab proteins from membranes after GTP hydrolysis and to return them to the donor membrane. As predicted from such a function, cytosolic Rab proteins are found in complex with Rab GDI [50-521 and a small but significant fraction of Rab GDI is associated with various organelles [50]. Thus, Rab GDI appears to be an essential component that ensures the cyclical and vecto- rial function of Rab proteins.

Is the function of GDI regulated? Analysis of the quartet mutation in D. mekznogaster suggests that it might be. Mutations in the quartet locus are associated with pleiotropic developmental defects in the larvae and distinguished by the presence of a basic shift in the isoelectric point of three abundant proteins, one of which has been identihed as a Drosopbilu homologue of Rab GDI [ 531. This suggests that Drosophila Rab GDI is post-translationally mod&d or demodiiied by the wild- type quartet gene product. Although there is no evidence that quurtetencodes a kinase, it is interesting to note that the cytosolic form of mammalian GDI seems to be highly phosphorylated (J Gruenbetg, personal communication). Phosphorylation or other post-translational modifications of Rab GDI might be necessary to inhibit reassociation of Rab proteins with the target membrane and direct them to the donor compartment.

Other components may also be involved in the Rab protein cycle. Geranylgeranylation of Rab proteins is cat- alyzed by geranylgeranyl transferase II, an enzyme that, in contmst to other known isoprenyl transferases [54], requires sequences amino-terminal to the target cysteine motif [55-571. This peculiarity may be due to a third enzyme subunit, component A [ 541. Perhaps component A recognizes Rab proteins and presents them to component B, the catalytic part of the transferase.

Intriguingly, the gene products of human and mouse choroideremia (CI-IM), a disease associated with degeneration of the retina and the underlying vasculature, the choroid, share sequence homology with rat component A 158,591. Two additional iindings support the possibility that CHM could be a member of a family of A components. First, in cytosol of lymphoblasts from CHM patients, geranylgeranylation activity is reduced and isoprenylation of Rab3a is more a&c&d than that of Rabla [&I. Second, another CHM-like gene, CHML, has been identiiied [ 611. Interestingly, CHM shares sequence similarity with Rab GDI [62]. This sequence conservation might reflect the ability of both GDI and component A to interact with geranylgeranylated Rab proteins. How- ever, unlike Rab GDI, component A can also recognize the non-prenylated form of Rab proteins, and it interacts

with both the GTP- and the GDP-bound forms [59’]. Therefore, it is thought that component A would be more likely to act as a chaperone for Rab proteins during the initial process of membrane association, than to remove F&b proteins from membranes. This function might be necessary to prevent interactions of the hydrophobic geranylgeranyl moiety with non-target membranes.

Rab proteins and the vesicle transport machinery

If Rab proteins are implicated in the regulation of vesicular transport, the real challenge now is to un- derstand how they function. For this task it is important to study the relationship between Rab proteins and other elements of the transport machinery. Differ- ent vesicle fusion processes in mammalian cells and in yeast require N-ethylmaleimide-sensitive fusion protein (NSF) and the soluble NSF attachment proteins (SNAPS) [ 1*,2*]. More recently, a search for SNAP receptors (SNARES) has yielded four different factors from bovine brain, all associated with the synapse [63**,64*]. VAMP/synaptobrevin-2 is found on transport vesicles (v- SNARE), while syntaxin A and B are on target membranes (t-SNARES). It is not clear where the fourth protein, SNAP-25 (synaptosome-associated protein of 25 kDa) is localized. These findings have led to the hypothesis that, in general, specificity in the process of vesicle fusion with its target compartment might be mediated by the interaction of a specific v-SNARE with its complementary t-SNARE through the NSF-SNAPS complex (Fig. 1).

Where do Rab proteins fit into this scheme? The answer is not there yet. However, recent progress suggests possible connections between Rab proteins and the vesicle fusion apparatus. Rabphilin-$4, a protein that interacts with Rab3a, has recently been puriiied and characterized [65,66*]. Structural analysis of the deduced amino acid sequence from the corresponding cDNA revealed high homology with synaptotagmin, a synaptic vesicle integral membrane protein [67]. However, unlike synaptotagmin, rabphilin3A contains neither a signal peptide nor a putative transmembrane domain. The precise function of synaptotagmin is unknown, although it has been postulated to be a regulatory component of calcium-dependent exocytosis in the presynaptic terminal [67,68]. The identification of rabphilin3A provides a potential link between Pab proteins and t-SNARES. In fact, syntaxin (a t-SNARE) can be co-immunoprecipitated with synaptotagmin suggesting an association of the two proteins [67], and rabphilin3A is a synaptotagmin-like molecule. In addition, the SLK’Z/SEC22 and SLY12IBETl genes, which are multicopy suppressors of functional loss of YPTl in S. cerevtie, encode two members of the synaptobrevin (v-SNARE) family [69-711. Therefore, it is possible to envisage interactions between a rabphilin and a t-SNARE and between a Rab protein and a v-SNARE (Fig. 1). The speciiicity of interactions between V-SNARES and t-SNARES in a complex with NSF and SNAPS might thus require further regulation by a Rab protein. This regula-

Rab GTPases in vesicular transport Zerial and Stenmark 617

z (a)

Fig. 1. Model for possible interactions between Rab proteins and known components of the membrane traffic machinery. Proper docking and fusion of a vesicle with its correct target membrane depends on interactions between specific components present on both compartments. (a) NSF and SNAPS interact with v- SNARE- and t-SNARE-type molecules 1671. (b) interaction between the vesicle-specific v-SNARE and its target-specific t-SNARE is still hypothetical. Cc) The vesicle-specific, CTP-bound Rab protein interacts with its corresponding rabphilin, and Cd) possibly, to a v- SNARE. (e) Rabphilin might interact with a t-SNARE as suggested by the interaction of syntaxin with synaptotagmin [671. In this chain of connections, the CTPase activity of Rab proteins could be utilized to proof-read I41 the interaction between the correct pair of SNARES.

tory mechanism might be used to proof-read the correct interactions between these components. As rabphilin- 3A can only recognize the GTP-bound form of Rab3A [66*], GTP hydrolysis would confer unidirectional@ to the docking/fusion process.

Conclusions

A comprehensive analysis of the molecules implicated in membrane trticking is beginning to establish which components are common to many transport processes, and which are specific. The data obtained so far suggest that, whereas NSF and SNAPS belong to the first cathe- gory, Rab proteins belong to the latter one. Although a number of studies have demonstrated the role of these GTPases as regulators of membrane traffic, their precise function remains unknown. Proteins that regulate the membrane association and nucleotide state of Rab proteins have now been identified, and further mecha- nistic studies will undoubtedly focus on three key ques- tions. First, what brings about the organelle-specific localization of Rab proteins? Second, which are their target molecules? Third, how does the nucleotide state (regu-

lated by GAPS and exchange proteins) control the interactions between Rab proteins and their target molecules? The synaptotagmin-like protein rabphilin-3A is certainly an interesting candidate for a Rab target [66*]. Com- plicated molecular machineries are often regulated by a number of different GTPases, and Rab proteins are not the only GTPases implicated in intracellular transport. Compelling evidence indicates that members of other subfamilies of GTPases function in regulating different aspects of membrane traffic, such as vesicle budding and vesicle coat assembly and disassembly [28,39]. Fu- ture studies will reveal whether there are links between Rab proteins and these other GTPases.

The identification of several tissue- and cell type specific Rab proteins suggests that the complexity of this fan- ily of GTPases could still increase, when one consid- ers the high heterogeneity of cell types from different organs. Learning more about the distribution of Rab proteins in various cell types will contribute to under- standing the relationship between structural features of organelles and their specialized functions, and between sub-compartments and transport routes.

Acknowledgements

We thank Jean Gruenberg, Bernard Hotlack, Anne Liitcke, Carol Mur- phy, Rob Parton, Sanjay Pimplikar and Kai Simons For comments and suggestions.

References and recommended reading

Papers of particular interest, published within the annual period of review, have been highlighted as: . of special interest . . of outstanding interest

1. ROTHMAN JE, ORCI L: Molecular Dissection of the Secretory Pathway. Natrrre 1992, 355:4OW15.

This review focuses on the transport pathways in the ER and GO&$, highlighting important molecular components, such as vesicle coat proteins, the KDEL receptor of the ER, NSF and SNAPS.

2. PRYER NK, WUESTEHUBE LJ, SCHEKMAN R Vesicle-Mediated . Protein Sorting. Ann~r Reu Biochenz 1992, 61:471-516. A comprehensive oveniew of the constituents of the vesicle transport machinery.

3. KREIS TE: Regulation of Vesicular and Tubular Membrane . Traffic on the Go&i Complex by Coat Proteins. Curr @in

Cell Bioi 1992, 4:6Osr615. This review discusses the role of coat proteins in intro-Golgi transport and in transport berzeen the Go@ apparatus and other organelles.

4. BOLWE HR: Do GTPases Direct Membrane Traffic in Secre- tion? CeN 1988, 53:66%671.

5. TOUCHOT N, CHARDIN P, TAUTLAN A: Four Additional Members of the Ras Gene Superfamily Isolated by an Oligonucleotide Strategy: Molecular Cloning of YFT-Related cDNAs from a Rat Brain Library. Proc Nad Acad Sci USA 1987, 84:821&8214.

6. PLUTNER H, COX AD, PwD S, KHOSRAVI-FAR R, BOURNE JR, SCHVANINCER R. DER CJ, BALCH B: Rablb Regulates Vesicular Transport Between the Endoplasmic Reticulum and Suc- cessive Golgi Compartments. J Cell Biol 1991, 115:3143.

618 Membranes

7. TBDAE E, BOWINE JR, KHOSR&%FAR R, DER tZJ, BALCH WE: GTP-Binding Mutants of Rabl and Rab2 Are Potent In- hibitors of Vesicular Transport from the Endoplasmic Retic- ulum to the GoIgI Complex. J Cell Bid 1992, 119749-761.

8. PADFIELD PJ, BALCH WE, JAMIESON JD: A Synthetic Peptide of the Rab3a Etiector Domain Stimulates hmylase Release from Permeabiied Pancreatic Acini. Proc Nafl Acad Sci USA 1992, 89~16561660.

9. SENYSHYN J, BALCH WE, HOIZ RW: Synthetic Peptides of the EtTector-Biding Domain of Rab Enhance Secretion from Digitonin-Permeabiid Chromffi CeUs. FEBS Len 1992, 30961-46.

10. OBEIXHAUSER AF, MONCK JR, B,uc~ WE, FERNANDEZ JM: Exo- cytic Fusion is Activated by Rab3a Peptides. Nahtre 1992, 360:27&273.

11. VAN DER SLIJIJS P, HUIL M, ZAHRAOUI 4 TAHITIAN A, GOUD B, MELMAN I: The SmaU GTP Binding Protein Rab4 Is Associ- ated with Early Endosomes. Pruc Nat1 Acad Sci USA 1991, 88: 6313-6317.

12. CHAWUER P, PARTON RG, IL~uRI HP, SlMo~s K, ZElU4L M: Localization of Low Molecular Weight GTP-Binding Pro- teins to Exocytic and Endocytic Compartments. Cell 1990, 62:317-329.

13. BUCCI C, PARTON RG, MATHER IM, STLJNNENBERC H, SIMONS K, . HOFLXK B, ZERLU M: The SmaII GTP-ase RabS Functions as

a Regulatory Factor in the Early Endocytic Pathway. Cell 1992, 70:715-728.

Transient overexpression of wild-type ItabS in baby hamster kidney cells led to an increased rate of endocytosis and to the formation of unusually large early endosomes. By contrast, cells expressing a nucleotide binding-delicient RabS mutant displayed a reduced rate of en- doqtosis and contained numerous smaU vesicles and tubules dose to the plasma membrane. The resuhs indicate that RabS is rate-limiting and regulates the membrane t&c between the plasma membrane and the early endosome. 14. GORVEL JP, CHAMUER P, ZER~AL M, GRUEN~ERG J: RabS Controls

EarIy Endosome Fusion fn Vh-o. Cell 1991, 64:91S-925.

15. VAN DER SLLIIJS P, Huu M, WEBSTER P, MALE P, GOUD B, . MEUMAN I: The SmaIl GTP-Binding Protein gab4 Controls

an Early Sorting Event on the Endocytic Pathway. Cell 1992, 70:729-740.

Whereas normal Chinese hamster ovaq cells contained only 20% of their transferring receptors on their surface, this number was raised to 80 % In stable transfectants overexpressing Rab4. Overexpression also prevented delivery of intemalized transferrin to acidic endosomes, and a large fraction of internalized transferrin was associated with small, tubular dusters incapable qf efficient acidI6cation. Overexpression of Rab4 did not affect initial rates of transfenin internalization, but the rate of transferrin recycling and a fluid phase marker to the plasma membrane were increased. The results suggest that Rab4 controls the rate of recycling, possibly by controlling the rate of transport between the early endosome and a putative recycling compartment. 16. ~MBARDI D, SOIDATI T, R~EDERER w GODA Y, ZERVU. M,

PF’EFFER SR: Rab9 Functions in Transport Between Late Endosomes and the Trans Golgi Network. EMBO J 1993, 12:677-682.

17. WKHMANN H, HENGST L, GALLWITZ D: Endocytosis in Yeast: . Evidence for the InvoIvement of a SmslI GTP-Binding Pro-

tein (Ypt7p). Cell 1992, 71:1131-1142. A S. cerevtiegene encoding Ypt7, a protein with 63 % sequence identity to mammalian Rab7, was cloned. Ypt7 and gab7 have identical effector regions. Gene disruption of Ypt7 had essentially no effect on the growth properties of yeast ceUs at various temperatures. However, the Ypt7 null mutants were characterized by a highly fragmented vacuole. while uptake of the pheromone alpha factor occurred at a normal rate, its degradation was severely inhibited in the null mutants.

18. SINGER-KRUGER B, FRANK R, CRAUSAZ F, RlEzMAN H: PattIal PurItication and Characterization of Early and Late Endo- somes hrn Yeast: Identilication of Four Novel Proteins. / Bfoi cbe?n 1993, 268: in press.

19.

20.

21.

22.

23.

24.

25.

26. .

KOUKA T, IKEDA K: Reversible Blockage of Mem- brane Retrieval and Endocytosis In the Garland Ceils of the Temperature-Sensitive Mutant of Drosopblla melanogastw, shibire-tsl. J Cell Biol 1983, 97~499-507.

VAN DER BUEK AM, MEYERO~TIZ EM: Dynamin-Like Protein Encoded by the Drosophila slnibire Gene Associated with Vesicular TraEic. Nature 1991, 351:411414.

CHEN MS, OUR RA, SCHROEDER CC, AUK TW, POODRY Cq WNXWARTH SC, VAUEE RB: Multiple Forms of Dynamin Are Encoded by sbibltq a Drosophfla Gene Involved in Endo- cytosls. Nature 1991, 353585586.

VAIENCIA A, Corn P, W~GHOFER A, SANDER C: The ras Protein Family: Evolutionary Tree and Role of Conserved Amino Acids. Biocbem~ty 1991, 30:4637-4648. CHAMUER P, SMONS K, ZERIAL M: The CompIexity of the gab and Rho GTP-Biding Protein Subfamilies Revealed by a PCR Cloning Approach. Gene 1992, 112:261-264.

ELFER~NK L4, ANZAI K, SCHELIER RI-I: RablS, a Novel Low Molecular Weight GTP-Binding Protein SpeciftcaIIy Ex- pressed in Rat Brain. J Biol them 1992, 2671-8.

BAILLY E, MCCAFFREV M, TOUCHOT N, ZAHRAOUI A, GOUD B, BORNENS M: Phosphorylation of Two Small GTP-biid- ing Proteins of the Rab Family by p34cdc*. Nafure 1991, 350:71S-718,

VAN DER SLUIJS P, Huu M, HUBER IA, m P, GOUD B, MEUMAN I: Reversible Phosphorylation - Dephosphorylation Deter- mines the Localization of Rab4 during the CeU Cycle. EMBO J 1992, 11:4379-4389.

RaMa is known to become cytosolic during mitosis as a consequence of phosphorylation 125). This regulation of membrane association may play a role in the arrest of membrane traffic observed during metaphase. In this paper, it is shown that mutation of the consensus site for p34dd kinase in the carboxyl terminus of RaMa prevents its phosphorylation during mitosis and blocks its appearance in the cytosol. Post-translational modilication of the carboxyl-terminal cysteine motif of RaMa was shown to be unafTected by phosphorylation, and membrane association was found to re-occur upon dephosphorylation after mitosis.

27.

28.

29.

30.

31.

32.

33.

34.

35. .

CHAVRIER P, VmGRoN M, SANDER C, SlhlONS K, ZERLU M: Molec- ular Cloning of YPTl/SEC4-Related cDNAs from an Epithe- liaI Cell Line. Mel Cell Biol 1990, 10:6578-6585.

GRUENBERG J, CIAGUE MJ: Regulation of Intracellular Mem- brane Transport. Curr @in Cell Biol 1992, 4:593-W.

AYE J, OLOFSSON B, TAVTMN A, PROCHL+N-IZ A: Developmen- tal and Regional Regulation of Rab3, a New Brain Specific Ras-Like Gene. J Neurasci Res 1989, 22:241-246.

SANO K, KIKUCI 4 ~WIWI Y, TERAN~SHI Y, TAKAI Y: Tissue Specitic Expression of a Novel GTP-Binding Protein (Smg p2SA) mRNA and Its Increase by Growth Factor and Cyclic AMP in Rat Pheochromocytoma PC12 Cells. Biochm Bio pbys Res Comm 1989, 158:377-385. MIZOGUCHI A, Km S, UEDA T, KIKUCHI A, YORJFUJI H, HIROKAWA N, TAUI Y: Localization and S&cellular Distribution of Smg ~254, a ras p21-Like GTP-Binding Protein, in Rat Brain. J Biol Cbem 1989, 265:11872-11879. BALDINI G, HOHL T, Lm HY, LODISH HF: Cloning of a Rab3 Isotype Predominantly Expressed in Adipocytes. Proc NatI Acad Sci USA 1992, 89504HO52.

MOSTOV K, APOCADA B, AROETI B, OKAMOTO C: Plasma Mem- brane Protein Sorting in Polarized Epithelial Cells. J Cell Biol 1992, 1163577-583. NELSON J: Regulation of Ceil Surface Polarity from Bacteria to Mammals. Sc&nce 1992, 258:94&955.

LC~TCKE A, JANSSON S, PARTON RG, CHAMUER P, VALENCLA A, HIJBER L4, LEHTONEN E, ZERIAL M: Rabl7, a Novel SmaU GTPase, Is Specific for EpitheIial CeUs and Is Induced during Ceil Polarization. J Cell Biof 1993, 121:553564.

Rab CTPases in vesicular transport Zerial and Stenmark 619

Cloning of a mouSe cDNA for a novel Rab protein, Rabl7, is reported. Northern blot analysis revealed that Rab17 is expressed exclusively in polarized epitheiial cells and is thus the first example of an epitheiiai ceii speciUc Rab protein. In situ hybridization on murine developing kidneys showed that rabl7 expression was turned on concomittantIy with the onset of epitheiiai differentiation. In the adult kidney, Rab17 was found to be associated with the basolaterai plasma membrane and with apical tubules. Localization data suggest that Rabl7 might be involved in reguiating transcytosis.

36.

37.

38.

39.

40. .

HUBER IA, PIMPUKAR S, PARTON RG, VIRTA H, ZERIAL M, SIMONS K: Rab8, a SmaU GTpase Involved In Vesicular Traffic between the TGN and the Basolateral Plasma Membrane. / cell Biof 1993, 121: in press.

HUBER u DE HOOP MJ, DUPREE P, ZERML M, SIMONS K, Do’rn C: Protein Transport to the Dendritic P!asma Membrane of Cultured Neurons Is Regulated by Rab8p. / Celf Biol 1993, 121: in press.

DCnn CG, SIMONS K: Polarized Sorting of VII Glycoprotekts to the Axon and Dendrites of Hippocampai Neurons in CuI- ture. Cell 1990, 62~63-72.

PFEFFER SR GTP-BindIng Proteins in IntraceUuIar Transport. Trenak Cell Biol 1992, 2:4145.

WALWORTH NC, BRENNWAU) P. KAFICENEU AK, GARRETT M, NOLICK P: Hydrolysis of GTP by Sed Protein Plays an Im- portant Role in Vesicular Transport and Is StimuIated by a GTPase-Activating Protein in Succharomyces cerevisiue. Mol Cell Biol 1992, 12:2017-2028.

A mutation (Gin79-+Leu) in the Sec4 protein strongiy inhibited its in trinsic GTPase activity. A partially purified GAP activity stimulated the hydrolysis rate of the mutant to about 30% of the GAP-stimulated, wild-type Sec4. The se&Leui’9 allele can function as the oniy copy of sec4 in yeast cells, but the cells are cold-sensitive, exhibit slowed in- vertase secretion and accumulate secretory vesicles. It thus appears that the mutation in Sec4 causes a partial loss of function. This is the best evidence so far that GTP hydrolysis by a member of the Sec4/ypt/Rab famiiy is necessary for vesicle fusion, as has been suggested earlier 14).

41. BU~~TEIN ES, MAC~RA IG: Characterization of a Guanlne Nu- cleotide Releasing Factor and GTPase Activating Protein that Are Specitic for the Ras-Related Rotein p25 Rab3a Proc Nat1 Acad Sci USA 1992, 89:1154-1158.

42. MOYA M. ROBERTS D, NOV~CK P: DSS4-I Is a Dominant Sup . pressor of sec4-8 that Encodes a Nucleotide Exchange Pro-

tein that Aids Sec4p Function. Nature 1993, 361:460-463. DSS41 was isolated as a dominant suppressor of the temperature. sensitive s&z+8 mutation. The gene product, Dss4-lp, is a protein with a molecular mass of 17kDa. It was found to stimulate GDP dissociation from Sec4p as weU as from Yptlp in vitro. Post-translational modification of Sec4p was not required for its interaction with Dss4-lp. The activity on Sec4p was about twofold higher than that on Yptl. Dss4-lp was not active on Rab3a and Ra.52. Disruption of the wild-type dss4 gene was not lethal. However, synthetic lethality was observed when ds4 disruption was combined with certain ret mutants showing lethality in the presence of the se&-8 mutation. It is therefore suggested that the normal role of Dss4-lp is to aid Sec4p function.

43. BURTON J, ROBERTS D, MO~~IDI M, NOV~CK P, DE CAMILIJ . P: A Mammalian Guanine-Nucleotide Releasing Protein En-

hances Function of Yeast Secretory Protein Sec4. Nature 1993, 361:464-467.

A rat-brain cDNA library was expressed in temperature-sensitive se&8 yeast ceUs, and screened for suppression of the growth defect at the re- strictive temperature. A temperature-resistant clone was recovered, and the mss4 cDNA was isolated. The Mss4 protein has a molecular mass of 14kDa and has sequence homology with Dss4-lp [41]. Ms.&p stimulated GDP release from Sec4p, and to a lesser extent from Rab3a and Yptl, but was inactive on Ras2. The mss4 gene is expressed in a vari- ety of tissues, suggesting that its natural target is one or more widely expressed Sec4-iike proteins.

44. STROM M, VOUMER P, TAN TJ, GALLWTIZ D: A Yeast GTPase- . Activating Protein that Interacts SpeciIicaUy with a Member

of the Ypt/Rab Family. Nurure 1993, 361:73&739.

Yeast cells were transformed with a yeast genomic library on a multicopy plasmid, and extracts from the individual transformants wem assayed for their ability to stimulate the GTPase activity of Ypd. A single, positive colony was isolated, and from this the Cl?6 (GAP of Ypt6) gene was cloned. The gene product, Gyp6, has no sign&ant homology to other known proteins, Including other GTPase-activating proteins. Purified Gyp6 strongly stimuiated the GTPase activity of Ypt6, and had a weak activity on Ypt7. It did not stimulate the GTPase activ ity of Yptl, Sec4, YpUA, H-ras or Rab2. Whereas @tideletion mutants are temperature-sensitive, disruption of gyp6resulted in no phenotypic alterations.

45. MAGEE T, N!zwhr~~ C: The Role of Lipid Anchors for SmalI G Proteins In Membrane TraBickIng. Trends 011 Bid 1992, 2:318-323.

46. CLARKE S: Protein IsoprenyIatIon and Methylation at Carboxyl-Terminal Cystelne Residues. Annu Rev Biocbem 1992, 61:355-386.

47. S~sw T, KIKUCHI 4 A&w S, HATA Y, Isomuw M, KURODA S, TAKA~ Y: PurIIication and Characterization from Bovine Brain Cytosol of a Protein that Inhibits the Dissociation of GDP from and the Subsequent BInding of GTP to Smg p25A, a Ras p21-Like GTP-Binding Protein. J Eiol Ckm 1990, 265:233+2337.

48. ARNU S, KIKuCHI A, HATA Y, l%MURA M, TIUW Y: Regula- tion of Reversible Binding of Smg p25A a Ras p21-like GTP-Bmding Protein to Synaptic Plasma Membranes and Vesicles by Its SpeciIic Regulatory Protein, GDP DIssocia- tion Inhibitor. J Biol &em 1990, 265:13007-13015.

49. SASAKJ T, KAIBUCHI K, KA~CENEU AK, NOV~CK P, TAJLAI Y: A Mammalian Inhibitory GDP/GTP Exchange Protein (GDP Dissociation Inhibitor) for Smg p25A Is Active on the Yeast SEC4 Protein. Mol Celf Biof 1991, 11:2909-2912.

50. UUIUCH 0, STENh4ARK H, ALEXANDROV K, HUBER IA, KA~UCHI K, SA%KI T, TAKAI Y, ZEIUAL M: Rah GDI as a General Regulator for the Membrane Association of Rab Proteins. J Bid C%e?n 1993, 268: in press.

51. REGAZZI R, K~KUCHI A, TAXAI Y, WO~IM CB: The SmaU GTP- Siding Proteins in the Cytosol of InsulinSecretIng CelIs Are Complexed to GDP Dissociation Inhibitor Proteins. J Biol c%em 1992, 26717512-17519.

52. SOIDATI T, REIDERER MA, PFEFFER SR Rab GDI: a Solubiliaing and Recycling Factor for Rab9 Protein. Mel Bid Cell 1993, 4:425434.

53. Z.AHNER JE, CHENEY CM: A Dnxopbiiu Homo& of Bovine Smg p25a GDP Dissociation Inhibitor Undergoes a Wft in Isoelectric Poiit In the Developmental Mutant Quartet. Mel Cdl Biol 1993, 13217-227.

54. SEABRA MC, GOLDSTEIN JI, SUDHOF TC, BROWN MS: Rab Geranylgeranyl Transferase. A Multisubunit Enzyme that Prenylates GTP-Binding Proteins Terminating in Cys-X-Cys or Cys-Cys. J Bid Cbem 1992, 26714497-14503.

55. KINSEUA TB, MALTESE WA: Rab GTP-Biding Roteins with Three Different Carboxyl-Terminal Cystelne Motifs Are Moditied in Vivo by 20Carbon Isoprenoids. J Biol C&m 1992, 267:394&3945.

56 KH~~RAVI-FAR R, Ct&tx GJ, ABE K, COX AD, MCLUN T, LUtZ RJ, S~NENSKY M, DER CJ: Ras (CXXX) and Rab (CCKXC) Prenylation Signal Sequences Are Unique and FunctlonaIIy Distinct. J Biol C&m 1992. 267:2436324368.

57. P!ZTER M, CHAVR~ER P, NIGG EA, ZE~UAI. M: Isoprenylation of Rab Proteins on StructuraIly Distinct Cysteine Motifs. J Cd Sci 1992, 102:857-865.

58 Cit!ZMERS FPM, VAN DE POL DJR, VAN KERXHOFF LPM, WtElUNGA B, ROPERS H-H: Cloning of a Gene that Is Rearranged In Patients with Choroideremia Nuhtre 1990, 347~674-677.

59. SEABRA MC, BROWN MS, SIAUGHTER (2% SUDHOF TC, GOID!%IN . JL PuriIicatIon of Component A of Rab GeranyIgeranyI

620 Membranes

Transferase: Possible Identitv with the Choroideremia Gene Product. Cell 1992, 70:104P-1057.

Partial peptide sequences of purllied rat component A of geranylgetanyl transferase II showed high similarity to the human and mouse CHM gene products. 60. SEABRA MC, BROW MS, GOIDSTEIN JL: Retinal Degeneration . in Choroideremia: Deficiency of Rab Geranylgeranyl Trans-

femse. Science 1992, 259:377+X. Cytosol prepared from lymphoblasts of patients suiTering from the X- linked eye disease CHM exhibited reduced ability to geranylgeranylate RabIa and Rab3a in c&n The ability to geranylgeranylate Rab3a was more alfected than that of Rabla The results are consistent with the idea that the CHM gene product may be member of a family of A components of Rab getanylgeranyl transferases. 61. CREMERs FPM, MOUOY CM, VAN DE POL DJR, VAN DEN HURK

JAJM, BACH I, GEURT~ VAN KERNEL AHM, ROPERS H-H: An Auto- somal Homologue of the Choroideremia Gene Colocalizes with the Usher Syndrome ‘Qpe II Locus on the Distal Part of Chromosome lq. Hum Mol Gene1 1992, 1:71-75.

62. FODOR E, LEE RT, O’DONNELL JJ: Analysis of Choroderemia Gene [Letter]. Nufure 1991, 351:614.

63. SOMR T, WHI?EHEART SW, BRUNNER M, ERDJUMEN~-BROMAGE . . H, GEROMANOS S, TEMPST P, ROTHMAN JE: SNAP Receptors

Implicated in Vesicle Targeting and Fusion. Nahtre 1993, 362:318324.

SNAPS are soluble receptors for the general fusion protein NSF. In this paper, SNAP receptors (SNARES) from bovine brain were isolated by the use of al??nity columns containing NSF/SNAP complexes. Remark- ably, of the four SNARES identified in this manner, all are proteins that have been previously found to be associated with the synapse. One of them, VAMP/synaptobrevln 2, is a synaptic vesicle protein, whereas two others, syntaxin A and syntaxln B, are found on the synaptic plasma membrane. The fourth SNARE, SNAP-25, is also thought to be associated with synaptic vesicles or plasma membrane. The results thus indicate that SNAP/NSF can intemct with SNARES in the vesicular membrane (V-SNARES) as well as with SNARES in the target membrane (t.SNARFs). Whereas NSF and SNAPS appear to be universal components of the vesicle fusion apparatus, it is possible that families of V-SNARES and t- SNARES ensure speciiic interactions between vesicles and their target membranes. 64. Ww SW, Gatrr IC, BRUNNER M, CIAR( DO, MAYER T, . B~JHROW SA, ROTHMAN JE: SNAP Family of NSF Attachment

FToteins Includes a Brain-Specific Isoform. Nufure 1993, 362:353355.

This paper reports the cloning and sequencing of cDNAs encoding a., g. and ~-SNAPS. a- and B-SNAPS were found to be 83% identical to each other, whereas -+SNAP showed about 25% identity to the other

SNAPS. a- and Y-SNAPS are ubiquitously expressed and act synerglsti- ally, whereas B-SNAP is bralnspeclftc. B-SNAP competes for the same biding site on SNAP receptors as a-SNAP, but unlike a-SNAP it cannot restore cell-free Golgl transport in cytosol from secl7 mutant yeast. It is most abundant in areas of high neuron density and may be involved in neurosecretion.

65. SH~RATAJCI H, KA~~UCHI K, YAMAGUCHI T, WA~A K, HO~UUCHI H, TAKAI Y: A Possible Target Protein for smg p25A/Rab3A Small GTP-Biding Protein. J Biol Ckm 1992, 267:10946-10949.

66. SHIRATAKI H, KAIBUCHI K, SAKODA T, KISHIDA S, YAMAGUCHI . T, WADA K, MN- M, TAKA~ Y: RabphiIin-3A, a Putative

Target Protein for Smg p25A/Rab3A ~25 Small GTP-Biid- ing Protein Related to Synaptotagmin. Mol Cell Biol 1993, 13:2061-2068.

This paper reports the cloning of a cDNA encoding tabphilin3A a protein isolated from bovine brain crude membranes by chemi- cal crosslinking to Rab3a. The carboxyl-terminal part of rabphilln3A shows significant sequence homology to the synaptic vesicle protein synaptotagmln, and contains two Cz domains like those found in protein klnase C and synaptotagmin. Unlike synaptotagmin, rabphilin3A does not contain any putative transmembrane region. Rabphliin3A was found to form a complex with the GTPyS form of Rab3a, but not with Rab3a-GDP and not with the GTPyS-form of Rabll. 67. BENNET MK, SCHELUR RI-I: The Molecular Machinery for

Secretion Is Conserved from Yeast to Neurons. Proc Nat1 Acud Sci USA 1993, 90:25592563.

68. SHOJ~-KAQI Y, YOSH~DA A SATO K, HOSHINO T, OGURA A, KONDO S, FUJIMOTO Y, Kuw~u~a~ R KATO R, T~HI M: Neurotransmitter Release from Synaptotagmin-Deficient Clonal Variants of PC12 CelIs. Science 1992, 256:182&1823.

69. NEWMAN AP, SHIM J, FERRO-NOVICK S: BETf, BOSI, and SEC22 are Members of a Group of Interacting Yeast Genes Re- quired for Transport from the Endoplasmic Reticulum to the Golgl Cotnlex. Mof Cell Biol 1990, 10:3405-3414.

70.

71.

DASCHER C, Osstc R G~uwrtz D, SCHM~I-~ HD: Identiiication and Structure of Four Yeast Genes (SLY) that Are Able to Suppress the Functional Loss of KPTI, a Member of the RAS Supedamily. Mol Cell Biol 1991, 11:872+385.

NEWMAN AP, GIW J, MANctm P. ROSSI G, UAN JP, FEatto- NO~~CK S: SEC22 and SLY2 are Identical. Mol Cell Biol 1992, 12:36633664.

M Zerial and H Stenmark, European laboratoty of Molecular Biology, Postfach 10.229, Meyerhofstrasse 1,6900 Heidelberg, Germany.

rab gtpases in vesicular transport

Documents