HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 36, 32722-32729, September 6, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/36/32722    most recent M205101200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Strick, D. J. Articles by Elferink, L. A. Search for Related Content PubMed PubMed Citation Articles by Strick, D. J. Articles by Elferink, L. A. Social Bookmarking                      What's this?

Mammalian Suppressor of Sec4 Modulates the Inhibitory Effect of Rab15 during Early Endocytosis^*

David J. Strick, Dina M. Francescutti, Yali Zhao, and Lisa A. Elferink

From the Department of Physiology and Biophysics, Marine Biomedical Institute, University of Texas Medical Branch, Galveston, Texas 77555-1069

Received for publication, May 23, 2002, and in revised form, June 24, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Rab15 is a novel endocytic Rab that counters the stimulatory effect of Rab5-GTP on early endocytic trafficking. Rab15 may interfere with Rab5 function directly by sequestering Rab5 effectors or indirectly through novel sets of effector interactions. To distinguish between these possibilities, we examined the effector binding properties of Rab15. Rab15 does not interact directly with the Rab5 effectors rabex-5 and rabaptin-5 in a yeast two-hybrid binding assay. Rather mammalian suppressor of Sec4 (Mss4) was identified as a binding partner for Rab15. Mss4 preferentially binds GDP-bound (T22N) and nucleotide-free (N121I) Rab15, consistent with the proposed role of Mss4 as a chaperone that stabilizes target Rabs in their nucleotide-free form. Mutational analysis of Rab15 indicates that lysine at position 48 (K48Q) is important for the binding of Rab15-GDP to Mss4. Moreover, the mutation K48Q counters the inhibitory phenotype of wild type Rab15 on receptor-mediated endocytosis in HeLa cells and homotypic endosome fusion in vitro without altering the relative amount of cell surface-associated transferrin receptor. Together, these data indicate a novel role for Mss4 as an effector for Rab15 in early endocytic trafficking.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Endocytosis of cell surface receptors regulates both the intensity and duration of receptor signaling by controlling the location of signaling interactions and the desensitization and recycling of activated receptors (reviewed in Refs. 1 and 2). Accordingly, endocytic compartments are highly specialized both in terms of their organization and function. The early/sorting endosome is a major trafficking compartment from which several trafficking pathways emerge. Recently, Rab GTPases have emerged as potent regulators of membrane trafficking through early/sorting endosomes. Rabs do not regulate membrane trafficking per se but function as regulatory throttles impacting the kinetics of membrane transport steps through the recruitment of specific effectors that in turn mediate membrane transport (reviewed in Refs. 3-5). For example, Rab5 mediates the internalization and fusion of incoming endocytic vesicles in vivo (6-8) and the homotypic fusion of endosomes in vitro (8-12). Overexpression of the constitutively active GTP-bound mutant Rab5-Q79L in baby hamster kidney cells results in a dramatic increase in fluid phase and receptor-mediated endocytosis and leads to formation of enlarged early/sorting endosomes. Conversely, overexpression of GDP-bound Rab5 (S34N) reduces endocytic uptake and results in the formation of a diffuse network of small endocytic vesicles (6-8).

Following activation on endosome membranes, Rab5-GTP drives the organization of a specialized membrane domain with distinct functional characteristics (13-15). Rab5-GTP forms this domain by recruiting the phosphatidylinositol 3-kinase hVPs34, which catalyzes the local production of PI-3-phosphate (PI3P)¹ (16-18). Rab5-GTP and hVPs34 activities are essential for the subsequent recruitment of rabenosyn-5 and the docking protein early endosome antigen (EEA1) to early endosomal membranes through PI3P (17-22). EEA1 also interacts directly with syntaxin 13, a SNARE implicated in the fusion of early endosomes (13, 23). Thus a model is emerging in which Rab5-GTP functions as a regulatory protein, driving assembly of specific effector complexes on endosomal membranes leading to membrane fusion (24). Consistent with this model, the early endocytic GTPases, Rabs 4 and 11, have also been shown to organize into distinct domains on early endosomes through the local recruitment of effectors (14, 15). Moreover, Rab4 and Rab5 function are linked through the shared effectors rabaptin-5 (25) and rabenosyn-5 (15). Thus Rab-specific domains appear to coordinate endosomal trafficking directly by communicating via shared effector complexes.

The early endocytic GTPase Rab15 exhibits distinct endocytic localization and activity. Rab15 distributes between two early endosomal compartments, colocalizing with Rabs 4 and 5 on early/sorting endosomes and with Rab11 on pericentriolar recycling endosomes (26). Overexpression of activated Rab15 (Rab15-GTP) inhibits both fluid phase and receptor-mediated endocytosis in vivo and the homotypic fusion of early endosomes in vitro. Conversely, mutations that constitutively inactivate Rab15 (Rab15-GDP) stimulate early endocytosis and fusion of homotypic endosomes in vitro (27). These data suggest that Rab15 functions to reduce endocytic trafficking, primarily at the level of early/sorting endosomes. Consistent with an inhibitory role, overexpression of Rab15-GTP reverses the stimulatory effect of Rab5-GTP on early endocytosis, whereas coexpression of Rab15-GDP with activated Rab5 increased internalization of the fluid phase marker horseradish peroxidase relative to cells expressing activated Rab5 alone (27).

Given the opposing effects of Rab15 and Rab5 on early endocytosis, the transport steps regulated by these GTPases likely intersect at some point within the endocytic network, presumably through the action of a shared effector or accessory protein. Rab15 may interfere with Rab5 function directly by sequestering Rab5 effectors or indirectly through a unique set of effector interactions. To distinguish between these possibilities, we examined the effector-binding properties of Rab15. By using a yeast two-hybrid binding assay, we demonstrate that Rab15 does not directly interact with the Rab5 effectors rabaptin-5 or rabex-5. Rather, mammalian suppressor of Sec4 (Mss4) was identified as a binding partner for Rab15. Mss4 specifically binds GDP-bound Rab15 (T22N) and the nucleotide-free mutant N121I, consistent with the proposed role of Mss4 as a chaperone mediating GDP removal, stabilizing its target Rab in a nucleotide-free state (28-30). Our functional analyses indicate that interactions with Mss4 are required for the inhibitory effect of Rab15 in early endocytosis, suggesting a novel role for Mss4-mediated interactions in early endocytic trafficking.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reagents and Plasmids-- General cell culture reagents and chemicals were obtained from Invitrogen and Fisher, respectively, unless specified otherwise. All restriction enzymes were purchased from New England Biolabs, and Na¹²⁵I was purchased from Amersham Biosciences. The following antibodies were obtained as indicated: anti-HA monoclonal antibody, 12CA5 (Roche Molecular Biochemicals); anti-human transferrin receptor monoclonal antibody, H68.4 (Zymed Laboratories Inc.). pET15B expressing rat Mss4 and an anti-rat Mss4 polyclonal antibody (37) were kindly provided by Pietro De Camilli (Yale University). Plasmids encoding Rab5 (7, 8), rabex-5 (31), rabaptin-5 (32), and rabenosyn-5 (22) were generous gifts from Marino Zerial (Max Planck Institute for Molecular Cell Biology and Genetics, Germany) and Harold Stenmark (Norwegian Radium Hospital, Norway). A HeLa cell cDNA library pre-transformed into EGY187 (MAT) was kindly provided by Russell Finley, Jr. (Wayne State University). The cDNAs for wild type and mutant Rab15 (Q67L, N121I, and T22N) containing an amino-terminal HA epitope have been described elsewhere (27) and were cloned directly into pCI-Neo (Invitrogen). Site-directed mutagenesis was performed using the QuikChange^TM Mutagenesis kit (Stratagene) according to the manufacturer's instructions and verified by DNA sequencing (Applied Biosystems).

Yeast Two-hybrid Binding Assays-- Bait strains were prepared by cloning wild type Rab15 and its respective mutants into pLexA (CLONTECH); GTP-bound Rab15 (Q67L), nucleotide-free Rab15 (N121I), GDP-bound Rab15 (T22N), Rab15-T22N containing the single mutations K46L or K48Q, and Rab15-T22N in which the motifs DN (residues 30-31), DFKMK (residues 44-48), and TITK (residues 72-75) were substituted with the corresponding Rab5a sequences KG, AFLTQ, and SLAP, respectively. cDNAs encoding rabaptin-5, rabex-5, and rabenosyn-5 were cloned into pB42AD (CLONTECH). Bait and prey constructs were transformed into RFY206 (MATa) and EGY187 (MAT), respectively, using established techniques (CLONTECH). Expression of the indicated bait and prey constructs was confirmed by SDS-PAGE and Western analysis.² Yeast two-hybrid binding assays were performed by mating bait strains with prey strains (33) as specified in the text. Positive diploids were identified by growth on quadruple synthetic dropout media (Trp/His/Ura/Leu) and LacZ activation. For LacZ activation assays, the appropriate diploids were grown in 5 ml of triple synthetic dropout media (Trp/His/Ura) overnight at 30 °C and subcultured 1:10 in fresh dropout media for 7 h at 30 °C. Cells were pelleted at 1000 × g for 5 min at 4 °C, washed once in 5 ml of Z Buffer (113 mM Na₂HPO₄·7H₂O, 39 mM NaH₂PO₄·H₂O, 10 mM KCl, 1 mM MgSO₄·7H₂O, and 35 mM -mercaptoethanol), resuspended in 120-150 µl of Z Buffer, and subjected to 3 cycles of 1-min freeze/thaws in liquid nitrogen. The lysates were centrifuged at 20,000 × g for 5 min at 4 °C, and 15 µl of the clarified supernatant was incubated with 150 µl of CUG substrate (Molecular Probes) for 30 min at room temperature in darkness. The reactions were terminated with 75 µl of 0.2 M Na₂CO₃, and the relative fluorescence was measured according to the manufacturer's specifications (Molecular Probes). Each assay was performed in triplicate and repeated at least twice. Relative fluorescence units were normalized to the amount of protein in each sample (Bradford, Bio-Rad) and are reported as a measure of relative -galactosidase activity.

A HeLa cell library was screened by mating EGY187 cells with a RYF206 strain expressing Rab15-T22N as indicated above. Plasmid DNA prepared from positive diploids identified in the library screen was isolated by transformation into Escherichia coli KC8 to isolate the library construct. Inserts from the resulting cDNAs were PCR amplified using the primers 5'-CGTAGTGGAGATGCCTCC-3' and 5'-CTGGCAAGGTAGACAAGCCG-3' and analyzed by HaeIII digestion and DNA sequence analysis (Applied Biosystems). DNA and predicted protein sequences were further analyzed using BLAST searches.

Cell Culture and Transfections-- All cells were cultured in Dulbecco's modified Eagle's medium supplemented with penicillin/streptomycin and maintained at 37 °C with 5% CO₂. HeLa media was supplemented with 10% Cosmic Calf Sera (HyClone) and penicillin/streptomycin. Overexpression studies using T7 recombinant vaccinia virus (vTF7-3) were performed as described previously (26, 27). Transient expression using LipofectAMINE^TM (Invitrogen) was performed as described previously (26, 27).

Biochemical Pull-down Assays-- Recombinant rat Mss4 was expressed as a His₆ fusion in BL21-DE3 pLys (Stratagene) and purified by NiNTA affinity chromatography (Qiagen). For pull-down studies, HeLa cells were transfected with Rab5 or HA-tagged Rab15 using LipofectAMINE^TM (Invitrogen) as described elsewhere (27). Transfected cells were resuspended in 200 µl of ice-cold lysis buffer (10 mM HEPES, pH 7.4, 1.5% IGEPAL (Sigma), 0.1 mM MgCl₂, 150 mM NaCl, 10 µg/ml each aprotinin, leupeptin, and pepstatin A), and cell lysates were clarified at 16,000 × g for 5 min at 4 °C. Supernatants were adjusted to 1.0 mM MgCl₂, incubated with 1.0 mM GTPS (Sigma) or 1.0 mM GDPS (Sigma) for 60 min at 4 °C, and the extracts incubated for 2.0 h at 4 °C with 5 µg of purified His₆-Mss4. Mss4 and its associated proteins were isolated by binding to NiNTA beads in lysis buffer containing 10 mM imidazole at 4 °C for 60 min. Beads were washed three times in cell lysis buffer, three times in 150 mM NaCl, 10 mM HEPES-KOH, pH 7.5, 0.1 mM MgCl₂, and analyzed by SDS-PAGE followed by Western analysis by enhanced chemiluminescence (Amersham Biosciences).

Functional Analysis of Rab15 and Mss4 Interactions-- ¹²⁵I-Tfn internalization studies and in vitro homotypic endosome fusion assays were performed on transiently transfected HeLa cells as described previously (27). Surface-bound Tfn was assessed as follows. Transiently transfected HeLa cells were depleted of endogenous Tfn for 1 h at 37 °C and chilled at 4 °C for 30 min to stop endocytosis. The cells were incubated in IM (Dulbecco's modified Eagle's medium with 20 mM HEPES, pH 7.4, and 20 mg/ml bovine serum albumin) containing 3 µg/ml ¹²⁵I-Tfn for 1 h at 4 °C to enable binding of the ¹²⁵I-Tfn to cell surface-associated TfR. The cells were washed in PBS containing 0.1% bovine serum albumin at 4 °C, and the total amount of cell-associated ¹²⁵I-Tfn was measured with a gamma counter (Packard Instrument Co.). 80-90% of the cell-associated ¹²⁵I-Tfn was routinely removed with three successive acid washes as described previously (27) (data not shown). All numerical results were subjected to a one-way ANOVA with a Neuman-Keuls post hoc test to determine statistical significance between selected groups (Prism GraphPad).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Does Rab15 Bind Rab5 Effectors?-- Given the opposing effects of Rab15 and Rab5 on early endocytosis, it is highly likely that the transport steps regulated by these Rabs overlap at some point within the endocytic network, possibly in terms of a common effector or target. We first examined the ability of Rab15 to interact with the Rab5 effectors rabex-5 (31), rabaptin-5 (32), and rabenosyn-5 (22). Rabex-5 is a guanine nucleotide exchange factor for Rab5 originally identified as a component in a complex with rabaptin-5. Rabaptin-5 increases the exchange activity of rabex-5 on Rab5, promoting early endosome fusion in vitro (34). Rabaptin-5 also interacts with Rab4 through a distinct binding site (25) suggesting a role for this effector in recycling from early endosomes to the cell surface. Rabenosyn-5 preferentially interacts with GTP-bound Rab5 (Rab5-Q79L) and PI3P on early endosomes. Consistent with the stimulatory role of Rab5-GTP in endocytosis, rabenosyn-5 promotes homotypic endosome and clathrin-coated vesicle fusion in vitro (22). In addition, rabenosyn-5 interacts directly with Rab4-GTP and promotes transferrin (Tfn) recycling from early sorting endosomes when overexpressed in HeLa cells (15). Rabaptin-5 is a cytosolic protein that was originally identified as an effector for Rab5-GTP using a yeast two-hybrid approach (32). Thus interactions with shared effectors may functionally couple otherwise distinct Rab-mediated transport steps within early sorting endosomes (25, 32).

To determine whether Rab15 binds Rab5 effectors, cDNAs encoding rabex-5, rabaptin-5, and rabenosyn-5 were cloned into the plasmid pB42AD, in-frame with the activation domain of the bacterial transcription factor, B42, and conditionally expressed from the Gal1 promoter in the presence of galactose. Wild type and mutant Rab15 were expressed in-frame with the DNA binding domain of the bacterial transcription factor LexA, and protein-protein interactions were assayed using leu2 (data not shown) and lacZ (Table I) activation as reporter genes. Negligible binding was detected between Rabex-5 and wild type or mutant Rab15. Similarly, we detected no interaction between rabaptin-5 and wild type or mutant Rab15 (Table I). Western analysis indicates that the absence of any notable interaction between Rab15 and rabex-5 or rabaptin-5 in this assay is not related to differences in the relative amounts of the expressed prey and bait proteins (data not shown). Moreover, Rab5 strongly interacted with rabex-5 and rabaptin-5, indicating the specificity of the results. Specifically, rabex-5 bound GDP-bound (S34N) and the nucleotide-deficient mutant Rab5-N133I. Conversely, rabaptin-5 interacted directly with Rab5-GTP (Q79L) and the nucleotide-free Rab5 mutant, N133I. Interestingly, a weak interaction was detected between rabenosyn-5 and wild type Rab15 as well as the GTP-bound and nucleotide-free Rab15 mutants, Rab15-Q67L and Rab15-N121I, respectively (Table I). Given the modest binding observed between Rab15 and rabenosyn-5 in the absence of any discernible nucleotide dependence, the physiological significance of this interaction remains uncertain. Taken together, these data indicate that Rab15 does not directly interact with the Rab5 effectors rabex-5 and rabaptin-5.

View this table: [in this window] [in a new window]   Table I -Galactosidase activity (relative fluorescence units)

Mammalian Suppressor of Sec4 (Mss4) Binds to GDP-bound Rab15-- Because Rab15 does not interact with the Rab5 effectors rabaptin-5 and rabex-5, we reasoned that the inhibitory effect of Rab15 during early endocytosis is regulated by a unique set of effector molecules. Therefore, we screened a HeLa cell cDNA library using Rab15-T22N as "bait" in a yeast two-hybrid system. HeLa cells are known to express Rab15 (26) supporting the premise that Rab15 effectors are represented in this library. Eleven of twelve strongly positive clones were identified as Mss4. All Mss4 clones isolated from this screen minimally encoded residues 1-55. Mss4 and its yeast homologue Dss4p were originally identified in genetic screens for proteins that suppressed the secretory defect of sec4-8 mutants (33, 34). Mss4 specifically binds to and promotes the release of GDP from a subset of Rabs including Sec4, Rabs 1a, 3, 8 10, and 13. Mss4 does not bind Rabs 2, 4 or 7; moreover, no interaction is observed between Mss4 and Rab5 (29, 37). Although Mss4 and Dss4p facilitate the release of GDP from their target Rabs, they lack the ability to promote GTP binding (28, 38). Therefore, Mss4 does not function as a bona fide GEF as reported previously (39) but rather stabilizes Rabs in a nucleotide-free state (30, 38).

Accordingly, we examined the nucleotide dependence of the interaction between Rab15 and full-length Mss4. High levels of -galactosidase activity were observed in strains expressing Mss4 and the inactive Rab15 mutant Rab15-N121I, and to a lesser extent Rab15-T22N (Fig. 1). Mss4 failed to bind wild type Rab15 and its GTP-bound mutant Q67L under these conditions. Mss4 did not interact with wild type Rab5 or its guanine nucleotide-binding mutants Q79L, S34N, and N133I as judged by control levels of -galactosidase (Fig. 1B) and lack of growth on Leu plates (Fig. 1A). No interaction was observed when the diploids were grown on plates containing glucose, indicating that galactose-driven expression of Mss4 was essential for the interaction with Rab15 (data not shown). The absence of -galactosidase activity in cells expressing Mss4, wild type, or mutant Rab15 alone demonstrates that reporter expression was dependent on a two-hybrid protein-protein interaction (data not shown). These data indicate that Mss4 directly interacts with Rab15, preferentially nucleotide-free Rab15-N121I, and the constitutively inactive GDP-bound mutant Rab15-T22N.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Mss4 preferentially binds constitutively inactive Rab15 mutants, N121I and T22N. A, yeast strains expressing Mss4 fused to the activation domain of B42 (pB42AD:Mss4) with wild type (wt) or the indicated Rab mutants expressed as fusions with the DNA binding domain of LexA (pLEXA) were mated on synthetic media lacking tryptophan and histidine but containing X-gal (X-Gal) or synthetic media lacking tryptophan, histidine, and leucine (Leu). Blue colonies on X-gal and growth on Leu plates indicate specific interactions between Mss4 and inactive Rab15 mutants. B, reporter -galactosidase activity was determined and represents the means ± S.E. of triplicate experiments and are normalized with respect to protein concentration. Significant differences were observed between experimental and control conditions (one-way ANOVA, p < 0.001) as described in the text.

We confirmed the interaction between Rab15 and Mss4 using pull-down assays. Recombinant Mss4 was expressed as a His₆-tagged fusion in E. coli, purified by affinity chromatography on NiNTA-agarose, and the purified protein incubated with cell lysates prepared from HeLa cells overexpressing wild type HArab15. The nucleotide dependence of this interaction was examined by first priming the cell lysates with GDPS or the non-hydrolyzable GTP analog, GTPS. Mss4·Rab15 complexes were recovered by binding NiNTA-agarose and analyzed by Western analysis. As shown in Fig. 2A, wild type HArab15 binds Mss4 in the absence and presence of GDPS. Because wild type HArab15 binds and hydrolyzes GTP when transiently overexpressed in Trvb-1 (25) and HeLa cells,² the interaction observed between Rab15 and Mss4 in the absence of GDPS likely reflects binding to GDP-bound and nucleotide-free forms of wild type Rab15. Consistent with this hypothesis, priming the lysates with GTPS prior to the addition of recombinant Mss4 prevents the interaction between Mss4 and Rab15. Moreover, a strong interaction occurs between Mss4 and the nucleotide-free Rab15 mutant, N121I, in the presence and absence of GTPS and GDPS (Fig. 2B). No HArab15 immunoreactivity was detected using beads alone or cell lysates expressing wild type Rab5, confirming the specificity of these results (Fig. 2). Taken together, these data corroborate our yeast two-hybrid studies and indicate that Mss4 directly interacts with constitutively inactive Rab15 mutants, T22N and N121I.

View larger version (42K): [in this window] [in a new window]   Fig. 2.   Mss4-Rab15 interactions are guanine nucleotide-dependent. A, cell lysates prepared from HeLa cells transiently overexpressing wild type HArab15 or Rab5 were incubated with (+) or without () GTPS or GDPS, prior to incubation with purified, recombinant His6-tagged Mss4. Mss4-Rab complexes were immobilized on NiNTA-agarose beads and analyzed by Western analysis (Wn) using antibodies (Ab) for HArab15, Rab5, and Mss4. Expression of the indicated proteins in cell lysates were confirmed by Western analysis (C). B, Mss4 binding studies using lysates prepared from cells overexpressing constitutively inactive HArab15-N121I indicate that Mss4 preferentially binds nucleotide-free HArab15.

Identification of Mss4-binding Sites on Rab15-- Interpreting the effect of overexpressing dominant Mss4 mutants on Rab15-mediated endocytosis is confounded by Mss4 binding multiple Rabs (29). Therefore, we used a yeast two-hybrid approach to identify potential loss-of-function Rab15 mutants that do not bind Mss4. Comparison of the Rab15 peptide sequence with other Mss4-binding Rabs (including Rabs 1a, 3a, 8 and 10, and Sec4p) reveals three conserved signature motifs that may comprise Mss4-binding sites in these proteins (37). In Sec4p and Rab3a, these sites reside within or juxtaposed to the switch I and II regions, regions on the surface of these proteins that change conformation during guanine nucleotide binding and GTP hydrolysis (39, 41). Moreover, these two switch regions are major sites of interaction with regulator and effector molecules in general, including GEFs and GTPase-activating proteins. In Rab15, the putative Mss4-binding motifs include DN (m1), DFKMK (m2), and TITK (m3) (comprising amino acid residues 30-31, 44-48, and 72-75, respectively) (37). To determine whether these three regions contribute to the interaction between Rab15 and Mss4, we individually substituted these motifs in Rab15-T22N with the corresponding regions of Rab5, a GTPase known not to bind Mss4 (see Fig. 1). The three Rab15-GDP mutants (m1, m2, and m3) were tested for their ability to bind Mss4 in a yeast two-hybrid binding assay. As shown in Fig. 3, binding of Mss4 to Rab15-T22N was abolished by substitution of the DN (m1) and DFKMK (m2) motifs. The motif TITK (m3) was not required for binding Mss4 to Rab15, because comparable levels of -galactosidase reporter activity were detected relative to strains expressing Rab15-T22N and Mss4. -Galactosidase activity was reduced 60-70% in the double mutant m1/m3 relative to yeast lysates prepared from strains coexpressing Mss4 with Rab15-GDP (T22N) or the nucleotide-free form of Rab15 (N121I) confirming the importance of the m1 region for Rab15-Mss4 interactions. These data indicate that the DN (m1) and DFKMK (m2) motifs are required for Mss4 binding to Rab15.

View larger version (43K): [in this window] [in a new window]   Fig. 3.   K48Q abolishes the interaction between HArab15 and Mss4. A, the relative positions of the T22N mutation in the first of four GTP-binding motifs (gray boxes) and putative Mss4-interacting motifs m1, m2, and m3 are shown. B, substitution of the indicated Rab15 amino acid residues with the corresponding Rab5 peptide sequence generates the Rab15-T22N mutants m1, m2, and m3. Asterisks indicate the single mutations K46L and K48Q in m2. C, -galactosidase reporter activity (means ± S.E. of triplicate experiments) produced by interactions between Mss4, T22N, N121I, and the indicated single and double Rab15-T22N mutants. Significant differences were observed between experimental and control conditions (one-way ANOVA, p < 0.001) as described in the text.

Structural analysis of the corresponding m2 motifs in Rab3a and Sec4p suggests that the lysine residues, at positions 46 and 48 of Rab15, probably reside on the surface of Rab15 consistent with a role in binding Mss4 (42). To test this experimentally, we generated two additional mutations in Rab15-T22N in which Lys-46 and Lys-48 were substituted with leucine and glutamine, respectively (i.e. the corresponding residues of Rab5) and assayed Mss4 binding using a yeast two-hybrid assay. -Galactosidase reporter activity was comparable in strains expressing Mss4 with Rab15-T22N or Rab15-K46L. Conversely, binding is disrupted between Mss4 and Rab15-K48Q indicating that lysine at position 48 is critical for the interaction between Rab15 and Mss4.

K48Q Counters the Inhibitory Effect of Rab15 on Early Endocytosis-- We demonstrated previously that wild type HArab15 binds and hydrolyzes GTP and reduces receptor-mediated endocytosis when transiently expressed in Trvb-1 and baby hamster kidney cells (27). We suspect that Rab15 mutants may exert their effect on endocytic trafficking by sequestering effectors or by blocking essential Rab-effector interactions. If Mss4 functions as a chaperone stimulating GDP release and stabilizing Rab15 in a nucleotide-free state, overexpression of Rab15 mutants, which do not bind Mss4, would be predicted to suppress the inhibitory phenotype of wild type Rab15 on endosomal trafficking and endosome fusion (27). To test this, we compared the internalization of ¹²⁵I-labeled transferrin (¹²⁵I-Tfn) in HeLa cells transiently expressing HA epitope-tagged forms of Rab15 (26), which bind Mss4 (wild type HArab15, HArab15-K46L, and HArab15-m3), with cells expressing the mutant HArab15-K48Q, which does not bind Mss4 (Fig. 4). Internalization studies were not performed with HArab15-m1, due to technical limitations associated with poor expression of this mutant. Transfected HeLa cells were depleted of endogenous Tfn, incubated with ¹²⁵I-Tfn for 1 h on ice, followed by extensive washing to remove unbound ¹²⁵I-Tfn. The cells were then incubated at 37 °C for 15 min to allow internalization of ¹²⁵I-Tfn. Expression of wild type HArab15 in HeLa cells resulted in a 44% reduction in the maximal level of internalized ¹²⁵I-Tfn relative to mock-transfected cells, consistent with our previous studies (27). A comparable reduction in the level of internalized ¹²⁵I-Tfn was observed in the transiently expressing wild type HArab15 of HeLa harboring the mutations K46L or m3 (38-42%, respectively). However, cells transiently expressing HArab15-K48Q internalized significantly more ¹²⁵I-Tfn (50%) than cells expressing wild type HArab15 or HArab15-m3. Western analysis indicates that these differences are not a consequence of discernible differences in the relative amount of endogenous TfR or exogenously expressed wild type and mutant HArab15 (Fig. 4B). Similarly, Western analysis confirmed that endogenous levels of Mss4 were not affected by the expression of wild type and mutant forms of HArab15 (data not shown). Taken together, these data suggest that the mutation K48Q counters the inhibitory effect of wild type Rab15 on receptor-mediated endocytosis.

View larger version (42K): [in this window] [in a new window]   Fig. 4.   K48Q counters the inhibitory effect of wild type HArab15 on receptor-mediated endocytosis. A, HeLa cells were transiently transfected with wild type (wt) HArab15 and the indicated mutants, depleted of endogenous Tfn and incubated in medium containing 125I-Tfn for 1 h at 4 °C. The cells were extensively washed to remove unbound ligand and incubated at 37 °C for 15 min to promote the internalization of bound 125I-Tfn. Following internalization, the cells were washed to remove residual surface associated 125I-Tfn and counted to quantitate total levels of internalized 125I-Tfn as described previously (27). All values represent means of triplicate experiments and are expressed as a percent of mock-transfected cells. B, Western analysis (Wn) using antibodies (Ab) against HArab15 and TfR indicate negligible differences in the relative amounts of transiently expressed HArab15 and endogenous TfR. Con represents mock-transfected cells. Significant differences observed between experimental conditions described in the text were verified using a one-way ANOVA with a Neumann-Keuls post hoc test (p < 0.05) and are indicated by an asterisk.

Mss4 has been reported to interact with a specific subset of Rabs that regulate distinct steps in exocytosis (29). To ensure that the observed increase in internalized ¹²⁵I-Tfn in cells expressing HArab15-K48Q does not result from a corresponding increase in the trafficking of the TfR through exocytic compartments, we directly compared the effect of this mutant with wild type HArab15 on the relative amount of cell-surface associated TfR. HeLa cells transiently overexpressing wild type HArab15, HArab15-m3, HArab15-K46L, or the mutant HArab15-K48Q (which does not bind Mss4) were depleted of endogenous Tfn and incubated on ice for 30 min to reduce endocytic trafficking of the TfR. The cells were subsequently incubated with ¹²⁵I-Tfn for 1 h to allow binding of the ligand to surface-associated TfR and washed with PBS, and the level of cell-associated ¹²⁵I-Tfn was determined. Comparable levels of surface-bound ¹²⁵I-Tfn were detected in cells overexpressing HArab15, HArab15 m3, and the non-Mss4 binding mutant HArab15-K48Q (Fig. 5). Interestingly, a 2-fold increase in the amount of surface-associated ¹²⁵I-Tfn was observed in cells expressing HArab15-K46L relative to cells expressing wild type HArab15 or the mutants m3 and K48Q (Fig. 5). Taken together, these data indicate that interaction between Mss4 and Rab15 specifically affects early endocytosis rather than other aspects of membrane trafficking including exocytosis.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   HArab15-K46L increases the relative amount of surface TfR. HeLa cells were transiently transfected with wild type (wt) HArab15 and indicated mutants, depleted of endogenous transferrin for 1 h at 4 °C, incubated at 4 °C for 30 min to stop endoctyosis, and subsequently incubated with 125I-Tfn for 1 h. Cells were washed with PBS and surface-associated 125I-Tfn measured in a gamma counter. Con denotes control, mock-transfected cells. Significant differences were observed between experimental and control conditions (one-way ANOVA, p < 0.001) as described in the text.

We reported previously (27) that wild type HArab15 reduces the level of homotypic early endosome fusion in vitro. To determine whether interactions with Mss4 modulate Rab15 activity at the level of endosome fusion, we compared the effect of overexpressing forms of HArab15 that bind Mss4 with the mutant HArab15-K48Q on homotypic early endosome fusion. Endosomal fractions labeled with either biotinylated horseradish peroxidase or avidin were mixed with cytosolic fractions prepared from HeLa cells overexpressing wild type and mutant HArab15, under conditions that modulate membrane fusion of the labeled endosomal populations. Homotypic endosome fusion is monitored by the immunoisolation of biotinylated horseradish peroxidase-avidin complexes, which are assayed for avidin activity. Expression of wild type HArab15 results in 13.3% reduction in endosome fusion relative to control untransfected cells (Fig. 6). Similarly, expression of HArab15-m3 and HArab15-K46L (Rab15 mutants which bind Mss4) results in a 17.0 and 24.9% reduction, respectively, in endosome fusion relative to control conditions. In contrast, expression of HArab15-K48Q reverses the inhibitory phenotype of wild type HArab15 resulting in a 5% increase in endosome fusion relative to control cells (Fig. 6). Thus, interactions with Mss4 modulate the inhibitory phenotype of wild type Rab15 on early endosome fusion in vitro, consistent with our observations on receptor-mediated endocytosis.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   K48Q counters the inhibitory effect of HArab15 on early endosome fusion in vitro. Early/sorting endosomes labeled with biotinylated horseradish peroxidase or avidin were prepared from HeLa cells and incubated with cytosol prepared from untransfected control cells or HeLa cells transiently overexpressing wild type (white bar) or mutant (black bars) HArab15 as indicated. All values represent the means of triplicate experiments and are normalized with respect to protein concentration. Values are expressed as a percentage of base-line fusion observed using cytosols prepared from untransfected HeLa cells ± S.E.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Early/sorting endosomes are the nexus for several membrane trafficking pathways (reviewed in Refs. 1 and 2). Accordingly, trafficking through this compartment is tightly regulated and relies on a fine balance involving multiple Rabs functioning to either facilitate or inhibit membrane trafficking at discrete membrane transport steps. Several recent studies (reviewed in Refs. 3-5) indicate that Rabs function largely by promoting the recruitment of effectors, which in turn regulate distinct trafficking events. Given the inhibitory phenotype of Rab15 on early endocytosis and the functional relationship between Rab-effector interactions, we examined the effector-binding properties of Rab15. Rabaptin-5 and rabenosyn-5 have been reported previously (15, 25) to bind Rab4 and Rab5 through distinct binding domains. The multivalent binding properties of these Rab5 effectors and the opposing effects of Rab15 and Rab5 on early endocytic trafficking in cultured cells are reconcilable with a model in which Rab15 and Rab5 compete for shared effectors (27). No interaction was detected between Rab15 and rabaptin-5 or rabex-5 in a yeast two-hybrid binding assay. Although rabaptin-5 and rabex-5 do not bind Rab15 directly, we cannot disregard the possibility that they may functionally interact via additional unidentified binding partners. For example, the recruitment of rabaptin-5 with Rab5-GTP on early endosomes is promoted by the guanine nucleotide exchange activity of rabex-5 (34). Recently, however, rabex-5 was also shown to bind early endosomes and clathrin-coated vesicle membranes independently of rabaptin-5 and Rab5-GTP, indicating that additional binding partners for rabex-5 exist (34). By using a yeast two-hybrid binding assay, we detected a potential interaction between Rab15 and rabenosyn-5. Although it is tempting to speculate on the impact of Rab15 interactions with rabenosyn-5, the physiological significance of this interaction will require further verification.

This study identified Mss4 as a direct binding partner for Rab15. Mutational analysis indicates that lysine at position 48 (K48Q) is important for the binding of Rab15 to Mss4. Expression of HArab15-K48Q partially reverses the inhibitory phenotype of wild type HArab15 on receptor-mediated endocytosis and homotypic early endosome fusion in vitro without altering the relative amount of cell surface-associated TfR, indicating that interactions with Mss4 specifically modulate the inhibitory effect of Rab15 on early endocytosis. Functional analysis of Mss4 and its yeast homologue Dss4p demonstrated previously (35, 36) that these proteins promote GDP removal and stabilization of their target Rabs in guanine nucleotide-free states. Accordingly, our binding studies confirm that Mss4 preferentially interacts with GDP-bound and nucleotide-free Rab15 mutants. In contrast to the GEF rabex-5, Mss4 does not efficiently promote GTP recruitment (38). Therefore, Mss4 appears to function as a chaperone stabilizing its target Rabs for subsequent interactions with additional factors that promote GTP binding and Rab activation. Indeed, our functional data demonstrating that expression of HArab15-K48Q cannot fully compensate for the inhibitory effect of wild type Rab15 on receptor-mediated endocytosis suggest that additional factors contribute to Rab15 activation. In this context, Mss4 may facilitate these interactions and increase the relative rate of Rab15 activation.

The three-dimensional structure of Mss4 and two of its target GTPases (Rab3a and Sec4p) has been determined; however, identification of the residues mediating binding of Mss4 and its yeast homologue Dss4p to their target Rabs remains elusive (43, 44). Structural analysis of Rab3a and Sec4p demonstrated that lysine residues at positions 46 and 48 reside on their surfaces accessible to effector molecules (40, 43). Our mutational analysis of Rab15 reveals that lysine 48 is essential for the interaction between Rab15 and Mss4. Conversely, mutation of the lysine at position 46 results in minimal loss of Mss4 binding to Rab15. This mirrors the observations in Rab3a where the corresponding mutation K60A did not impact the ability of Mss4 to promote GDP release from Rab3a in vitro (43). Modeling of the K46L and K48Q mutations on the surface of Rab3a and Sec4p does not impart significant conformational changes in the predicted structure of these proteins,³ supporting the premise that the absence of Mss4 binding to HArab15-K48Q is not a result of major perturbations in the overall structure of Rab15.

Interestingly, an increase in the relative amount of TfR was observed on the surface of cells expressing HArab15-K46L. Yet overexpression of HArab15-K46L did not counter the inhibitory effect of wild type Rab15 on receptor-mediated endocytosis or in vitro endosome fusion. We reconcile this observation with our earlier studies showing that overexpression of constitutively inactive GDP-bound HArab15-T22N in cultured cells promotes recycling of the TfR directly from early/sorting endosomes (27). A similar mechanism may account for the relative increase in surface-associated TfR observed in cells expressing HArab15-K46L reported in this study.

Our observation that Mss4 binds Rab15 and regulates its activity during endocytosis is wholly consistent with the emerging concept that exocytosis and endocytic trafficking are functionally linked through shared components. For example, the Rab5 effector EEA1 has been shown to bind syntaxin 6, a SNARE implicated in trafficking between the trans-Golgi network and early endosomes (45). In addition to binding Rab5, rabex-5 and rabaptin-5 directly interact with Rab33b, a GTPase implicated in retrograde transport from the Golgi to the endoplasmic reticulum (46, 47). Recently, rabaptin-5 was reported to form a complex with rabphilin-3, a protein implicated in the control of exocytosis and endocytosis in the nerve terminal (48-51). Mutational analysis of rabphilin identified a point mutation (V61A) that disrupted binding to the exocytic GTPase Rab3a, but not rabaptin-5. Moreover, expression of rabphilin-V61A in cultured cells promoted receptor-mediated endocytosis, implicating a role for rabphilin-rabaptin-5 interactions in early endocytic events (47). Similarly, several genetic and molecular studies support a dual role for the synaptic vesicle protein synaptotagmin 1 as a calcium-responsive trigger that drives neurotransmitter release as well as synaptic vesicle recycling within the nerve terminal (52-56).

In conclusion, our observation that Mss4 binds Rab15 and modulates its function in vivo is, to the best of our knowledge, the first report of an interaction between Mss4 and an endocytic protein. Our studies are consistent with an emerging model in which exocytic and endocytic pathways are functionally linked, expanding the role of Mss4 as a chaperone in early endocytic trafficking. Although our results indicate that Rab15 function involves at least one unique effector interaction, given the opposing effects of Rab15 and Rab5 on endocytic trafficking, we anticipate that the transport steps regulated by these GTPases will likely intersect at some point in terms of a shared effector or accessory protein. However, the nature of these interactions remains to be determined and will require the functional analysis of additional interacting partners for Rab15.

	FOOTNOTES

^* This work was supported in part by National Science Foundation Grant IBN-9974517 (to L. A. E.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed. Tel.: 409-772-1942; Fax: 409-772-2789; E-mail: Lisa.Elferink@utmb.edu.

Published, JBC Papers in Press, June 24, 2002, DOI 10.1074/jbc.M205101200

² D. J. Strick, D. M. Francescutti, Y. Zhao, and L. A. Elferink, unpublished observations.

³ D. Strick and L. Elferink, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: PI3P, phosphatidylinositol 3-phosphate; GEF, guanine nucleotide exchange factor; HA, hemagglutinin; Mss4, mammalian suppressor of Sec4; SNARE, soluble NSF attachment protein receptor; Tfn, transferrin; TfR, transferrin receptor; X-gal, 5-bromo-4-chloro-3-indolyl--D-galactopyranoside; GTPS, guanosine 5'-3-O-(thio)triphosphate; ANOVA, analysis of variance; PBS, phosphate-buffered saline; NiNTA, nickel-nitrilotriacetic acid; GDPS, guanosine 5'-O-2-(thio)diphosphate.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. J. Strick and L. A. Elferink Rab15 Effector Protein: A Novel Protein for Receptor Recycling from the Endocytic Recycling Compartment Mol. Biol. Cell, December 1, 2005; 16(12): 5699 - 5709. [Abstract] [Full Text] [PDF]

				C. Cans, B. J. Passer, V. Shalak, V. Nancy-Portebois, V. Crible, N. Amzallag, D. Allanic, R. Tufino, M. Argentini, D. Moras, et al. Translationally controlled tumor protein acts as a guanine nucleotide dissociation inhibitor on the translation elongation factor eEF1A PNAS, November 25, 2003; 100(24): 13892 - 13897. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/36/32722    most recent M205101200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Strick, D. J. Articles by Elferink, L. A. Search for Related Content PubMed PubMed Citation Articles by Strick, D. J. Articles by Elferink, L. A. Social Bookmarking                      What's this?