Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M402183200 on May 13, 2004

J. Biol. Chem., Vol. 279, Issue 30, 31409-31418, July 23, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/30/31409    most recent
M402183200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mattera, R.
Right arrow Articles by Bonifacino, J. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mattera, R.
Right arrow Articles by Bonifacino, J. S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

The Trihelical Bundle Subdomain of the GGA Proteins Interacts with Multiple Partners through Overlapping but Distinct Sites*

Rafael Mattera, Rosa Puertollano, William J. Smith, and Juan S. Bonifacino{ddagger}

From the Cell Biology and Metabolism Branch, NICHD, National Institutes of Health, Bethesda, Maryland 20892

Received for publication, February 27, 2004 , and in revised form, May 3, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The Golgi-localized, {gamma}-adaptin ear-containing, ARF-binding (GGA) proteins are monomeric clathrin adaptors that mediate the sorting of cargo at the trans-Golgi network and endosomes. The GGAs contain four different domains named Vps27, Hrs, Stam (VHS); GGAs and TOM1 (GAT); hinge; and {gamma}-adaptin ear (GAE). The VHS domain recognizes transmembrane cargo, whereas the hinge and GAE regions bind clathrin and accessory proteins, respectively. The GAT domain is a polyfunctional module that interacts with various partners including the small GTPase ARF, the endosomal fusion regulator Rabaptin-5, ubiquitin, and the product of the tumor susceptibility gene 101 (TSG101). Previous x-ray crystallographic analyses showed that the GAT region is composed of two subdomains, an N-terminal helix-loop-helix containing the ARF binding site, and a C-terminal triple {alpha}-helical (trihelical) bundle. In this study, we define the Rabaptin-5 binding site on the GGA1-GAT domain and its relationship to the binding sites for ubiquitin and TSG101. Our observations show that Rabaptin-5, ubiquitin, and TSG101 bind to overlapping but distinct binding sites on the trihelical bundle. The different GAT binding partners engage in both competitive and cooperative interactions that may be important for the function of the GGAs in protein sorting.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The Golgi-localized, {gamma}-ear-containing, ARF-binding (GGA)1 proteins are a family of monomeric clathrin adaptors that bind to membranes in an ARF-dependent fashion and mediate the sorting of specific transmembrane cargo at the trans-Golgi network and endosomes (reviewed in Ref. 1). Three GGAs (i.e. GGA1, GGA2, and GGA3) have been described in humans, and 1–3 GGAs have been described in other organisms from yeast to mammals. The GGAs have a modular structure consisting of four domains or regions named Vps27, Hrs, Stam (VHS), GGAs and TOM1 (GAT), hinge, and {gamma}-adaptin ear (GAE) (Fig. 1A). The VHS domain binds acidic cluster-dileucine sorting signals present in the cytosolic domains of transmembrane cargo proteins such as the mannose 6-phosphate receptors (2–4), sortilin (4, 5), the lipoprotein receptor-related protein 3 (4), and {beta}-secretase (6). The GAT domain was initially shown to interact with the GTP-bound form of ARF (ARF-GTP) (7, 8), which is critical for the regulated recruitment of the mammalian GGAs to membranes. The long, unstructured hinge region has peptide motifs that interact with the terminal domain of the clathrin heavy chain (9). Finally, the GAE domain binds to a set of accessory proteins including Rabaptin-5 (3, 10), Enthoprotin/EpsinR/Clint (11–14), {gamma}-synergin (15, 16), p56 (17), NECAP1, NECAP2, and aftiphilin (18), all of which contain peptide motifs fitting the {Psi}G(P/D/E)({Psi}/L/M) consensus (where {Psi} is an aromatic residue) (18).



View larger version (39K):
[in this window]
[in a new window]
  		FIG. 1.
Structure of the N-terminal hook and trihelical bundle subdomains in GGA1-GAT. A, domain organization of human GGA1. B, ribbon representation of the GGA1-GAT domain showing the side chains of residues important for the binding of its different partners. The left and right images show enlarged views of the N-terminal hook and the C-terminal bundle, respectively. C, surface representation of the GGA1-GAT domain. Highlighted in yellow (left image) are the residues in the N-terminal hook that are important for ARF1 binding. These include the residues analyzed in the experiments shown in Fig. 3A as well as four others (Leu178, Lys183, Leu190, and Val201) characterized in Refs. 27 and 29. Highlighted in the right image are the residues in the trihelical bundle involved in binding to Rabaptin-5, ubiquitin, and TSG101. Shown in red, blue, and green are the residues involved in the interaction with Rabaptin-5 only, with both Rabaptin-5 and ubiquitin, or with all three partners, respectively. Leu281 was assayed only against Rabaptin-5 in this study and may represent a blue residue (important for binding of both Rabaptin-5 and ubiquitin) based on the lack of ubiquitin binding observed after substitution of the corresponding residue in GGA3-GAT (Leu280) (25). The GGA1-GAT residue Asp285 (not analyzed in this study) may represent another blue residue based on the reported lack of Rabaptin-5 binding following its substitution by Arg (37) and on the alteration in ubiquitin binding arising from mutation of the corresponding residue in GGA3-GAT (Asp284) (25). Surface and ribbon representations were prepared with GRASP (42), Bobscript (43, 44), and Raster3D (45) using the structure of GGA1-GAT (Protein Data Bank accession code 1NAF [PDB] ).

 
Recent studies have revealed that, like the GAE domain, the GAT domain of the GGAs has multiple binding partners. In addition to ARF, the GAT domains of GGA1 and GGA2, but not GGA3, interact with Rabaptin-5 (19). Since the GAE domain also binds Rabaptin-5, this indicates that the interactions of GGA1 and GGA2 with Rabaptin-5 are divalent. Whereas the GAE domain binds the {Psi}G(P/D/E)({Psi}/L/M)-fitting sequence FG-PLV (residues 439–443) in the central unstructured region of Rabaptin-5, the GAT domain of GGA1 and GGA2 binds to a C-terminal region of Rabaptin-5 predicted to form two coiledcoils (referred to as C2-1 and C2-2) (19). Rabaptin-5 is in turn a divalent Rab4/Rab5 effector that exists in a complex with Rabex-5, a guanine nucleotide exchange factor for Rab5 (20–23). The Rabaptin-5-Rabex-5 complex participates in endosomal tethering/fusion events, and its interaction with the GGAs could promote tethering and fusion of GGA-coated carrier vesicles with endosomes. Furthermore, the GAT domains of the GGAs, particularly GGA1 and GGA3, bind ubiquitin (24–26). Finally, the tumor susceptibility gene 101 product (TSG101), which is a ubiquitin-binding component of the machinery involved in multivesicular body formation, also interacts with GGA-GAT domains (24). These observations indicate that the GAT domain of some GGAs, such as GGA1, has the ability to bind at least four partners: ARF-GTP, Rabaptin-5, ubiquitin, and TSG101.

The three-dimensional structure of the GGA1-GAT domain has recently been solved by x-ray crystallography (27–30). It consists of an elongated structure having an N-terminal helix-loop-helix "hook" subdomain and a C-terminal triple {alpha}-helical bundle subdomain (Fig. 1B). Interestingly, this "trihelical" subdomain resembles the regulatory N-terminal domain of the SNAREs, syntaxin 1a and 6 (31). The GAT domains of two other proteins, target of myb1 (TOM1) and TOM1-like 1 (TOM1-L1) do not contain the N-terminal hook subdomain but are also predicted to have a similar trihelical subdomain based on sequence similarity. Biochemical and structural analyses have shown that residues in the hook subdomain of GGA-GAT bind ARF-GTP (9, 27–30, 32, 33). The trihelical subdomain of GGA-GAT, on the other hand, has been shown to bind Rabaptin-5 (30) and to contain residues that bind ubiquitin and TSG101 (24, 25). The binding site for Rabaptin-5 within the trihelical subdomain and its relationship to that of the other GAT binding partners, however, remain to be delineated. It also remains to be determined whether Rabaptin-5 binds the trihelical domains of TOM1, TOM1-L1, and the syntaxins 1a and 6. Here we demonstrate that Rabaptin-5, ubiquitin, and TSG101 bind to overlapping but distinct sites on the trihelical subdomain of GGA1-GAT and that Rabaptin-5 does not interact with the trihelical domains of other proteins. The different GAT binding partners engage in both competitive and cooperative interactions that may be important for the function of the GGAs in protein sorting.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
DNA Constructs
pGAD and pGBT9 Constructs—Mutations in pGAD424-human GGA1-GAT, pGAD424-human GGA1-VHS+GAT, and pGBT9-human GGA1-GAT were introduced using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) using primers (35-mers) purchased from Sigma or Qiagen Inc. (Valencia, CA). Mutations in pGBT9-human GGA3-GAT were introduced using the same procedure. The pGBT9 constructs encoding human GGA1-GAT (residues 148–314), human GGA2-GAT (residues 164–330), and human GGA3-GAT (residues 147–313) as well as the pGAD constructs encoding full-length human Rabaptin-5 and the 551–661 and 551–862 fragments of this protein were described in Ref. 19. An EcoRI/SalI fragment encoding human GGA1-VHS+GAT (residues 5–314, containing also a M233V substitution) was obtained by digestion from pGFP-GGA1-VHS+GAT and subcloned into the corresponding sites of pGAD424 and pGBT9. The pGBT9 constructs encoding human ARF1 Q71L and human ARF1 T31N (subcloned into EcoRI/SalI sites) were described in Ref. 8. The pGBT9 constructs encoding human ubiquitin (subcloned into EcoRI and BamHI sites) and human TSG101 (subcloned into EcoRI and SalI sites) were described in Ref. 24). An EcoRI/SalI cDNA fragment encoding human TOM1-GAT (residues 215–299) was obtained by digestion of pGFP-TOM1-GAT (Ref. 9) and subcloned into the corresponding sites of pGBT9. The EcoRI/SalI and EcoRI/XhoI cDNA fragments encoding TOM1-L1-GAT (residues 200–284) and full-length synthaxin were obtained by PCR amplification from human brain and human liver cDNA libraries, respectively, and subcloned into the EcoRI/SalI sites of pGBT9.

pET-28a(+) Constructs—An EcoRI/XhoI fragment encoding human Rabaptin-5-(551–862) was obtained by PCR amplification from pGAD-myc-Rabaptin-5 (19), digested, and subcloned into the corresponding sites of pET-28a(+) (Novagen, Madison, WI). The His6-tagged Rabaptin-5-(551–862) fragment was expressed in Escherichia coli strain BL21(DE3) (Novagen) and purified using Ni2+-nitrilotriacetic acid columns (Qiagen).

pGEX Constructs—EcoRI/SalI fragments encoding human GGA1-GAT, human GGA1-VHS+GAT, human GGA3-GAT, and human GGA3-VHS+GAT were subcloned into the corresponding sites of pGEX-5X-1 (Amersham Biosciences).

Antibodies
The mouse monoclonal anti-Rabaptin-5 antibody was purchased from Transduction Laboratories (Lexington, KY). The rabbit anti-GGA1 and anti-Rabex-5 antisera were kind gifts from M. S. Robinson (University of Cambridge, UK) and M. Zerial (Max Planck Institute, Dresden, Germany), respectively. The rabbit anti-GST antiserum was described in Ref. 34.

Yeast Two-hybrid Analysis
The AH109 yeast reporter strain was maintained on YPD agar plates. Transformation of AH109 cells with pGAD- and pGBT9-based constructs by the lithium acetate method was performed following the instructions for the Matchmaker two-hybrid system (Clontech). Double transformants were isolated on synthetic defined medium lacking leucine and tryptophan. Interaction of fusion proteins was monitored by activation of HIS3 gene transcription following plating on medium lacking histidine, leucine, and tryptophan, in the presence or absence of the indicated concentrations of the competitive inhibitor of the HIS3 protein 3-amino-1,2,4-triazole (3-AT). The use of plates with different concentrations of 3-AT provided a semiquantitative evaluation of the strength of the specific interactions under study. The presence of 3-AT also minimized nonspecific interactions between constructs and prevented background growth.

Pull-down Assays Using Beads with Immobilized Ubiquitin or GST Fusions
The preparation of bovine brain cytosol and the general conditions used in the pull-down protocols by immobilized GST fusion proteins (Fig. 7B) were described previously (19). The effect of ubiquitin on the pull-down of Rabaptin-5 from bovine brain cytosol by GGA1-GAT (Fig. 7B) was analyzed using a GST-GGA1-VHS-GAT fusion, since we did not observe binding of endogenous Rabaptin-5 to the GST-GGA1-GAT alone (results not shown). This difference may reflect a spacer effect that improves the folding and/or the accessibility of the GAT domain in the fusion protein to Rabaptin-5. Consistent with the results obtained in the yeast two-hybrid analysis (Fig. 5B and Ref. 19), we did not observe binding of Rabaptin-5 to either GST-GGA3-VHS-GAT or GST-GGA3-GAT alone. The glutathione-Sepharose beads (Amersham Biosciences) with immobilized GST fusions (30 µg) were washed and preincubated with 0–100 µg of His6-human ubiquitin (Sigma) or irrelevant proteins for 40 min at 4 °C in a final volume of 750 µl of 15 mM HEPES, pH 7.0, 75 mM NaCl, and 0.25% Triton X-100 supplemented with protease inhibitors (EDTA-free CompleteTM; Roche Applied Science) (binding buffer). At the end of this period, 600 µl of bovine brain cytosol diluted in binding buffer (1.5 mg of protein) were added, followed by further incubation for 90 min at 4 °C. The beads were washed, and the bound proteins were eluted and subjected to SDS-PAGE and immunoblotting with anti-Rabaptin-5 or anti-His tag antibodies as previously described (19).



View larger version (23K):
[in this window]
[in a new window]
  		FIG. 7.
Competitive and cooperative interactions in the binding of Rabaptin-5 and ubiquitin to the GGA1-GAT domain; interaction of ubiquitin-agarose with an endogenous complex containing GGA1, Rabaptin-5, and Rabex-5. A, GST fusion proteins were preincubated with the indicated amounts of His6 -Rabaptin-5-(551–862), followed by incubation with either ubiquitin-agarose or protein A-agarose beads. The proteins bound to the washed beads were eluted and subjected to SDS-PAGE and immunoblotting (IB) with anti-GST antiserum. The apparent molecular mass of the GST-{beta}3 ear fusion ({beta}3B isoform) is 54 kDa. B, GST fusion proteins were immobilized onto glutathione-Sepharose and preincubated with the indicated amounts of His6-ubiquitin followed by incubation with bovine brain cytosol. The proteins bound to the washed beads were eluted and subjected to SDS-PAGE and immunoblotting using anti-Rabaptin-5 or anti-His antibodies. Preincubation with His6-ubiquitin did not result in an artificial pull-down of Rabaptin-5 by the GST-{beta}3 ear construct; similarly, preincubation with His6-ubiquitin did not affect the pull-down of endogenous Rabaptin-5 by GST-GGA1-GAE (results not shown). C, GST fusion proteins were immobilized and preincubated with or without His6-ubiquitin as described for B, followed by incubation with recombinant His6-Rabaptin-5-(551–862) (apparent molecular mass ~40 kDa) and immunoblotting using anti-His antibody. D, ubiquitin-agarose or protein-A-agarose beads were incubated with bovine brain cytosol. The material bound to the washed beads was eluted and subjected to SDS-PAGE and immunoblotting with the indicated antisera. For technical details, see "Experimental Procedures." BSA, bovine serum albumin.

 



View larger version (85K):
[in this window]
[in a new window]
  		FIG. 5.
Identification of GGA1-GAT and GGA3-GAT mutations that result in the loss and gain of Rabaptin-5 binding, respectively. Experiments were performed as described in the legend to Fig. 4 using the single, double, or triple substitution mutants depicted in the panels.

 
The effect of Rabaptin-5 on the ubiquitin-GAT domain interaction (Fig. 7A) was studied with ubiquitin-agarose beads and recombinant GGA1-VHS-GAT and Rabaptin-5 fragments. Aliquots representing 1.5 µg of GST fusions were preincubated in the presence or absence of increasing concentrations of His6-Rabaptin-5-(551–862) for 60 min at 4 °C in a final volume of 500 µl of binding buffer supplemented with 0.15% bovine serum albumin. At the end of this period, the preincubation mixtures were added to 20 µl of ubiquitin-agarose (Affinity Bioreagents, Golden, CO) or protein A-agarose (Sigma) followed by incubation for 4 h at 4 °C. Beads were subsequently washed three times with binding buffer without bovine serum albumin, and the bound proteins were eluted and subjected to SDS-PAGE and immunoblotting with rabbit anti-GST antiserum.

Interaction of endogenous complexes with ubiquitin (Fig. 7D) was analyzed by incubating 40 µl of ubiquitin-agarose or protein A-agarose beads with 1.5 ml of bovine brain cytosol (15 mg of protein diluted in binding buffer) for 4 h at 4 °C. Beads were subsequently washed three times with binding buffer, and the eluted proteins were subjected to immunoblotting using anti-Rabaptin-5, anti-GGA1, or anti-Rabex-5 antisera.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Relationship between the Binding Sites for ARF1 and Rabaptin-5 in the GGA1-GAT Domain—We introduced various amino acid substitutions in the N-terminal helix-loop-helix "hook" of GGA1-GAT, which contains residues critical for ARF1 binding, and in a hydrophobic patch on the surface of the C-terminal trihelical bundle of GGA1-GAT (Fig. 1B; yellow and red dots, respectively, in Fig. 2). The mutations introduced in the N-terminal hook (L182A, N194A, I197A, K198A, M200A, D204A, and Q205A) were selected based on previous mutational analyses performed by us and by other laboratories (9, 32, 33) and on the analysis of the structure of the GGA1-GAT domain (27–30). The experiments aimed at examining GAT domain-ARF1 interactions were performed using constructs expressing the constitutively active, GTP-locked, ARF1 Q71L mutant and the inactive, GDP-locked, ARF1 T31N mutant as a negative control. We previously reported that the C-terminal Rabaptin-5-(551–739) fragment containing the C2-1 and C2-2 coiled-coil regions and excluding the Rab5 interaction site binds to GGA1-GAT and GGA2-GAT and that the Rabaptin-5-(551–661) fragment (including only the C2-1 region) is sufficient for this interaction (19). In order to detect relative differences in avidity, we evaluated the effect of GAT domain substitutions on the binding of Rabaptin-5 using three different constructs expressing either full-length Rabaptin-5 or the 551–661 and 551–862 fragments of this protein. The effects of the substitutions in the N-terminal hook and C-terminal region of GGA1-GAT on the binding of ARF1 and Rabaptin-5 were analyzed by the yeast two-hybrid system using different concentrations of 3-AT, a competitive inhibitor of the His3 protein, to evaluate the avidity of the interactions. Consistent with previous structural analyses (9, 27–30, 32, 33), the results in Fig. 3A (left panels) showed the importance of GGA1-GAT residues Leu182, Asn194, Ile197, Lys198, Met200, and Asp204 in interactions with GTP-bound ARF1. The role of the GGA1-GAT residue Lys198 was evident only when assaying the interactions under more stringent conditions (i.e. in the presence of 2 mM 3-AT), consistent with the partial reduction in the ability of this mutant to pull down ARF1 in vitro (27). Substitution of some of the residues proposed to adjoin the ARF binding site in the GGA1-GAT domain (28) (namely Met200 and Asp204 but not Gln205) resulted in a similar decrease in the interaction with ARF1 Q71L when the yeast two-hybrid analysis was performed in the presence of 2 mM 3-AT (Fig. 3A, left panels). Substitution of residues that are important for the binding of ARF1 to the N-terminal hook of GGA1-GAT did not affect the interaction of this domain with the different Rabaptin-5 constructs (Fig. 3B, left panels). An exception to this general pattern was the GGA1-GAT N194A mutant, which exhibited a weaker binding to full-length Rabaptin-5 in plates containing 2 mM 3-AT (Fig. 3B, left panels) and did not bind this construct in assays carried out in the presence of 15 mM 3-AT (not shown); however, interactions of this mutant with the smaller Rabaptin-5 fragments were normal under all conditions (Fig. 3B, left panels; results not shown).



View larger version (48K):
[in this window]
[in a new window]
  		FIG. 2.
Sequence alignment of the GAT domains from GGA1 and TOM1/TOM1-L1 proteins. The sequences shown correspond to human proteins. Identical and conserved residues in all five GAT domains are shown in red and green, respectively; residues identical in the GAT domains of all human GGAs but not in TOM1/TOM1-L1 are shown in blue. Residues in helices 1–4 of the GGA1-GAT structure (bars above the sequences) are shown as defined in Ref. 30. The yellow and red dots indicate the residues important for ARF1 and Rabaptin-5 binding, respectively, as analyzed in the experiments shown in Fig. 3. The asterisks show additional residues surrounding the Rabaptin-5 binding site in helices {alpha}3 and {alpha}4 of GGA1-GAT that are not conserved in GGA3-GAT and were analyzed in Figs. 4 and 5.

 



View larger version (86K):
[in this window]
[in a new window]
  		FIG. 3.
Effect of mutations in the helix-loop-helix "hook" (N terminus) and the C-terminal trihelical bundle (C terminus) of GGA1-GAT on the binding to ARF1 Q71L and Rabaptin-5. A, the indicated substitutions were introduced in the Gal4 transcription AD construct pGAD-GGA1-GAT, whereas the cDNAs encoding ARF1 Q71L and ARF1 T31N were subcloned in the Gal4 DNA-BD vector pGBT9. This specific design was required, since no interactions could be evidenced when the wild type GGA1-GAT and ARF1 Q71L cDNAs were subcloned in BD and AD vectors, respectively (results not shown). B, the indicated substitutions were introduced in pGBT9-GGA1-GAT, whereas the Rabaptin-5 constructs were subcloned in the AD vector pGAD. Co-transformants shown in both panels were plated on medium without histidine (–His), to detect HIS3 reporter gene activation upon interaction of constructs, and on medium with histidine (+His) as a control for loading and growth of the co-transformants. The –His plates containing 3-AT (a competitive inhibitor of the His3 protein) provided increased stringency and were used to compare the relative strength of the specific interactions under study. Controls for nonspecific interactions included co-transformation of pGAD constructs with a BD-p53 plasmid as well as co-transformation of the pGBT9 constructs with an AD-SV40 large T-antigen plasmid (T-Ag). Co-transformation with vectors encoding the BD-p53 and AD-SV40 large T-antigen provided a positive control for interactions. Lanes labeled wt show the interactions with the native GGA1-GAT domain.

 
Importantly, the substitution of all residues in the proposed Rabaptin-5 binding site located on the surface of the GGA1-GAT trihelical bundle (28) (Fig. 1, B and C, and red dots in Fig. 2), including the replacement of Arg260 by Glu but not by Ala, affected the interaction with Rabaptin-5 constructs (Fig. 3B, right panels). Whereas the F264A, A267D, and L277A substitutions in GGA1-GAT abolished binding to all of the Rabaptin-5 constructs, the R260E, L281A, and N284A substitutions exhibited weakened interactions with either full-length Rabaptin-5 or its 551–862 fragment but maintained recognition by the shorter fragment (residues 551–661) (Fig. 3B). However, the double substitution R260E/L281A also abolished the interaction with the Rabaptin-5-(551–661) fragment, which exhibits a higher avidity for the GGA1-GAT domain as compared with the full-length or 551–862 constructs (Fig. 3B, right panels). Significantly, the substitutions in the Rabaptin-5 binding site did not affect the interaction of GGA1-GAT with ARF1 Q71L in assays performed in the absence of 3-AT, although a slightly weakened interaction of some mutants could be observed in the presence of 2 mM 3-AT (Fig. 3A, right panels). Taken together, these observations demonstrate the existence of a binding site for Rabaptin-5 in the trihelical bundle of the GGA1-GAT domain. The experiments also show that the binding sites for ARF1 and Rabaptin-5 are largely independent, although the partial effects on Rabaptin-5 binding brought about by the N194A substitution and the weakened interaction with ARF1 Q71L seen with some GGA1-GAT trihelical bundle mutants hint at the possibility of interactions between these two separate binding sites.

Rabaptin-5 Does Not Interact with the GAT Domain of TOM1 or TOM1-L1 or with the Syntaxin 6 Trihelical Bundle—The GAT domains in TOM1 and TOM1-L1 lack the N-terminal helix-loop-helix present in the corresponding domains of the GGAs ({alpha}1 helix and N-terminal portion of {alpha}2) but exhibit sequence similarity in the trihelical bundles (C-terminal portion of {alpha}2 together with {alpha}3 and {alpha}4 helices) (Fig. 2). The absence of the sequence critical for binding of ARF1 to GGA1-GAT in the corresponding domains of TOM1 and TOM1-L1 (region with yellow dots in Fig. 2) is consistent with the inability of TOM1-GAT to associate with the trans-Golgi network in an ARF1-dependent manner (9). Nonetheless, it was of interest to investigate the possible interaction of Rabaptin-5 with the trihelical bundle of TOM1-GAT and TOM1-L1 GAT. However, we could not observe interactions of these GAT domains with any of the Rabaptin-5 constructs that we tested, including the "high avidity binder" fragment encompassing residues 551–661 (results not shown). Significantly, this lack of interaction is consistent with the presence of substitutions in the trihelical bundle of the TOM1-GAT and TOM1-L1-GAT domains in place of residues that are critical for the binding of Rabaptin-5 to GGA1-GAT (red dots in Fig. 2; Fig. 3B). These substitutions include R260Q, F264(L/M), A267(I/L), and L277T (residues are numbered according to the GGA1 sequence; residues between parentheses indicate substitutions in TOM1 followed by those in TOM1-L1) (Fig. 2). Although R260Q and L277T are the less conserved substitutions and likely to contribute more significantly to the lack of interaction of TOM1 and TOM1-L1 with Rabaptin-5, we did not assess their relative importance by introducing gain-of-function mutations in these GAT domains.

Interestingly, the trihelical bundles in the N-terminal domain of the SNAREs syntaxins 1a and 6 show a striking structural resemblance to those in GGA-GAT domains, including the presence of a hydrophobic binding site (for C-terminal SNARE motifs) between the second and third helices of the bundles (27, 28). The topological similarity between this site and the Rabaptin-5 binding site in GGA1-GAT suggested the possibility of interactions between Rabaptin-5 and the trihelical bundles in the N-terminal domain of syntaxins 1a and 6. However, experimental analysis of this hypothesis using the yeast two-hybrid system (BD-syntaxin 6 and AD-Rabaptin-5 constructs) showed that Rabaptin-5 is unable to bind to the trihelical bundle of syntaxin 6 (results not shown).

We have also previously demonstrated that, unlike GGA1-GAT and GGA2-GAT, the GAT domain of GGA3 does not bind to Rabaptin-5 (19). Taken together, these observations highlight the specificity of the binding of Rabaptin-5 to the GAT domains of GGA1 and GGA2, as opposed to a more promiscuous interaction with other GAT domains or similar trihelical bundles.

Basis for the Lack of Interaction of Rabaptin-5 with GGA3-GAT—The elucidation of the structural determinants responsible for the interaction of the trihelical bundles in the GAT domains of GGA1 and GGA2 with Rabaptin-5 should explain the lack of interaction of this protein with GGA3-GAT. Analysis of the residues required for binding of Rabaptin-5 to GGA1-GAT (Arg260, Phe264, Ala267, Leu277, Leu281, and Asn284; red dots in Fig. 2 and assay in Fig. 3B) shows that only Asn284 is replaced in GGA3-GAT (the corresponding residue is Ser283 in GGA3). We examined whether the N284S substitution affected the binding of GGA1-GAT to Rabaptin-5 and also tested the substitution of additional residues that surround the Rabaptin-5 binding site in the {alpha}3 and {alpha}4 helices of GGA1-GAT and that are not maintained in GGA3-GAT (asterisks in Fig. 2). This systematic analysis was also expected to extend the identification of residues that are critical for the binding of Rabaptin-5 to GGA1-GAT. The experiments in Fig. 4 show that, among all the mutants tested, only GGA1-GAT N284S (Fig. 4, A and B) and GGA1-GAT M259K (Fig. 4B) exhibit decreased binding of Rabaptin-5 as compared with that of native GGA1-GAT. We also tested the binding of a GGA1-GAT double mutant including the P261R and N284S substitutions. The GGA1 Pro261 is located at a kink near the middle of the {alpha}3 helix in close proximity to or within the Rabaptin-5 binding site. Interestingly, whereas the GGA1-GAT P261R substitution is inconsequential per se (Fig. 4B), the double mutant GGA1-GAT P261R/N284S showed a further reduction in binding to the Rabaptin-5 constructs compared with the N284S single mutant (see interactions of these mutants with the Rabaptin-5-(551–661) and -(551–862) constructs in Fig. 4B).



View larger version (63K):
[in this window]
[in a new window]
  		FIG. 4.
Binding of Rabaptin-5 constructs to GGA1-GAT mutants containing residues present in helices {alpha}3 and {alpha}4 of GGA3-GAT. Experiments were performed as described in the legend to Fig. 3 using AD-Rabaptin-5 and BD-GGA1-GAT constructs. All of the mutants assayed represent individual replacements except for the E275N/A276S (A) and P261R/N284S (B) double substitutions. The weakened interactions of the GGA1-GAT N284S, M259K, and P261R/N284S mutants were also observed in –His plates without 3-AT.

 
The relative importance of Met259, Pro261, and Asn284 in the binding of Rabaptin-5 to the GGA1-GAT was further investigated by the analysis of additional double and triple mutants. The results showed that, similar to the effect of GGA1-GAT P261R/N284S, the GGA1-GAT M259K/N284S double mutant displays reduced interaction with Rabaptin-5 as compared with the corresponding single mutants and that the triple mutant was unable to interact even with the high avidity binder Rabaptin-5-(551–661) under more stringent conditions (Fig. 5A). Finally, the importance of these residues in the failure of GGA3-GAT to bind Rabaptin-5 was investigated by attempting the corresponding gain-of-function substitutions (Fig. 5B). The results showed that the S283N substitution in GGA3-GAT conferred Rabaptin-5 binding, whereas the other individual substitutions (K258M and R260P) were ineffective (Lys258, Arg260, and Ser283 in GGA3 correspond to Met259, Pro261, and Asn284 in GGA1; see sequence alignment in Fig. 2). However, the GGA3-GAT K258M/S283N double substitution resulted in a further increase in Rabaptin-5 binding (Fig. 5B), which was indistinguishable from that of the native GGA1-GAT (Fig. 5A). These experiments highlight the importance of Met259 and Asn284 in the binding of Rabaptin-5 to GGA1-GAT and demonstrate that their substitution in GGA3-GAT (Asn284 more critically so than Met259) accounts for the lack of recognition of GGA3-GAT by this partner. These results also show an indirect role of Pro261 in the binding of Rabaptin-5 to GGA1-GAT, which was only evident upon assaying double or triple mutants of this domain.

Overlapping but Distinct Binding Sites for Rabaptin-5, Ubiquitin, and TSG101 in GGA1-GAT—It has been recently reported that ubiquitin and TSG101 bind to GGA3-GAT and, more weakly, to GGA1-GAT domains and that residues in the trihelical bundle of GGA3-GAT are important for these interactions (24–26). Given that GGA3-GAT does not bind Rabaptin-5 (Fig. 5B), we studied the possible relationship of the binding sites for Rabaptin-5, ubiquitin, and TSG101 using mutants of GGA1-GAT, a domain that recognizes all three partners. We analyzed various individual substitutions that are important for the interaction of Rabaptin-5 with GGA1-GAT, along with other mutations that affect the binding of ARF1 to the N-terminal hook. The substitutions were introduced in GGA1-VHS+GAT, as opposed to GGA1-GAT, because although GGA1-VHS+GAT binds ubiquitin very weakly, it does so more efficiently than the shorter GAT construct, possibly due to improved presentation of binding sites on the GAT domain (24). The results in Fig. 6A show that some, but not all, substitutions that affected the binding of Rabaptin-5 to GGA1-GAT (namely R260E, A267D, and L277A) also resulted in decreased binding to ubiquitin. Among these three mutants, only L277A affected the binding to TSG101 as well. Importantly, the two other substitutions that also diminished binding to Rabaptin-5 (F264A and N284A) (Fig. 3B) were inconsequential for recognition of either ubiquitin or TSG101 (Fig. 6A). The differential recognition of the F264A and N284A mutants by the three partners did not originate in the use of different constructs (GGA1-GAT versus GGA1-VHS+GAT), since the interactions of the Rabaptin-5 constructs with the GGA1-VHS+GAT mutants were similar to those previously observed with the GGA1-GAT substitutions, regardless of the subcloning of Rabaptin-5 constructs or GGA1-VHS+GAT in AD or BD vectors (Fig. 6, B and C). The analysis in Fig. 6A also shows that a mutation in the ARF1 binding site of GGA1-GAT (I197A but not N194A) affected the recognition of ubiquitin and TSG101 (the N182A substitution bound weakly to ubiquitin and did not bind TSG101; results not shown).



View larger version (43K):
[in this window]
[in a new window]
  		FIG. 6.
Interaction of GGA1-VHS+GAT mutants with ubiquitin and TSG101. A, experiments were performed using AD-GGA1-VHS+GAT and BD-ubiquitin or BD-TSG101 constructs. B and C, the interaction of Rabaptin-5 with selected substitutions introduced in the GGA1-VHS+GAT construct is similar to that observed when the same substitutions were introduced in GGA1-GAT (Fig. 3B, right panel). The interaction of Rabaptin-5 constructs with the GGA1-VHS+GAT L182A mutant (not shown) is also similar to that of the native domain, as observed with the corresponding GGA1-GAT constructs (Fig. 3B, right panel). Similar results were observed when the Rabaptin-5 and GGA1-VHS+GAT constructs were subcloned in the BD and AD vectors, respectively (B), or vice versa (C). Whereas BD-Rabaptin-5 FL displayed non-specific interactions in low stringency plates (–His and –His plus 1 mM 3-AT, not shown in B), an interaction of this construct with AD-GGA1-VHS+GAT N284A was nonetheless noticeable, although more weakly than those with the wt or the I197A constructs.

 
The information obtained with the GGA1-GAT mutants tested against all the GAT domain partners (ARF1, Rabaptin-5, ubiquitin, and TSG101) is summarized in Table I. This analysis demonstrates the existence of overlapping but distinct binding sites for Rabaptin-5, ubiquitin, and TSG101 on the trihelical bundle of GGA1-GAT. The results obtained with the F264A and the N284A substitutions, which bind ubiquitin but not Rabaptin-5, indicate that the interaction of Rabaptin-5 with the trihelical bundle of GGA1-GAT is not a prerequisite for the binding of ubiquitin to this subdomain. The reciprocal possibility, that ubiquitin binding may be necessary for the interaction of Rabaptin-5 to the trihelical bundle, could not be ruled out, since the panel of mutants that we tested in Fig. 6 included substitutions in the trihelical bundle affecting Rabaptin-5 binding. However, the results summarized in Table I indicate that, in addition to residues required for both Rabaptin-5 and ubiquitin binding (namely Arg260 and Leu277), the interaction with Rabaptin-5 requires other residues on the GGA1-GAT trihelical bundle (e.g. Phe264 and Asn284).


View this table:
[in this window]
[in a new window]
  		TABLE I
Binding of partners to GGA1-GAT mutants

Information presented in this table was derived from the experiments shown in Figs. 3,4,5,6 as well as results not shown (interaction of L182A with ubiquitin and TSG101). Only the substitutions affecting Rabaptin-5 binding that were also assayed for interactions with all other partners (Figs. 3 and 6A) are shown.

 
Another conclusion derived from this analysis is that the binding of ubiquitin to the trihelical bundle subdomain also involves residues independent from those required by TSG101; this is apparent from the fact the R260E and A267D substitutions in the GGA1-GAT bundle still allow binding of TSG101 but preclude binding of ubiquitin (Fig. 6 and Table I). Finally, and as discussed above for the effect of GGA1-GAT N194A on Rabaptin-5 binding (Fig. 3B), the lack of binding of TSG101 to the L182A and I197A mutants (Fig. 6A and Table I) also suggests the possibility of interactions between the N-terminal hook and trihelical bundle subdomains of GGA1-GAT. The relative importance of residues in the binding of Rabaptin-5, ubiquitin, and TSG101 to the GGA1-GAT trihelical bundle is illustrated in Fig. 1C (right image).

Competitive and Cooperative Interactions in the Binding of Rabaptin-5 and Ubiquitin to the GGA1-GAT Domain—The overlapping but distinct binding sites for Rabaptin-5 and ubiquitin on the GGA1-GAT trihelical bundle suggest the possibility of competitive effects by these ligands. We first analyzed this issue by studying the effect of preincubating His6-Rabaptin-5-(551–862) with GST-GGA1-VHS+GAT on the subsequent recognition of the GST fusion protein by ubiquitin-agarose. The results in Fig. 7A show that preincubation with increasing concentrations of this Rabaptin-5 fragment, containing the GAT domain interaction region, results in a proportional decrease in the recognition of GST-GGA1-VHS+GAT by ubiquitin-agarose. We did not observe any interactions between ubiquitin-agarose and the GST-{beta}3 ear control, regardless of the presence or absence of His6-Rabaptin-5-(551–862) during the preincubation, and we did not observe binding of GST-GGA1-VHS+GAT to the protein A-agarose used as control with or without the recombinant Rabaptin-5 fragment (Fig. 7A). In addition, we could not demonstrate specific binding of the recombinant Rabaptin-5 fragment to ubiquitin-agarose that could account for the inhibition of binding of GST-GGA1-VHS+GAT to ubiquitin (low binding to both ubiquitin-agarose and protein A-agarose was observed; results not shown). The lack of interaction of Rabaptin-5 with ubiquitin was also confirmed in yeast two-hybrid experiments (results not shown). The results in Fig. 7A indicate that preformation of a relatively stable complex between Rabaptin-5-(551–862) and GGA1-VHS+GAT occludes the sites on the GAT domain necessary for interaction with ubiquitin during the subsequent incubation. We also performed a reciprocal analysis and studied the effect of preincubating GST-GGA1-VHS+GAT with increasing concentrations of His6-ubiquitin on the subsequent pull-down of Rabaptin-5 from brain cytosol. Surprisingly, the results showed that the concentration-dependent binding of His6-ubiquitin to GGA1-VHS-GAT (Fig. 7B, lower blot) resulted in a proportional increase in pull-down of endogenous Rabaptin-5 (Fig. 7B, upper blot). This effect was not mimicked by preincubation of GST-GGA1-VHS+GAT with a high concentration of bovine serum albumin or other irrelevant proteins (Fig. 7B) (results not shown). The specificity of the effect of His6-ubiquitin on the binding of Rabaptin-5 to GST-GGA1-VHS+GAT was also confirmed by the observation that preincubation with His6-ubiquitin did not result in an artificial pull-down of Rabaptin-5 by GST-{beta}3 ear, a construct that does not recognize this protein (Fig. 7B) (results not shown). Importantly, preincubation with ubiquitin did not increase the binding of GST-GGA1-VHS+GAT to recombinant His6-Rabaptin-5-(551–862) (Fig. 7C). This observation suggests that the ubiquitin-induced increase in the binding of endogenous Rabaptin-5 to GST-GGA1-VHS+GAT depends on the presence of other cytosolic proteins associated with Rabaptin-5 and/or GGA1-VHS+GAT. This possibility was supported by the demonstration that ubiquitin interacts with an endogenous complex containing GGA1, Rabaptin-5, and Rabex-5 (Fig. 7D). The presence of Rabex-5 is consistent with the association of this Rab5-specific guanine nucleotide exchanger with Rabaptin-5 (22) and indicates that the physiologically relevant Rabaptin-5-Rabex-5 complex associates with GGA1 and ubiquitin.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Recent findings have shown that the GGA1-GAT domain is a multifunctional module capable of interacting with various partners including ARF (7, 8), Rabaptin-5 (19, 30), ubiquitin (24–26), and TSG101 (24). The crystal structure of both the ARF-bound (29) and unliganded GGA1-GAT domains (27, 28, 30), together with mutational analyses, have defined in detail the residues in the N-terminal hook that bind to ARF. Less information is available, however, on the interaction of the trihelical bundle with Rabaptin-5, ubiquitin, and TSG101 and on the possible relationship among the individual binding sites on this subdomain. The present study initiates the analysis of the molecular determinants responsible for the multifunctional nature of the GAT domain trihelical bundle.

We have defined residues that are critical for the binding of Rabaptin-5 by introducing substitutions in the proposed protein-binding site located on the surface of the GGA1-GAT trihelical bundle (27, 28). This analysis demonstrated the importance of Arg260, Phe264, Ala267, Leu277, and Asn284 for the binding of Rabaptin-5 to GGA1-GAT (Fig. 3B). The experiments carried out with other GAT domains confirmed and extended the conclusions of this mutational analysis. The three members of a subfamily of VHS domain-containing proteins, namely TOM1, TOM1-L1, and TOM1-L2, also contain GAT domains that display similarity to the C-terminal half of {alpha}2 and the {alpha}3 and {alpha}4 helices constituting the trihelical bundle of the GGA-GAT domains where Rabaptin-5 binds. TOM1 undergoes monoubiquitination, binds polyubiquitinated proteins, and interacts with clathrin and the ubiquitin-binding protein tollip (35). The GAT domain is critical for the interaction of TOM1 with tollip and other polyubiquitinated proteins (35). TOM1, but not TOM1-L1 or TOM1-L2, also interacts with the endosomal protein endofin (36). The endofin-dependent recruitment of TOM1 to endosomes (36), along with the interaction of ubiquitin with TOM1-GAT and the conservation of the sequence corresponding to the GGA1-GAT trihelical bundle in TOM1-GAT, suggests that this domain could also interact with the endosomal fusion regulator Rabaptin-5. However, our experiments showed that the GAT domains in TOM1 and TOM1-L1 do not bind Rabaptin-5 (results not shown). This observation confirmed the mutational analysis in Fig. 3 demonstrating that the native GAT domains in TOM1 and TOM1-L1, which have substitutions in residues critical for the binding of GGA1-GAT to Rabaptin-5, are unable to interact with this partner. The lack of binding of Rabaptin-5 to TOM1/TOM1-L1 GAT domains can be explained by the substitution of GGA1-GAT residues Arg260, Phe264, or Leu277 (replaced by Gln, Leu/Met, and Thr, respectively, in TOM1/TOM1-L1) (Fig. 2). Although not included in our experimental analysis, the strong identity of TOM1-L2-GAT with the corresponding domains in TOM1 and TOM1-L1 (including the above mentioned nonconserved substitutions when compared with GGA1-GAT) predicts that the former does not bind Rabaptin-5 either.

The residues required for Rabaptin-5 binding were also defined through the analysis of the structural differences responsible for the lack of interaction of GGA3-GAT with Rabaptin-5. The loss- and gain-of-function mutations in GGA1-GAT and GGA3-GAT, respectively (Figs. 4 and 5), confirmed the importance of Asn284 and revealed a role for Met259 in the binding of Rabaptin-5 to GGA1-GAT that was not anticipated in the initial mutational analysis (Fig. 3) based on inferences from the crystal structure of this domain (28). The lack of interaction of GGA1-GAT N284S is in agreement with the recent report by Zhai et al. (37), who observed a diminished pull-down of recombinant Rabaptin-5 by this mutant (similar to that resulting from their F264R and L277R substitutions). These authors found a partial recovery of Rabaptin-5 pull-down by a GGA3-GAT reversal mutant and suggested that additional structural differences may be responsible for the lack of binding of Rabaptin-5 to GGA3-GAT (37). Our results demonstrating the additional role of M259K and the complete recovery of Rabaptin-5 binding by the GGA3-GAT K258M/S283N when compared with the GGA1-GAT (Fig. 5) extend the definition of the molecular determinants responsible for the lack of binding of Rabaptin-5 to GGA3-GAT. Finally, the lack of interaction of syntaxin 6 with Rabaptin-5 (results not shown) demonstrates that the selectivity of Rabaptin-5 in the recognition of specific GAT domains (GGA1 and GGA2-GAT, but not GGA3-, TOM1-, or TOM1-L1-GAT) extends also to other structurally similar trihelical bundles.

We have not only defined the residues in the GGA-GAT trihelical bundle that are important for the interaction with Rabaptin-5 but also the relevance of these residues to the binding of ubiquitin and TSG101 to this subdomain. In this regard, we have found that some but not all residues in the GGA1-GAT trihelical bundle that are critical for recognition of Rabaptin-5 are also important for binding of ubiquitin and TSG101. The information summarized in Table I shows that whereas Arg260, Ala267, and Leu277 in the GGA1-GAT are important for binding of both Rabaptin-5 and ubiquitin, the residues Phe264 and Asn284 are only important for binding of Rabaptin-5. This indicates that the binding sites for these two partners overlap but are not identical. Similarly, a partial overlap of binding sites is also apparent when the interaction with TSG101 is considered. Of the three residues that we found important for both Rabaptin-5 and ubiquitin binding (Arg260, Ala267, and Leu277), only Leu277 is also critical for recognition of TSG101 (Table I and residues in red, blue, and green in Fig. 1C). The importance of the GGA1-GAT Leu277 in ubiquitin binding is consistent with the role of the equivalent GGA3-GAT Leu276 in the recognition of this partner (24, 25). The fact that this residue is not conserved in TOM1 (Fig. 2) seems inconsistent with recent reports showing that TOM1-VHS+GAT binds ubiquitin (25, 35). Whereas the binding of ubiquitin to GGA1 and TOM1 were not compared in these studies, it is possible that the interaction with the latter may reflect a lower affinity recognition or, alternatively, a different mode of recognition, as also suggested by the additional role of the C-terminal region of TOM1-VHS in the binding of ubiquitin (35). The existence of residues critical for the binding of ubiquitin to GGA1-GAT in addition to those also required for Rabaptin-5 binding (Arg260, Ala267, and Leu277; see Table I) is also highlighted by the conservation of these residues in GGA2-GAT, a domain reported not to bind ubiquitin efficiently (25).

A question arising from these studies is whether the binding of one partner to the GGA1-GAT trihelical bundle is dependent on the previous binding of another. Although we studied the binding of ubiquitin and TSG101 in GGA1-GAT mutants that did not interact with Rabaptin-5, we did not screen for independent mutations that may affect only the interaction with either TSG101 or ubiquitin, without affecting the recognition of Rabaptin-5 by this domain. However, our analysis shows that ubiquitin does not require the binding of Rabaptin-5 to the GGA1-GAT in order to interact with this subdomain (see mutations F264A and N284A in Table I). Similarly, TSG101 does not require the binding of ubiquitin to GGA1-GAT to interact with this subdomain (see R260E and A267D in Table I). Although we have not identified residues that interact with ubiquitin but do not interact with TSG101 (except for Leu182 and Ile197, which bind ubiquitin weakly), the existence of TSG101-independent contacts for ubiquitin on GGA1-GAT (such as Arg260 and Ala267) is supported by the finding that small interfering RNA-mediated inhibition of expression of TSG101 does not affect the binding of ubiquitin to GGA3 (24).

Another possibility arising from the observation that multiple partners interact with the GGA1-GAT trihelical bundle through overlapping but distinct sites is the existence of competitive effects between them. We analyzed this possibility by examining the interplay between the binding of Rabaptin-5 and ubiquitin to the GGA1-GAT trihelical bundle. The experiments uncovered a complex relationship in the binding of these partners. Preincubation with His6-Rabaptin-5-(551–862) (a fragment that contains the coiled-coil regions recognizing the GAT domain) decreased the interaction of the GST-GGA1-VHS+GAT with ubiquitin-agarose in a concentration-dependent manner (Fig. 7A). This result is consistent with the overlap between the binding sites and suggests that the interaction between recombinant Rabaptin-5 fragment and GGA1-VHS+GAT competes effectively with the low affinity interaction between ubiquitin and the GAT domain-containing construct during the subsequent incubation (complexes between ubiquitin and human and yeast GGAs display affinities in the 100–400 µM range (26), similar to the interaction between ubiquitin and the ubiquitin-interacting motifs (38)). However, a different effect was observed when the GST-GGA1-VHS+GAT construct was preincubated with ubiquitin before the incubation with brain cytosol. In this case, the preincubation with ubiquitin resulted in a concentration-dependent increase in the pull-down of endogenous Rabaptin-5 from the brain cytosol (Fig. 7B). Whereas this increase may seem paradoxical given the overlap of some of the residues critical for binding of Rabaptin-5 and ubiquitin (Table I), it is likely to derive from the presence in the cytosol of Rabaptin-5-associated proteins that also bind ubiquitin or undergo ubiquitin-dependent modification (Rabaptin-5 does not bind ubiquitin directly; results not shown). The formation of a complex between Rabaptin-5 and a ubiquitin-binding partner may allosterically increase the affinity of Rabaptin-5 for its site on the GAT domain and override the low affinity interaction between this domain and ubiquitin. This effect may be also magnified by the fact that GGA1-GAT not only binds ubiquitin and ubiquitinated proteins but also undergoes monoubiquitination (25). The monoubiquitination of GGA1-VHS-GAT in the presence of the cytosolic machinery may provide an independent attachment for Rabaptin-5-associated proteins that would result in the increased pull-down of Rabaptin-5 shown in Fig. 7B. The formation of networks based on the binding of ubiquitin, subsequent monoubiquitination, and interaction with other ubiquitin-binding proteins is, in fact, a recurrent theme in the regulation of the endocytic pathway (39). Two lines of evidence support the above hypothesis. First, the effect brought about by preincubation of GGA1-VHS-GAT with ubiquitin was seen only with endogenous Rabaptin-5 and could not be recapitulated by incubation with His6-Rabaptin-5-(551–862) instead of brain cytosol (Fig. 7C). Second, we observed the interaction of ubiquitin with an endogenous complex including GGA1, Rabaptin-5, and its partner the Rab5-specific guanine nucleotide exchanger Rabex-5 (Fig. 7D). Interestingly, Vps9, the yeast orthologue of Rabex-5, interacts with ubiquitin and undergoes ubiquitination through its Cue1-homologous domain (40, 41), which appears to be conserved in Rabex-5 (40). We are currently addressing the possible interaction of Rabex-5 with ubiquitin and its effects on the Rabaptin-5-GGA1-GAT interaction; however, we cannot exclude at this time the possibility of additional ubiquitin-binding proteins present in endogenous complexes containing Rabaptin-5 that may promote the interaction of this protein with GGA1-GAT.

A final issue that deserves consideration is the possible interaction of the two subdomains in GGA1-GAT, namely the N-terminal hook and the trihelical bundle. These subdomains are separated by >35 Å and have protein-binding sites that face opposite directions (37) (Fig. 1C). Several findings suggest a possible interaction between these domains, including the effect of the GGA1-GAT N194A on binding to full-length Rabaptin-5 (Fig. 3B) and the lack of interaction of TSG101 with the GGA1-VHS-GAT L182A and I197A mutants (Table I). These findings suggest that alterations in the N-terminal hook may allosterically affect the binding of partners to critical residues in the trihelical bundle. The possibility of interactions between the two subdomains of GGA1-GAT is also supported by the observation that deletion of the N-terminal hook region enhances the efficiency of binding of ubiquitinated proteins to the trihelical bundle and by the increased binding of multiubiquitin chains to GGA1-GAT following interaction of GTP{gamma}S-bound ARF with this domain in vitro (25).

The competitive and cooperative interactions observed following binding of Rabaptin-5 and ubiquitin to GGA1-GAT are likely to influence the proposed role of the GGAs in the sorting of ubiquitinated cargo at the endosomes and trans-Golgi network (24, 26). Interestingly, small interfering RNA-mediated decrease in the expression of GGA3, but not GGA1 or GGA2, induces defects in endosomal trafficking in HeLa cells (24). The various GGA-GAT domains show different properties in terms of binding to Rabaptin-5 and ubiquitin; GGA1-GAT binds both partners, GGA2-GAT binds Rabaptin-5 (minimal binding to ubiquitin was observed), and GGA3-GAT binds only ubiquitin (Fig. 5B) (24, 25). It is then possible that the exclusion of Rabaptin-5 from GGA3-GAT, despite the recognition of Rabaptin-5 by the separate GGA3-GAE domain (19), and the minimal binding of ubiquitin to GGA2-GAT are responsible for the unique role of GGA3 in endosomal trafficking. Future studies will be required to understand the functional interplay of ubiquitin, Rabaptin-5, and associated proteins in the regulation of endosome fusion by GGA1-GAT.


   FOOTNOTES
 
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} To whom correspondence should be addressed: Cell Biology and Metabolism Branch, NICHD, Bldg. 18T/Rm. 101, National Institutes of Health, Bethesda, MD 20892. Tel.: 301-496-6368; Fax: 301-402-0078; E-mail: juan{at}helix.nih.gov.

1 The abbreviations used are: GGA, Golgi-localized, {gamma}-ear-containing, ARF-binding protein; 3-AT, 3-amino-1,2,4-triazole; AD, Gal4 transcription activation domain; ARF, ADP-ribosylation factor; BD, Gal4 DNA-binding domain; GAE, {gamma}-adaptin ear; GAT, domain found in GGAs and TOM1; GST; glutathione S-transferase; SNARE, N-ethylmaleimide-sensitive factor attachment protein receptor; TOM1, target of myb1; TOM1-L1, TOM1-like 1; TOM1-L2, TOM1-like 2; VHS, Vps27, Hrs, Stam; TSG101, tumor susceptibility gene 101 product; GTP{gamma}S, guanosine 5'-3-O-(thio)triphosphate. Back


   ACKNOWLEDGMENTS
 
We are grateful to J. Hurley for critical reading of the manuscript and to A. San Miguel and X. Zhu for expert technical assistance. We thank Cecilia N. Arighi for the preparation of His6-Rabaptin-5-(551–862) and José Martina for helpful discussions. We are also indebted to Margaret S. Robinson and Marino Zerial for the generous gift of antisera.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Bonifacino, J. S. (2004) Nat. Rev. Mol. Cell. Biol. 5, 23–32[CrossRef][Medline] [Order article via Infotrieve]
  2. Puertollano, R., Aguilar, R. C., Gorshkova, I., Crouch, R. J., and Bonifacino, J. S. (2001) Science 292, 1712–1716[Abstract/Free Full Text]
  3. Zhu, Y., Doray, B., Poussu, A., Lehto, V. P., and Kornfeld, S. (2001) Science 292, 1716–1718[Abstract/Free Full Text]
  4. Takatsu, H., Katoh, Y., Shiba, Y., and Nakayama, K. (2001) J. Biol. Chem. 276, 28541–28545[Abstract/Free Full Text]
  5. Nielsen, M. S., Madsen, P., Christensen, E. I., Nykjaer, A., Gliemann, J., Kasper, D., Pohlmann, R., and Petersen, C. M. (2001) EMBO J. 20, 2180–2190[CrossRef][Medline] [Order article via Infotrieve]
  6. He, X., Chang, W. P., Koelsch, G., and Tang, J. (2002) FEBS Lett. 524, 183–187[Medline] [Order article via Infotrieve]
  7. Boman, A. L., Zhang, C. J., Zhu, X., and Kahn, R. A. (2000) Mol. Biol. Cell 11, 1241–1255[Abstract/Free Full Text]
  8. Dell'Angelica, E. C., Puertollano, R., Mullins, C., Aguilar, R. C., Vargas, J. D., Hartnell, L. M., and Bonifacino, J. S. (2000) J. Cell Biol. 149, 81–94[Abstract/Free Full Text]
  9. Puertollano, R., Randazzo, P., Hartnell, L. M., Presley, J., and Bonifacino, J. S. (2001) Cell 105, 93–102[CrossRef][Medline] [Order article via Infotrieve]
  10. Hirst, J., Lui, W. W., Bright, N. A., Totty, N., Seaman, M. N., and Robinson, M. S. (2000) J. Cell Biol. 149, 67–80[Abstract/Free Full Text]
  11. Wasiak, S., Legendre-Guillemin, V., Puertollano, R., Blondeau, F., Girard, M., de Heuvel, E., Boismenu, D., Bell, A. W., Bonifacino, J. S., and McPherson, P. S. (2002) J. Cell Biol. 158, 855–862[Abstract/Free Full Text]
  12. Kalthoff, C., Groos, S., Kohl, R., Mahrhold, S., and Ungewickell, E. J. (2002) Mol. Biol. Cell 13, 4060–4073[Abstract/Free Full Text]
  13. Hirst, J., Motley, A., Harasaki, K., Peak Chew, S. Y., and Robinson, M. S. (2003) Mol. Biol. Cell 14, 625–641[Abstract/Free Full Text]
  14. Mills, I. G., Praefcke, G. J., Vallis, Y., Peter, B. J., Olesen, L. E., Gallop, J. L., Butler, P. J., Evans, P. R., and McMahon, H. T. (2003) J. Cell Biol. 160, 213–222[Abstract/Free Full Text]
  15. Page, L. J., Sowerby, P. J., Lui, W. W., and Robinson, M. S. (1999) J. Cell Biol. 146, 993–1004[Abstract/Free Full Text]
  16. Takatsu, H., Yoshino, K., and Nakayama, K. (2000) Biochem. Biophys. Res. Commun. 271, 719–725[CrossRef][Medline] [Order article via Infotrieve]
  17. Lui, W. W., Collins, B. M., Hirst, J., Motley, A., Millar, C., Schu, P., Owen, D. J., and Robinson, M. S. (2003) Mol. Biol. Cell 14, 2385–2398[Abstract/Free Full Text]
  18. Mattera, R., Ritter, B., Sidhu, S. S., McPherson, P. S., and Bonifacino, J. S. (2004) J. Biol. Chem. 279, 8018–8028[Abstract/Free Full Text]
  19. Mattera, R., Arighi, C. N., Lodge, R., Zerial, M., and Bonifacino, J. S. (2003) EMBO J. 22, 78–88[CrossRef][Medline] [Order article via Infotrieve]
  20. Stenmark, H., Vitale, G., Ullrich, O., and Zerial, M. (1995) Cell 83, 423–432[CrossRef][Medline] [Order article via Infotrieve]
  21. Vitale, G., Rybin, V., Christoforidis, S., Thornqvist, P., McCaffrey, M., Stenmark, H., and Zerial, M. (1998) EMBO J. 17, 1941–1951[CrossRef][Medline] [Order article via Infotrieve]
  22. Horiuchi, H., Lippe, R., McBride, H. M., Rubino, M., Woodman, P., Stenmark, H., Rybin, V., Wilm, M., Ashman, K., Mann, M., and Zerial, M. (1997) Cell 90, 1149–1159[CrossRef][Medline] [Order article via Infotrieve]
  23. Lippe, R., Miaczynska, M., Rybin, V., Runge, A., and Zerial, M. (2001) Mol. Biol. Cell 12, 2219–2228[Abstract/Free Full Text]
  24. Puertollano, R., and Bonifacino, J. S. (2004) Nat. Cell Biol. 6, 244–251[Medline] [Order article via Infotrieve]
  25. Shiba, Y., Katoh, Y., Shiba, T., Yoshino, K., Takatsu, H., Kobayashi, H., Shin, H. W., Wakatsuki, S., and Nakayama, K. (2004) J. Biol. Chem. 279, 7105–7111[Abstract/Free Full Text]
  26. Scott, P. M., Bilodeau, P. S., Zhdankina, O., Winistorfer, S. C., Hauglund, M. J., Allaman, M. A., Kearney, W. R., Robertson, A. D., Boman, A. L., and Piper, R. C. (2004) Nat. Cell Biol. 6, 252–259[Medline] [Order article via Infotrieve]
  27. Collins, B. M., Watson, P. J., and Owen, D. J. (2003) Dev. Cell 4, 321–332[CrossRef][Medline] [Order article via Infotrieve]
  28. Suer, S., Misra, S., Saidi, L. F., and Hurley, J. H. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 4451–4456[Abstract/Free Full Text]
  29. Shiba, T., Kawasaki, M., Takatsu, H., Nogi, T., Matsugaki, N., Igarashi, N., Suzuki, M., Kato, R., Nakayama, K., and Wakatsuki, S. (2003) Nat. Struct. Biol. 10, 386–393[CrossRef][Medline] [Order article via Infotrieve]
  30. Zhu, G., Zhai, P., He, X., Terzyan, S., Zhang, R., Joachimiak, A., Tang, J., and Zhang, X. C. (2003) Biochemistry 42, 6392–6399[CrossRef][Medline] [Order article via Infotrieve]
  31. Misura, K. M., Bock, J. B., Gonzalez, L. C., Jr., Scheller, R. H., and Weis, W. I. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 9184–9189[Abstract/Free Full Text]
  32. Boman, A. L., Salo, P. D., Hauglund, M. J., Strand, N. L., Rensink, S. J., and Zhdankina, O. (2002) Mol. Biol. Cell 13, 3078–3095[Abstract/Free Full Text]
  33. Takatsu, H., Yoshino, K., Toda, K., and Nakayama, K. (2002) Biochem. J. 365, 369–378[CrossRef][Medline] [Order article via Infotrieve]
  34. Dell'Angelica, E. C., Klumperman, J., Stoorvogel, W., and Bonifacino, J. S. (1998) Science 280, 431–434[Abstract/Free Full Text]
  35. Yamakami, M., Yoshimori, T., and Yokosawa, H. (2003) J. Biol. Chem. 278, 52865–52872[Abstract/Free Full Text]
  36. Seet, L. F., Liu, N., Hanson, B. J., and Hong, W. (2004) J. Biol. Chem. 279, 4670–4679[Abstract/Free Full Text]
  37. Zhai, P., He, X., Liu, J., Wakeham, N., Zhu, G., Li, G., Tang, J., and Zhang, X. C. (2003) Biochemistry 42, 13901–13908[CrossRef][Medline] [Order article via Infotrieve]
  38. Raiborg, C., Bache, K. G., Gillooly, D. J., Madshus, I. H., Stang, E., and Stenmark, H. (2002) Nat. Cell Biol. 4, 394–398[CrossRef][Medline] [Order article via Infotrieve]
  39. Di Fiore, P. P., Polo, S., and Hofmann, K. (2003) Nat. Rev. Mol. Cell. Biol. 4, 491–497[CrossRef][Medline] [Order article via Infotrieve]
  40. Donaldson, K. M., Yin, H., Gekakis, N., Supek, F., and Joazeiro, C. A. (2003) Curr. Biol. 13, 258–262[CrossRef][Medline] [Order article via Infotrieve]
  41. Shih, S. C., Prag, G., Francis, S. A., Sutanto, M. A., Hurley, J. H., and Hicke, L. (2003) EMBO J. 22, 1273–1281[CrossRef][Medline] [Order article via Infotrieve]
  42. Nicholls, A., Sharp, K. A., and Honig, B. (1991) Proteins 11, 281–296[CrossRef][Medline] [Order article via Infotrieve]
  43. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946–950[CrossRef]
  44. Esnouf, R. M. (1997) J. Mol. Graph. Model 15, 132–134[CrossRef][Medline] [Order article via Infotrieve]
  45. Merritt, E. A., and Bacon, D. J. (1997) Methods Enzymol. 277, 505–524[Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
C. S. Chen, C. M. Nelson, D. Khauv, S. Bennett, E. S. Radisky, Y. Hirai, M. J. Bissell, and D. C. Radisky
Homology with Vesicle Fusion Mediator Syntaxin-1a Predicts Determinants of Epimorphin/Syntaxin-2 Function in Mammary Epithelial Morphogenesis
J. Biol. Chem., March 13, 2009; 284(11): 6877 - 6884.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Biol. CellHome page
J. Wang, H.-Q. Sun, E. Macia, T. Kirchhausen, H. Watson, J. S. Bonifacino, and H. L. Yin
PI4P Promotes the Recruitment of the GGA Adaptor Proteins to the Trans-Golgi Network and Regulates Their Recognition of the Ubiquitin Sorting Signal
Mol. Biol. Cell, July 1, 2007; 18(7): 2646 - 2655.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
R. Mattera, Y. C. Tsai, A. M. Weissman, and J. S. Bonifacino
The Rab5 Guanine Nucleotide Exchange Factor Rabex-5 Binds Ubiquitin (Ub) and Functions as a Ub Ligase through an Atypical Ub-interacting Motif and a Zinc Finger Domain
J. Biol. Chem., March 10, 2006; 281(10): 6874 - 6883.
[Abstract] [Full Text] [PDF]

Home page
 GENES CELLSHome page
M. Kawasaki, T. Shiba, Y. Shiba, Y. Yamaguchi, N. Matsugaki, N. Igarashi, M. Suzuki, R. Kato, K. Kato, K. Nakayama, et al.
Molecular mechanism of ubiquitin recognition by GGA3 GAT domain
Genes Cells, July 1, 2005; 10(7): 639 - 654.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
S. K. Mishra, P. A. Keyel, M. A. Edeling, A. L. Dupin, D. J. Owen, and L. M. Traub
Functional Dissection of an AP-2 {beta}2 Appendage-binding Sequence within the Autosomal Recessive Hypercholesterolemia Protein
J. Biol. Chem., May 13, 2005; 280(19): 19270 - 19280.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
R. Puertollano
Interactions of TOM1L1 with the Multivesicular Body Sorting Machinery
J. Biol. Chem., March 11, 2005; 280(10): 9258 - 9264.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
G. Prag, S. Lee, R. Mattera, C. N. Arighi, B. M. Beach, J. S. Bonifacino, and J. H. Hurley
Structural mechanism for ubiquitinated-cargo recognition by the Golgi-localized, {gamma}-ear-containing, ADP-ribosylation-factor-binding proteins
PNAS, February 15, 2005; 102(7): 2334 - 2339.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
P. S. Bilodeau, S. C. Winistorfer, M. M. Allaman, K. Surendhran, W. R. Kearney, A. D. Robertson, and R. C. Piper
The GAT Domains of Clathrin-associated GGA Proteins Have Two Ubiquitin Binding Motifs
J. Biol. Chem., December 24, 2004; 279(52): 54808 - 54816.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/30/31409    most recent
M402183200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mattera, R.
Right arrow Articles by Bonifacino, J. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mattera, R.
Right arrow Articles by Bonifacino, J. S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2004 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement