Traffic 2007; 8: 1385–1403 Blackwell Munksgaard # 2007 The Authors Journal compilation # 2007 Blackwell Publishing Ltd doi: 10.1111/j.1600-0854.2007.00612.x Rab6-interacting Protein 1 Links Rab6 and Rab11 Function Stéphanie Miserey-Lenkei1,2,†, Francxois Waharte1,3,†, Annick Boulet1,2, Marie-Hélène Cuif4, Danielle Tenza1,5, Amed El Marjou1,6, Gracxa Raposo1,5, Jean Salamero1,2,3, Laurent Héliot7, Bruno Goud1,2,* and Solange Monier1,2,8,* 1 Institut Curie, Centre de Recherche, Paris, 75248 France Molecular Mechanisms of Intracellular Transport, UMR CNRS 144, Institut Curie, 26 rue d‘Ulm, 75248 Paris Cedex 05, France 3 Cell and Tissue Imaging Facility, UMR CNRS 144, Institut Curie, 26 rue d’Ulm, 75248 Paris Cedex 05, France 4 UMR CNRS 8621, University of Paris 11, Orsay, France 5 Structure and Membrane Compartments, UMR CNRS 144, Institut Curie, 26 rue d’Ulm, 75248 Paris Cedex 05, France 6 Recombinant Protein Facilities, UMR CNRS 144, Institut Curie, 26 rue d’Ulm, 75248 Paris Cedex 05, France 7 Biophotonic Group, IRI, Lille, France 8 Present address: UMR CNRS 6061, Université de Rennes, 35043 Rennes, France *Corresponding author: Bruno Goud, bruno.goud@curie.fr or Solange Monier, solange.monier@curie.fr †These authors contributed equally to this work. 2 Rab11 and Rab6 guanosine triphosphatases are associated with membranes of the recycling endosomes (REs) and Golgi complex, respectively. Evidence indicates that they sequentially regulate a retrograde transport pathway between these two compartments, suggesting the existence of proteins that must co-ordinate their functions. Here, we report the characterization of two isoforms of a protein, Rab6-interacting protein 1 (R6IP1), originally identified as a Rab6-binding protein. R6IP1 also binds to Rab11A in its GTP-bound conformation. In interphase cells, R6IP1 is targeted to the Golgi in a Rab6-dependent manner but can associate with Rab11-positive compartments when the level of Rab11A is increased within the cells. Fluorescence resonance energy transfer analysis using fluorescence lifetime imaging shows that the overexpression of R6IP1 promotes an interaction between Rab11A and Rab6 in living cells. Accordingly, the REs marked by Rab11 and transferrin receptor are depleted from the cell periphery and accumulate in the pericentriolar area. However, endosomal and Golgi membranes do not appear to fuse with each other. We also show that R6IP1 function is required during metaphase and cytokinesis, two mitotic steps in which a role of Rab6 and Rab11 has been previously documented. We propose that R6IP1 may couple Rab6 and Rab11 function throughout the cell cycle. Key words: cytokinesis, FRET/FLIM, Golgi complex, intracellular transport, mitosis, Rab GTPases, recycling endosomes Received 16 February 2007, revised and accepted for publication 11 June 2007, uncorrected manuscript published online 15 June 2007, published online 24 August 2007 Small guanosine triphosphatases (GTPases) of the Rab family are recognized as key regulators of intracellular transport and membrane trafficking within eukaryotic cells (1). They function as molecular switches, cycling between GDP- and GTP-bound states and between cytosol and membranes. Through specific interactions with a wide variety of effectors, Rab GTPases participate in the formation of transport intermediates, in their movement along cytoskeletal tracks and in their docking/fusion with acceptor membranes. Rab proteins also play an important role in the organization and dynamics of membrane domains on organelles of the endocytic and exocytic pathways (2,3). The action of Rab GTPases needs to be co-ordinated in order to ensure the transport directionality and to maintain the identity of each intracellular compartment. Rab coupling can be accomplished by proteins that are able to interact with more than one Rab protein. Many examples of such proteins now exist in the literature, although functional data are still lacking for most of them. They can schematically be divided into two main classes. The first comprises ‘bifunctional’ Rab effectors that bind to Rabs associated with the same compartment or with compartments in dynamic continuity. Examples are Rabaptins 4 and 5 (4,5), Rabenosyn 5 (6) and Rabip 40 (7), which are each able to interact with both Rab4 and Rab5. These proteins are likely to be involved in the formation and organization of Rab4 and Rab5 domains on endosomal membranes as well as in the coordination of endocytic and recycling trafficking pathways (2,8). A protein, Rab coupling protein/Rab11-FIP1 (Family of Interacting Protein 1), that binds to both Rab4 and Rab11 has also been identified (9), but its role in coupling Rab4 and Rab11 function remains unclear (10). The second class comprises proteins that behave as an effector for one Rab and as an exchange factor (GTP exchange factor, GEF) for another Rab. The identification of these proteins led to the concept of a Rab cascade that explains how continuity can be ensured along transport pathways (1). The best characterized is Sec2p that serves as an effector for Golgiassociated Ypt31p/Ypt32p and as a GEF for Sec4p present on post-Golgi secretory vesicles (11). Another example is the class C VPS–HOPS complex (Vacuole Protein SortingHomotypic fusion and vacuole protein sorting), a Rab7 GEF, recently shown to be an effector of Rab5 and to participate in the maturation of early Rab5-positive endosomes into late Rab7-positive endosomes (12). www.traffic.dk 1385 Miserey-Lenkei et al. Rab11 and Rab6 are predominantly associated with membranes of the recycling endosomes (REs) and Golgi complex, respectively (13–15). Numerous studies have illustrated that members of the Rab11 subfamily, Rab11A, Rab11B and the closely related Rab25, function as regulators of endocytic membrane recycling in both polarized and nonpolarized cells [for reviews, see Zerial and McBride (2) and Hoekstra et al. (16)]. Work from our laboratory has also shown that the overexpression of the dominant-negative Rab11A mutant (S25N) or the addition of anti-Rab11 antibody to a permeabilized cell assay inhibits the delivery of internalized B fragment of Shiga toxin to trans Golgi network (TGN) membranes (17,18). This suggests that Rab11A, in addition to its function in recycling toward the plasma membrane, regulates transport events between REs and TGN. Rab6, for which three isoforms (Rab6A, A0 and B) have been described, regulates intra-Golgi and Golgi to endoplasmic reticulum retrograde transport (19–21). Rab6 also plays a role in endosome to Golgi trafficking, as documented by its participation in transport of Shiga and ricin toxins (18,22). Depletion of Rab6A0 by small interfering RNA (siRNA) or overexpression of the dominant-negative Rab6A0 mutant (T27N) strongly inhibit the delivery of Shiga toxin B fragment (STxB) to TGN (18,23). A likely scenario is that TGN-associated Rab6A0 participates in the tethering/docking of transport intermediates originating from early endosomes and/or REs and carrying STxB. This event requires the function of several effector molecules that are recruited onto Golgi membranes by Rab6:GTP, such as Rab6-interacting protein 2 (24), and may serve as a link between Rab6A0 and the TGNlocalized t-SNARE complex, composed of Syntaxin 6, Syntaxin 16 and Vti1a (18). Thus, Rab11 and Rab6 are both involved in the retrograde transport of STxB. Furthermore, the finding that modulating the activities of Rab6 and Rab11 using specific antibodies in the same semi-in vitro system does not result in additive inhibition of STxB transport suggests that these GTPases regulate two sequential retrograde transport steps between REs and TGN (18). In addition to their role in membrane trafficking in interphasic cells, recent data indicate that Rab11 and Rab6 also function in mitosis and cytokinesis. Rab11, in concert with its effector Rab11-FIP3, is required for the delivery of endocytic vesicles necessary for cell abscission at the apex of the cleavage furrow (25). Rab6A, through its interaction with Rabkinesin-6/MKlp2 and Rab6A0 through its interaction with GAPCenA and p150Glued, participates in signaling pathways involved in the metaphase/anaphase transition (unpublished results and 26). The aim of this study was to identify the proteins that could co-ordinate Rab6 and Rab11 function. Here, we report the functional characterization of a protein, Rab6interacting protein 1 (R6IP1), originally identified as a Rab6-binding protein in a yeast two-hybrid screen [clone 1386 B in (27)]. R6IP1, for which two isoforms are expressed (R6IP1A and R6IP1B), also binds to Rab11A. We provide evidence that R6IP1 acts in the Rab6 and Rab11 pathways both in interphase and during mitosis and cytokinesis. Results Expression and structure of R6IP1 We first cloned R6IP1 complementary DNA (cDNA) from a mouse brain library. Genome analyses revealed that mouse R6IP1 gene is composed of 23 exons. Two splice variants are found in the data banks, a short sequence (accession number AJ245569), corresponding to splicing of exon 2 that we named R6IP1A, and a longer sequence (accession number BC060230), which includes all exons, named R6IP1B. The 24 amino acid extension (exon 2) present in R6IP1B is located at the N-terminus of the protein (Figure 1A). In human Expressed Sequence Tag (EST) databases, only the sequence of R6IP1B is present, so we performed reverse transcriptase–polymerase chain reaction (RT-PCR) using primers flanking the 24 amino acid insertion in a variety of human cell lines to determine whether R6IP1A is also expressed in humans. In the human material tested, R6IP1B is by far the most abundant form of R6IP1. However, low amount of a RT-PCR product corresponding to R6IP1A could be detected in most cell lines, including HeLa cells (Figure 1B). The RTPCR experiments also revealed that in mouse, R6IP1A is predominantly expressed in brain, whereas R6IP1B is the major transcript in intestine (Figure 1B). R6IP1 is found in most species, except in yeast Saccharomyces cerevisiae. Both isoforms are present in chick, cow, zebra fish and tetraodon EST data banks. As in humans, only the R6IP1B isoform has been identified in rat and dog. On the contrary, insects such as bee, anopheles and fly appear to only express R6IP1A. Because of the divergence of the worm R6IP1 sequence compared with the mammal proteins, it was difficult to conclude which isoform is expressed in this organism. Figure 1A shows the structure of the R6IP1A/B protein. Its C-terminal half, which includes the Rab6-binding domain (Rab6BD) (see Figure 1), contains two RUN [RPIP8 (Rap2 interacting protein 8), UNC-14 and NESCA (new molecule containing SH3 at the carboxy-terminus)] domains separated by a PLAT (Polycystin-1, Lipoxygenase, Alpha Toxin) domain. RUN domains appear to be characteristic of proteins that interact with Ras-like GTPases (28). In particular, they are present in Rabip4, a Rab4 effector and Connexin 43-interacting protein (CIP-85), a Rab GTPase activating protein (GAP)-like protein (29,30). PLAT domains are b-sandwich structures thought to mediate interactions with lipids or membrane-associated proteins (31). Finally, the N-terminal half of R6IP1 possesses DENN (Differentially Expressed in Neoplastic versus Normal Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 Figure 1: Structure and expression of R6IP1. A) Schematic representation of R6IP1A and R6IP1B isoforms and some of the truncated forms used in this study. B) Reverse transcription–polymerase chain reaction (RT-PCR) analysis of R6IP1 expression in adult mouse tissues and in human material. Mouse RT products of different tissues and of R6IP1A and R6IP1B plasmid cDNAs were amplified using mousespecific primers (left), and human RT products and mouse R6IP1A plasmid cDNA in the last lane on the right were amplified using humanspecific primers. IMR32 and HTB13 are neuroblastoma- and glioma-derived cell lines, Ewing E and F are Ewing tumors Erg and Fli, neuroblast is neuroblastoma and De Luca is a rhabdoid tumor. Mr indicates molecular weight DNA markers, the size of which is indicated on the left. Expected sizes of R6IP1A and R6IP1B RT-PCR products are 204 and 276 nt, respectively. aa, amino acid; Ct, C-terminus; Nt, N-terminus. Cell) domains, which have been found in several proteins involved in Rab-mediated processes or regulation of mitogen-activated protein kinase signaling pathways (32). However, it should be pointed out that R6IP1 does only bind to Rab6-GTPgS in GST pull-down experiments performed at higher stringency (>250 mM NaCl, data not shown). R6IP1 interacts with both Rab6 and Rab11A Yeast two-hybrid assay and biochemical pull-down experiments were performed to study the interaction of R6IP1A/B with various Rab proteins. As shown in Table S1, the two isoforms strongly interact in the yeast two-hybrid system with the GTP-bound mutants of Rab6A, Rab6A0 and Rab6B (Q72L) but not with dominant-negative Rab6A and Rab6A0 (T27N) mutants. A weak interaction was found with Rab6A wild-type (wt) and a variable interaction with Rab6A0 wt. This suggests that R6IP1 interacts with Rab6 in a GTPdependent manner. However, no nucleotide specificity was found in glutathione S-transferase (GST) pull-down experiments performed in standard conditions. As shown in Figure 2A, 2B (right panel), the same amount of R6IP1 present in mouse brain post-nuclear supernatant (PNS) was bound to Rab6 incubated in the absence of nucleotide or loaded with GDP or GTPgS. A similar result was obtained using in vitro-translated R6IP1A or R6IP1B (data not shown). No interaction was found between R6IP1 and Rab4, Rab5 or Rab7 (Table S1). In contrast, a specific interaction was obtained with Rab11A by the yeast two-hybrid and GST pull-down assays. However, in contrast to Rab6, Rab11A binds to R6IP1A but not to R6IP1B (Table S1 and data not shown). R6IP1A interacts with the two GTP-locked mutants of Rab11A, Rab11AQ70L and Rab11AS20V but not with Rab11A wt and the dominant-negative mutant, Rab11AS25N (Table S1). The nucleotide dependence of Rab11A/R6IP1A interaction was preserved in GST pulldown experiments, as only Rab11A loaded with GTPgS is able to bind R6IP1 present in mouse brain PNS (Figure 2A, 2B right panel) or to purified R6IP1A (data not shown). We next compared the ability of Rab6 and Rab11 to bind R6IP1A. As shown in Figure 2B (left panel), approximately four times more R6IP1 can be bound to Rab6A/A0 :GTP compared with Rab11A:GTP, suggesting that R6IP1 has a better affinity for Rab6 than for Rab11A. Traffic 2007; 8: 1385–1403 1387 Miserey-Lenkei et al. Figure 2: R6IP1 interacts with Rab6 with no nucleotide specificity, whereas its binding to Rab11A is GTP dependent. Pull-down experiments were performed using mouse brain PNS, mostly producing R6IP1A, as shown in Figure 1B. A) Glutathione S-transferase (GST)–Rab fusion proteins were loaded with GDP or GTPgS or incubated in the absence of nucleotide (empty form). After incubation with PNS in the presence or absence of nucleotide, R6IP1 bound to GST or GST–Rab was revealed by Western blot analysis using anti-R6IP1 antibody. Input represents a 5% load of the total PNS used in all conditions. B) Comparison of R6IP1 binding to Rab6 and Rab11. On the left panel, R6IP1 bound to Rab loaded with GTPgS is shown, as either the ratio of Rab6A to Rab11 or that of Rab6A0 to Rab11. The means of two experiments are shown. On the right panel, the ratio of R6IP1 bound to Rab:GTPgS to that bound to Rab:GDP is presented. For Rab6A, the ratio is presented as means ! SEM of six experiments, for Rab6A0 nine experiments and for Rab11A eleven experiments. C) HeLa cells were transfected with GFP–R6IP1A (lanes a–f) or with GFP–R6IP1B (lanes g–l) together with myc-Rab6AQ72L (lanes a and b, g and h) or myc-Rab11AQ70L (lanes c and d, i and j) or myc-Rab11AS25N (lanes e and f, k and l). Untransfected cells were used as control (lanes m and n). Cells lysates were immunoprecipitated using a goat anti-myc antibody. GFP fusion proteins were detected using an anti-GFP antibody on the upper half of the blot; immunoprecipitation efficiency was controlled using goat anti-myc on the lower half of the gel. To, 1/30 of the cell lysate before immunoprecipitation; IP, immunoprecipitate; WB, western blotting. To further establish the interaction between R6IP1 and either Rab6 or Rab11A, we performed a series of coimmunoprecipitation experiments in cells coexpressing myc-Rab6A or myc-Rab11A constructs and either green fluorescent protein (GFP)–R6IP1B or GFP–R6IP1A. GFP– R6IP1A (Figure 2C, lanes a and b) and GFP–R6IP1B (Figure 2C, lanes g and h) could be coimmunoprecipitated with myc-Rab6AQ72L and myc-Rab6A‘Q72L (data not 1388 shown). GFP–R6IP1A was also coimmunoprecipitated with myc-Rab11AQ70L but not with myc-Rab11AS25N, as expected (Figure 2C, lanes c and d, e and f). Lower but detectable amounts of GFP–R6IP1B were found in complexes with myc-Rab11AQ70L (Figure 2C, lanes i and j). In contrast, no GFP–R6IP1B was immunoprecipitated with myc-Rab11AS25N (Figure 2C, lanes k and l). Although no interaction was detected in vitro between Rab11A and Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 Figure 3: The recruitment of R6IP1 to Golgi membranes is Rab6-dependent. A) HeLa cells were transfected with GFP–R6IP1A. Endogenous Rab6 was revealed with a specific antibody. B, C) HeLa cells were treated for 2 days with Rab6A/A0 siRNA and then transfected with GFP–R6IP1A alone (B) or with GFP–R6IP1A and wild-type (wt) Rab6A (C). Endogenous (B) and overexpressed (C) Rab6 was revealed with a specific antibody. D, E) HeLa cells were cotransfected with wt CFP–Rab6A and either YFP–R6IP1A (panel D) or YFP–R6IP1B (panel E). Each image represents a single confocal section (A–C) or a projection of confocal sections (D, E). Scale bars, 10 mm. R6IP1B, these experiments suggest that both proteins can directly or indirectly interact in vivo. Higher amounts of GFP–R6IP1A were found associated with Rab6AQ72L than with Rab11AQ70L, in good agreement with the results obtained in GST pull-down assays (Figure 2B, left panel). We next constructed a series of truncation mutants to map the Rab-binding domain(s) of R6IP1 (Figure 1A and data not shown). The minimal Rab6BD was identified using the yeast two-hybrid assay (Table S1) and is constituted by the segment encompassing residues 683–1066 (R6IP1A) or 707–1090 (R6IP1B). This domain includes a RUN–PLAT Traffic 2007; 8: 1385–1403 domain (Figure 1A), present in both R6IP1A and R6IP1B. Any further truncation of the Rab6BD led to a loss of Rab6/ R6IP1 interaction. Strikingly, an interaction between R6IP1A and Rab11A could only be obtained with the entire protein (Table S1). This suggests that secondary structures of R6IP1A, including both N and C extremities, are required for Rab11A recognition. R6IP1 is recruited by Rab6 on Golgi membranes and by Rab11 on REs Endogenous R6IP1 was visualized by immunofluorescence on HeLa cells using a polyclonal rabbit antibody 1389 Miserey-Lenkei et al. raised against recombinant full-length R6IP1A. This antibody also recognizes R6IP1B (data not shown). A faint and diffuse cytosolic staining was observed (data not shown). Fractionation experiments confirmed that endogenous R6IP1 is mainly found in the cytosol (Figure S1). The cytoplasmic localization of R6IP1 is consistent with the fact that R6IP1 does not bear any putative transmembrane domain. However, by moderately overexpressing GFP– R6IP1A or GFP–R6IP1B fusion proteins, part of GFP– R6IP1 was found to colocalize with Rab6 at the Golgi complex (Figure 3A). As analyzed by immunogold labeling on ultrathin cryosections, the majority of GFP–R6IP1A and GFP–R6IP1B revealed by anti-GFP antibodies were detected on Golgi stacks and on closely apposed tubulovesicular elements positive for the TGN marker TGN46 (see Figures 6A–C and S3, for quantification). In addition, the expression at moderate levels of Rab6A/A’ wt or Q72L mutants (data not shown) together with R6IP1A/B resulted in a nearly complete colocalization of both proteins at the Golgi complex, as illustrated in Figure 3D,E for cyan fluorescent protein (CFP)–Rab6A wt and yellow fluorescent protein (YFP)–R6IP1A or YFP– R6IP1B. In contrast, overexpressed R6IP1A/B remained cytosolic in cells cotransfected with Rab6A/A0 T27N mutants (data not shown). To investigate whether Rab6 is required to localize R6IP1A/B to the Golgi complex, endogenous Rab6A and Rab6A0 were depleted by siRNAs. In Rab6-depleted cells, overexpressed GFP–R6IP1A was found to be strictly cytosolic and no concentration was observed in the Golgi region (Figure 3B). In contrast, the re-expression of Rab6A wt (Figure 3C) or Rab6A0 wt (data not shown) in Rab6A/A0 -depleted cells restored the localization of GFP–R6IP1A to the Golgi (Figure 3C). Similar results were found for GFP–R6IP1B (data not shown). Finally, cells depleted of Rab11A by siRNAs showed no alteration in the Golgi localization of R6IP1A (Figure 4A). Altogether, the above experiments indicate that R6IP1A/ B is recruited to the Golgi in a Rab6-dependent manner. In addition, the recruitment of R6IP1A/B to the Golgi does not depend on its interaction with endogenous Rab11A. Figure 4: R6IP1 binds to Rab11-positive compartments in cells overexpressing Rab11A. A, B) HeLa cells were treated for 2 days with Rab11A siRNA and then transfected with GFP– R6IP1A alone (panel A), or with GFP– R6IP1A and wild-type (wt) Rab11A (panel B). Endogenous (A) and overexpressed (B) Rab11 was revealed with a specific antibody. C, D) HeLa cells were cotransfected with wt CFP–Rab11A and YFP– R6IP1A (panel C) or wt CFP–Rab11A and YFP–R6IP1B (panel D). Each image is a single confocal image (A) or a projection of confocal sections (B, C, D). Arrowheads in insets point to peripheral structures costained with Rab11 and R6IP1. Scale bars, 10 mm. 1390 Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 Figure 5: R6IP1 overexpression alters the distribution of peripheral REs. GFP–R6IP1A (panels A, C and E) or GFP–R6IP1B (panels B and D) constructs were overexpressed for 24 h. Cells were then processed for immunofluorescence using specific antibodies against Rab11 (panels A and B), TfR (panels C and D) or EEA1 (panel E). Images are single confocal sections. Scale bars, 10 mm. A strikingly different picture was obtained in cells coexpressing Rab11A wt or Rab11AQ70L with R6IP1A/B. In these cells, a strong overlap was observed between R6IP1A/B and Rab11 staining. This is illustrated in Figure 4B, showing a cell cotransfected with Rab11A wt and GFP– R6IP1A after Rab11-depletion by siRNA. GFP–R6IP1A was found to colocalize with Rab11A both in the pericentriolar region of the cells and in more peripheral structures characteristic of the recycling compartment in HeLa cells (Figure 4B, arrowheads in insets). Thus, higher amounts of Traffic 2007; 8: 1385–1403 Rab11A within the cells favor the recruitment of R6IP1A to Rab11-positive compartments. To further confirm this observation, we coexpressed YFP–R6IP1A with CFP– Rab11A wt. As illustrated in Figure 4C, a nearly complete colocalization was observed between the two proteins, both in pericentriolar and in peripheral REs (arrowheads in insets). As the Golgi complex and the Rab11-positive pericentriolar recycling compartment are located in the same area of the cell (illustrated in Figure S2 for a cell expressing YFP–R6IP1A and CFP–Rab11A), it was difficult, 1391 Miserey-Lenkei et al. Figure 6: R6IP1 is mainly localized to the Golgi apparatus. Ultrathin cryosections of HeLa cells expressing GFP–R6IP1A (A, C and D) and GFP– R6IP1B (B) were single immunogold labeled for GFP (A and B) or double immunogold labeled for GFP and TGN46 (C) or GFP and HRP (D). Antibodies were visualized with protein A conjugated to 15- or 10-nm gold particles as indicated. In D, cells were allowed to internalize Tf coupled to HRP for 45 min before fixation. A, B) Green fluorescent protein–R6IP1A and GFP–R6IP1B are detected not only in the cisternae of the Golgi apparatus (GA) but also in closely apposed vesicular elements (arrows in B). C) As in A, note the localization of GFP–R6IP1A in the GA, including in vesicles at the trans side of the GA (arrows). In addition, GFP–R6IP1A is also present in vacuolar endosomal structures displaying intraluminal membranes (arrowheads). D) Internalized Tf–HRP can be detected close to the GA (arrowheads), but it is not present in the GA cisternae, containing the bulk of GFP– R6IP1A. PM, plasma membrane. Scale bars, 200 nm. however, to determine whether R6IP1A was still associated with Golgi in cells overexpressing active Rab11. Similar results were obtained after coexpression of Rab11A and R6IP1B (Figure 4D, cell on the left). Nevertheless, the colocalization between YFP–R6IP1B and CFP–Rab11A appeared less complete than the one observed between YFP– R6IP1A and CFP–Rab11A (Figure 4D, cell on the right). Altogether, the above results suggest that R6IP1 behaves as a Rab6-interacting protein in vivo and is targeted to the Golgi in a Rab6-dependent manner. Moreover, R6IP1 can associate to Rab11-positive compartments when levels of Rab11A are increased within the cells. The overexpression of R6IP1 affects the spatial distribution of the REs We noticed that the overexpression of R6IP1 alters endogenous Rab11 staining. As shown in Figure 5A,B (outlined cells), R6IP1 overexpression resulted in a depletion of 1392 peripheral Rab11-positive compartments, including those present in cell tips, and in a concomitant concentration in the pericentriolar area. Transfected cells also displayed an accumulation of transferrin receptor (TfR) in the pericentriolar area (Figure 5C,D). In contrast, the distribution of the early endosome marker Early Endosomal Antigen 1 (EEA1) (Figure 5E) as well as the Golgi markers CTR433 and GM130 (data not shown) remained unaffected. At the ultracryomicroscopy level, endosomal compartments with a vacuolar appearance and with very few internal membranes were found close to the Golgi in cells overexpressing R6IP1 (illustrated in Figure 6C,D for R6IP1A). R6IP1A molecules were present on the limiting membrane of these compartments that often contained internalized Horse Radish Peroxydase (HRP)– transferrin (Tf) (Figure 6D). However, labeled membranes with R6IP1A/B and HRP–Tf appeared distinct from Golgi cisternae, and no HRP labeling could be found within them. Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 Thus, the overexpression of R6IP1A/B leads to concentration of Rab11- and TfR-positive recycling compartments to the pericentriolar area but does not seem to induce fusion events between endosomal and Golgi membranes. In vivo interactions between Rab6 and Rab11A in the presence of R6IP1 by two-photon fluorescence resonance energy transfer/fluorescence lifetime imaging The above results suggest that the overexpression of R6IP1A/B brings closer Rab6- and Rab11-positive compartments. To investigate whether an interaction between Rab6 and Rab11A after R6IP1A/B overexpression could mechanistically explain this topological gathering, we set up twophoton fluorescence resonance energy transfer (FRET)/ fluorescence lifetime imaging (FLIM) experiments in living cells. A first series of FLIM experiments was performed in cells expressing only CFP–Rab6A to obtain a reference value for the mean fluorescence lifetime of CFP without any FRET (Figure 7A, top left, and Figure 7B, first row). An average value of 2.4 nanoseconds was obtained, which corresponds to the expected and published value for the CFP (33). We next measured the fluorescence lifetime of CFP in cells coexpressing CFP–Rab6A or CFP-Rab6A0 and YFP–Rab11A. Image analysis was performed in the pericentriolar region of the cells, where the major part of Rab6 and Rab11A signal is detected (Figure 7A, top right). However, the low signal intensity and the acquisition time of FLIM images (300 seconds) precluded a reliable analysis of the fast-moving peripheral Rab11-positive vesicles. CFP lifetime displayed a value similar to the reference value for CFP alone (Figure 7B, first row, and Figure 7C), indicating that no interaction between Rab6A/A0 and Rab11A in transfected cells could be detected. In cells cotransfected with CFP–Rab6A or CFP–Rab6A0 and YFP–R6IP1A, values lower than 2.4 nanoseconds (Figure 7B, second row) were obtained, indicating that FRET occurred between the two fluorophores CFP and YFP and thus that Rab6A/A0 and R6IP1A interact. The broad distribution of the values shown on histograms may be due to variations in the relative position of the proteins and fluorophores. As the FRET efficiency varies as the inverse 6th power of the distance between the donor and the acceptor, slight changes can lead to significant variations in fluorescence lifetime (34). Similar experiments were performed in CFP–Rab11A-expressing cells cotransfected with YFP–R6IP1A (Figure 7A, bottom left). As shown in Figure 7B (second row), values lower than 2.4 nanoseconds were also found. Thus, the above results confirm the ability of Rab6A/A0 and Rab11A to interact in vivo with R6IP1A. We then measured the lifetime in cells overexpressing R6IP1A together with CFP–Rab6 and YFP–Rab11A. The CFP lifetime value was reduced in about 40% of cells, indicating that CFP–Rab6 and YFP–Rab11A were close Traffic 2007; 8: 1385–1403 enough for FRET to occur (Figure 7A, bottom right, Figure 7B, third row and Figure 7C). Similar results were obtained in cells expressing YFP–Rab11AQ70L instead of YFP– Rab11A wt (Figure 7C). In contrast, no FRET was observed in cells expressing YFP–Rab11AS25N (Figure 7B,C), in good agreement with previous data showing the GTP nucleotide specificity of the Rab11A/R6IP1A interaction (Figure 2). As a control for specificity, we performed experiments with cells expressing YFP–Rab4A, a Rab protein that partially localizes to the same compartments as Rab11A (8). No FRET occurred between YFP–Rab4A and CFP– Rab6A when R6IP1A was overexpressed (Figure 7C), supporting the specificity of the ternary association between Rab6, Rab11A and R6IP1. The above results thus indicate that R6IP1A can specifically trigger in vivo the formation of complexes containing Rab6 and Rab11A. R6IP1 may oligomerize to bind both Rab6 and Rab11A There are several possibilities to explain how R6IP1 can promote an interaction between Rab6 and Rab11A. One is that a single R6IP1 molecule is able to bind simultaneously to Rab6 and Rab11A. To investigate this, GST pull-down assays were performed by adding R6IP1A/B from various sources and Rab11A loaded with GTPgS sequentially to GST-Rab6 or by adding simultaneously the two proteins. However, no concomitant binding of R6IP1A/B and Rab11A was obtained (data not shown). Another possibility is that R6IP1 has to oligomerize in order to bind both Rab6 and Rab11A. To test this hypothesis, we performed a series of coimmunoprecipitation experiments using various R6IP1A/B constructs. As shown in Figure 8 (lanes a and b, c and d), full-length R6IP1A and R6IP1B can be coimmunoprecipitated with their respective N-terminal halves (Nt, Figure 1A). The N-terminal regions of R6IP1 can also interact in vivo with each other, as illustrated by coimmunoprecipitation of myc-tagged R6IP1A-Nt with GFP-tagged R6IP1A-Nt in cells expressing both constructs (Figure 8, lanes e and f). Under the same conditions, the C-terminal regions (Ct, Figure 1A) did not show any interaction (data not shown). This suggests that R6IP1 molecules can form in vivo homo-oligomers via their Nt domain. R6IP1A and R6IP1B also have the capability to form hetero-oligomers. The two proteins were found to weakly interact by yeast two-hybrid assay (Table S1), and GFP– R6IP1B can be coimmunoprecipitated with myc-tagged R6IP1A from cell lysates expressing both constructs (Figure 8, lanes g and h). Finally, we tested whether intramolecular interactions within R6IP1 take place. Weak but significant interactions over background were found between Ct and Nt halves of R6IP1A or B (Table S1). This interaction was confirmed by coimmunoprecipitation experiments (Figure 8, lanes i and j). These data suggest that R6IP1 form in vivo homo- or hetero-oligomers that allow binding of Rab6 and Rab11A 1393 Miserey-Lenkei et al. Figure 7: Legend on next page. 1394 Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 Figure 8: R6IP1 can form in vivo homo- or hetero-oligomers. HeLa cells were cotransfected with GFP– R6IP1B (g and h) or GFP–R6IP1B Nt (c and d) or GFP–R6IP1A (a and b) or GFP–R6IP1A Nt (e and f) or GFPR6IP1Ct (i and j), together with mycR6IP1B Ct (c and d) or myc-R6IP1A (g and h) or with myc-R6IP1A Nt (a and b, e and f, i and j). Cell lysates were then immunoprecipitated with a goat antimyc antibody, and GFP fusion proteins were detected by western blotting (WB) using an anti-GFP antibody. Molecular weight markers are indicated on the left (in kDa), the position of proteins of interest noted on the right. Nt, N-terminal; Ct, C-terminal; To, 1/30 of the cell lysate before immunoprecipitation; IP, immunoprecipitate. within the same complex. They also provide an explanation about the observation that GFP–R6IP1B, which does not interact in vitro with Rab11A, still associates to Rab11positive compartments in cells overexpressing Rab11A wt or Rab11AQ70L (Figure 4). R6IP1B may indeed form a complex with R6IP1A, which allows its binding to Rab11A. Finally, the existence of interactions between the Nt and the Ct regions suggests that R6IP1 may adopt a folded conformation. Overexpression of R6IP1 increases the rate of Rab11-dependent recycling of Tf We then investigated whether the redistribution of peripheral Rab11-positive compartments toward the pericentriolar area affects transport events associated with REs. We first monitored the transport of STxB between endosomes and Golgi, shown to involve in part Rab11-positive compartments (17). No significant changes in the kinetics of delivery of internalized Cy3-STxB to the TGN was observed after overexpression of R6IP1A/B (data not shown). The kinetics of the appearance of ts-O45 vesicular stomatitis virus G protein to the cell surface, that transits in part through the REs on its way to the plasma membrane (35), also remained unaffected (data not shown). However, an effect was observed on the rate of Tf recycling. After moderate overexpression of GFP–R6IP1A or GFP–R6IP1B, Tf uptake was comparable to that in control cells, but after a chase time of 30 min, Tf recycling was found to be slightly but consistently higher (10%) (Figure 9; Table S2). The colocalization of internalized Tf and Rab11 has been shown to be maximal after 30 min to 1 h in various cell types (14,36). In addition, an effect of Rab11 mutants on Tf recycling can be clearly detected only after 30-min chase period following Tf uptake (14,17). This suggests that the overexpression of R6IP1A/B affects the Rab11-dependent step of Tf recycling. It has been documented that about 20% of Tf recycles through the Rab11 pathway in HeLa cells (17) and other cell types (37). Thus, this indicates that the overexpression of R6IP1 increases by about 50% the rate of recycling of internalized Tf that passes through the Rab11-positive endosomes. The above results suggest that the concentration of Rab11-positive compartments in the pericentriolar area on R6IP1 overexpression increases the efficiency of TfR recycling but does not affect the transport pathway between endosomes and Golgi/TGN monitored by Shiga toxin. R6IP1 function is required for mitotic progression Interestingly, a strong effect of R6IP1 overexpression was observed during mitosis and cytokinesis. For these experiments, cells were transfected with R6IP1A/B and analyzed by time-lapse phase-contrast videomicroscopy. As shown in Table 1, a significant proportion of cells were blocked in Figure 7: Interaction between Rab6, R6IP1 and Rab11A in living cells by FRET/FLIM. A) Intensity (left) and FLIM images (middle) of a HeLa cell expressing CFP–Rab6A (top left), CFP–Rab11A/YFP–R6IP1A (bottom left), CFP–Rab6A/YFP–Rab11A (top right) or CFP–Rab6A/ R6IP1A/YFP–Rab11A (bottom right). The color coding of the FLIM image indicates the fluorescence lifetime value following the total distribution of lifetime in the image between 0 and 4000 picoseconds (right). B) Distributions of lifetime values over n cells expressing from left to right and from top to bottom CFP–Rab6A alone, CFP–Rab6A/YFP–Rab11A, CFP–Rab6A0 /YFP–Rab11A, CFP–Rab6A/YFP–R6IP1A, CFP–Rab6A0 /YFP–R6IP1A, CFP–Rab11A/YFP–R6IP1A or CFP–Rab6A/R6IP1A/YFP–Rab11A, CFP–Rab6A0 /R6IP1A/YFP–Rab11A and CFP– Rab6A/R6IP1A/YFP–Rab11AS25N. C) Graphs represent the synthesis of the mean filtered values of fluorescence lifetimes (left) and estimated FRET efficiency (right) ! SD obtained for the combinations of protein overexpression shown in B. Graphs also included data obtained for cells expressing CFP–Rab6A/R6IP1A/YFP–Rab11Q70L (n ¼ 20), CFP–Rab6A/R6IP1A/YFP–Rab4A (n ¼ 16) and CFP–Rab6A0 / R6IP1A/YFP–Rab4A (n ¼ 11). Traffic 2007; 8: 1385–1403 1395 Miserey-Lenkei et al. Figure 9: The overexpression of R6IP1 affects Tf recycling. HeLa cells transfected with GFP–R6IP1A or GFP–R6IP1B were loaded with Alexa633–Tf, as described in Material and Methods and chased for the various times indicated on the figure. Cells producing either GFP– R6IP1A (left) or GFP–R6IP1B (right) were sorted as no GFP–R6IP1 expression (control, blue curves) or as moderately expressing cells (GFP–R6IP1A or GFP–R6IP1B, red curves). Results were expressed as the percentage of Alexa633–Tf associated to cells at each timepoint compared with the maximum Alexa633–Tf associated to cells at t ¼ 0. Values are expressed as means ! SEM of four experiments for GFP–R6IP1A-expressing cells and of five experiments for GFP–R6IP1B-expressing cells. Paired Student’s t-tests are indicated in Table S2. metaphase 24–48 h after transfection. This phenotype was reminiscent to that observed after depletion of Rab6A0 or overexpression of the dominant negative form of Rab6A0 (26). In addition, the percentage of binucleated cells was increased after 3 days, suggesting a defect in cytokinesis (Table 1). To further understand these phenotypes, endogenous R6IP1 was depleted by siRNAs. As estimated by Western blotting, the transfection of cells with two different siRNAs (siR6IP1-1 and siR6IP1-2, see Material and Methods) led to an efficient knockdown of R6IP1 expression (Figure 10A). No major effects on the morphology of the Golgi complex or of endocytic compartments using various markers was Table 1: Effects of the overexpression of GFP–R6IP1A and GFP– R6IP1B on cell cycle. Cells were transfected with a control plasmid, GFP–R6IP1A or GFP–R6IP1B and imaged by time-lapse phase-contrast videomicroscopy. The percentage of cells arrested in metaphase was measured by analyzing the recorded movies. In parallel, cells were fixed 3 days after transfection and multinucleated cells were counted. Results are representative of two to three independent experiments. The number of cells analyzed ranges from 59 to 224. Results are expressed as means ! SEM Control plasmid GFP–R6IP1A GFP–R6IP1B 1396 % Of cells arrested in mitosis % Of multinucleated cells 3.9 ! 0.1 25 ! 5 13 ! 0.1 2.3 ! 0.3 14.2 ! 1.7 9.2 ! 1.2 observed in interphasic cells present in the R6IP1-depleted cell population (data not shown). Transport of internalized STxB toward the Golgi and Tf uptake and recycling did not also seem to be significantly affected. However, time-lapse videomicroscopy experiments showed that about 50% of R6IP1-depleted cells with siR6IP1-1 and siR6IP1-2 were arrested in metaphase and were unable to exit mitosis compared with 16% in control cells (Figure 10B, a). Arrested cells, which exhibited a normal alignment of their chromosomes at the metaphase plate (Movie S3), died after a few hours (Figure 10B, b and Movie S2 and S3). We analyzed this phenotype in further detail using the siRNA approach. Cells arrested in metaphase displayed a normal mitotic spindle, as judged by b-tubulin staining (data not shown). The majority of the cells (67%) showed three or more kinetochores labeled with the Mad2 antibody, suggesting an activation of the Mad2 spindle checkpoint (Figure 10C). This was further confirmed by silencing Mad2 by siRNA, which resulted in a 50% decrease of the metaphase arrest induced by the treatment with siR6IP1-1 (Figure 10D). Thus, the mitotic arrest observed in R6IP1-depleted cells is very similar to that obtained after alteration of Rab6A0 function, shown to induce a Mad2-dependent metaphase arrest (26). In addition, R6IP1-depleted cells displayed cytokinesis defects. We observed cells that start to divide normally but undergo re-fusion after an ineffective cleavage, giving rise to a binucleated cell (Figure 10E). Following a 72 h depletion with both R6IP1 siRNAs, 5% of cells were binucleated (Figure 10F). Interestingly, Wilson et al. have reported that Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 Figure 10: Legend on next page. Rab11 depletion or overexpression of Rab11 S25N in HeLa cells also results in cytokinesis defects, although the percentage of binucleated cells was significantly higher in their experiments (25). Traffic 2007; 8: 1385–1403 Figure S4 illustrates the localization of GFP–R6IP1A in mitotic cells. Up to anaphase, R6IP1A displays a diffuse cytosolic localization. Such a localization is also found for Rab6 (26; Figure S4) and Rab11, even though some 1397 Miserey-Lenkei et al. concentration can be detected around the spindle poles, as previously reported (38; Figure S4). Interestingly, in telophase and cytokinesis, GFP–R6IP1A associates to reforming Golgi and concentrates at the intracellular bridge where it is found colocalized with Rab6- and Rab11-positive structures. although the overall identity between the two proteins is only 5.2%. However, we have not been able to detect any GEF activity in vitro on Rab6 (data not shown). It should also be pointed out that the exact domain performing GEF function in Rab3 GEF has not been precisely mapped and that DENN domains are also not found in other Rab GEFs. Altogether, our data suggest that low amounts of R6IP1 are sufficient for its function in interphase and that highest levels of depletion give rise to strong mitotic phenotypes. They show that R6IP1 function is required during metaphase and cytokinesis, two mitotic steps in which the role of Rab6 and Rab11 has been previously documented. Without ruling out the above hypothesis, it is more likely that R6IP1 acts as a Rab6 effector recruited by Rab6 on Golgi membranes. This is consistent with the findings that R6IP1 is a cytosolic protein and that Rab6 depletion inhibits targeting of overexpressed R6IP1 to Golgi membranes. Along with this hypothesis, it would be the Rab6–R6IP1 complexes that interact, once formed, with Rab11. We obtained no evidence that a single R6IP1 molecule can bind simultaneously to Rab6 and Rab11, suggesting that Rab6/Rab11 interaction occurs through homo- and/or hetero-oligomerization of R6IP1 molecules. Rab6 binds to the C-terminal part of R6IP1, which has the potential to interact with its own N-terminus. It is thus possible that Rab6 binding leads to the opening of one R6IP1 molecule, which allows the binding of another R6IP1A or R6IP1B molecule and then of Rab11. R6IP1A and B are expressed at different levels depending on cell types and organisms. In addition, differences exist in the ability of Rab11A to bind either R6IP1A or R6IP1B in vitro, indicating that the insertion of a 24 amino acid stretch in R6IP1B affects the R6IP1/Rab11A interaction. These results suggest that Rab6/Rab11A interaction could be subjected to additional levels of regulation through the formation of R6IP1 homoor hetero-oligomers of different composition. Further experiments will be required to clarify these hypotheses. Discussion The relationships that exist between the Golgi complex and the REs are still poorly understood. However, it is now clear that the connection between these compartments occurs both in the retrograde and in the anterograde directions. REs may serve as a transit station toward the TGN for internalized toxins such as Shiga toxin and for resident TGN proteins such as TGN38 recycled back from the plasma membrane (39,40). Recent studies have highlighted that REs may also serve as a hub for the sorting of newly synthesized and endocytosed proteins (35). The identification of two proteins, R6IP1A and R6IP1B, that interact with both Rab6 and Rab11 provides to our knowledge the first molecular link between the two compartments. Several hypotheses are drawn to explain how R6IP1 may function in concert with Rab11 and Rab6. Taking into account the experimental evidence that Rab11A and Rab6A0 regulate two sequential transport steps of STxB fragment between endosomes and Golgi (18), R6IP1 could function in a Rab cascade mechanism between these two compartments. R6IP1 could be recruited on REs by Rab11A:GTP and then acts as an exchange factor for Rab6. It is noteworthy that the N-terminal half of R6IP1 shows similarity through its DENN domains with Rab3 GEF (41), In addition to REs, Rab11 was found to be associated with Golgi and/or TGN membranes in some cell types (14,42). However, endogenous Rab11 is not detected at the immunofluorescence level on Golgi/TGN membranes in HeLa cells (17). In these cells, the Rab11-positive compartment appears more scattered in the cytoplasm than in other cell types, such as Chinese Hamster Ovary (CHO) or PC12 cells, where it is predominantly concentrated in the Figure 10: R6IP1 is required for metaphase/anaphase transition and completion of cytokinesis. A) 72 h after transfection with siRNA that target R6IP1 (namely R6IP1-1 or -2) or control siRNA, cell lysates were subjected to Western blotting analysis and probed with an antiR6IP1 antibody (top) or anti-Rab11 antibody (bottom). B) a: Percentage of cells arrested in metaphase after transfection of control and R6IP1-1 or -2 siRNA. b: HeLa cells were transfected with control (top) or R6IP1-1 siRNA (bottom) and imaged for 72 h using time-lapse phase-contrast videomicroscopy. Arrowheads point to R6IP1-depleted cells arrested in metaphase. Note the alignment of chromosomes on the metaphase plate (Movie S3). The numbers correspond to the time in hours after the beginning of the recording. The corresponding movies (Movies S1,S2 and S3) are presented in supplementary online material. C) Maximal intensity projection through the z-axis of deconvolved images stacks of metaphasic cells 48 h after transfection with R6IP1 siRNA. Cells were costained with a CREST (calcinosis, Raynaud phenomenon, esophageal dismotility, sclerodactyly, telangiectasia) serum (left, green), anti-Mad2 antibody (middle, red) and DAPI (40 ,6-diamidino-2-phenylindole) (right, blue). Higher magnification for details of localization of CREST and Mad2 staining is shown in insets. Bar, 10 mm. D) Percentage of cells arrested in metaphase after cotransfection of R6IP1-1 siRNA with either control or Mad2 siRNA. Results are expressed as a ratio to the percentage found in cells cotransfected with R6IP1 and control siRNA. E) Details of a cell undergoing cytokinesis defect after R6IP1-1 siRNA. The numbers correspond to the time in hours after the beginning of the recording. Arrowheads point to a cell undergoing a cytokinesis defect. The corresponding movie (Movie S4) is presented in supplementary online material. F) Percentage of multi-nucleated cells (counted on fixed sample) after transfection of control or R6IP1-1 and -2 siRNA. In B, D and F, number of cells analyzed ranged from 48 to 86. Results are representative of two to three independent experiments and are presented as means ! SEM. 1398 Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 pericentriolar region. We observed that the overexpression of R6IP1 results in a depletion of peripheral Rab11-positive REs, indicating that increased amount of R6IP1 affects their dynamics. A likely explanation of this effect is that Rab6–R6IP1 complexes interact with Rab11 when Rab11positive peripheral structures move from the periphery to the cell center. In this case, excess amount of R6IP1 would be expected to shift the steady-state repartition of peripheral and pericentriolar REs. Thus, a tempting hypothesis is that Rab6–R6IP1 complexes participate in the tethering/ docking of transport intermediates that carry STxB between endosomes and Golgi/TGN membranes. This hypothesis is strongly supported by the results of FLIM– FRET analysis performed on living cells. It would also fit with previous results, indicating that part of STxB en route to TGN transits through REs (18). We do not exclude, however, other mechanisms accounting for the redistribution of peripheral REs. This phenotype is reminiscent of that obtained by overexpressing GTPase-deficient mutant of Golgi-associated Rab34, shown to mediate a long-range regulation of the spatial distribution of lysosomes (43). Rab34 was proposed to recruit the Rab7 effector RabInteracting Lysosomal Protein (RILP), which results in the modulation of the activity of the dynein–dynactin complex to promote the migration of lysosomes toward the minus end of microtubules (43). As Rab6 has been shown to interact with the same motor protein complex (26,44,45), a similar scenario could be envisioned for the effect of R6IP1 on the Rab11-positive compartment distribution. The concentration of Rab11-positive recycling compartments in the pericentriolar area on R6IP1 was correlated to a significant acceleration of TfR recycling. One hypothesis is that this effect is because of a shunting of the TfR into the recycling pathway that does not require passage through REs, i.e. early endosomes. However, the effect of R6IP1 overexpression was detected after a 30-min chase period, which is not compatible with the fast Rab4-dependent recycling step of internalized Tf. Another possibility is that TfR recycles via an alternative pathway. For instance, TfR could recycle faster because it enters the secretory pathway at the Golgi/TGN level. However, we obtained no evidence at the ultracryomicroscopy level that R6IP1 overexpression induces a fusion between REs and Golgi/TGN membranes. A concentration of the REs in the perinuclear region may also favor the association of transport intermediates carrying TfR toward the cell surface with microtubules. Finally, it should be pointed out that the concentration of Rab11 and TfR in the pericentriolar area was documented after the overexpression of arfophilin-2 and a tail chimera of myosin Vb, two other Rab11 effectors (46–48). Whereas the recycling of TfR is not significantly perturbed by the overexpression of arfophilin2, it is retarded in cells expressing myosin Vb mutant. A better understanding of the molecular mechanisms involved in slow TfR recycling should help us to understand the various effects induced by the three Rab11 effectors. In particular, it will be of interest to investigate the relationship Traffic 2007; 8: 1385–1403 between R6IP1 and myosin Vb, shown to function in traffic between peripheral and pericentriolar REs (48). Our results indicate that R6IP1 also plays a role in the Rab6 and Rab11 pathways during mitosis and cytokinesis. The block in metaphase induced by R6IP1 depletion is very similar to the one obtained after depletion of Rab6A0 , leading to an activation of the Mad2 spindle checkpoint (26). This suggests that R6IP1 acts as a Rab6A0 effector at the metaphase/anaphase transition. The observation that R6IP1 overexpression has the same effect as its depletion could be explained by an inhibitory effect on the interaction between Rab6A0 and another effector required for Rab6A0 function. We have shown that p150Glued, a subunit of the dynein–dynactin complex, acts in the Rab6A0 pathway and that Rab6A0 /p150Glued interaction is important for the dynamics of the dynein–dynactin complex at the kinetochores (26). Rab6A0 pathway also involves GAPCenA (49), acting in this case not as a GAP but rather as an effector protein. We have proposed that GAPCenA interacts through its N-terminal phosphotyrosine-binding domain with other regulatory proteins involved in Rab6A0 function (26). R6IP1 could be one of these proteins. R6IP1 is present in the interphasic cytosol in fractions of very high molecular mass containing GAPCenA and p150Glued (data not shown). Future studies should establish whether it is also the case in mitosis. A significant proportion of R6IP1-depleted cells or R6IP1overexpressing cells, likely those that succeeded to pass the metaphase block, were binucleated. This suggests that R6IP1, as previously documented for Rab11 (25), is required for cytokinesis. The exact role of Rab11 in cytokinesis is not yet fully understood. It has been proposed that Rab11–FIP3 complexes are involved in the tethering/fusion of vesicles at the apex of the cleavage furrow, allowing the delivery of proteins/membranes necessary for abscission of daughter cells (25,50). R6IP1 could be then another component of this Rab11-driven machinery. Finally, it should be pointed out that Rab11 and Rab6 could function at other stages of mitosis and cytokinesis than those previously described. For instance, Rab11 and several of its effectors can be found associated with vesicles of unknown composition and function clustered around the spindle poles after metaphase (38 and this study). Rab6 (this study) and several effectors of Rab6, including GAPCenA, also accumulate at the bridge during cytokinesis (26; unpublished data). Thus, a tentative hypothesis is that R6IP1 links the functions of Rab6 and Rab11 at metaphase and during cytokinesis, as described in interphase. Materials and Methods Cloning of R6IP1 The partial cDNA for R6IP1 was originally identified from a yeast two-hybrid screen of a mouse brain library using Rab6A Q22V as a bait [clone B; (27)]. The full-length cDNA was obtained by a combination of a new screen of mouse brain l-ZAP2 library and a pseudo-RACE (rapid amplification of cDNA ends) using a mouse brain cDNA library to get the 50 region of the cDNA. 1399 Miserey-Lenkei et al. Oligonucleotides and siRNAs The following oligonucleotides were synthesized using accession numbers NM_015312 for human R6IP1B and NM_021494.1 for mouse R6IP1B: sense oligonucleotide on1 corresponding to nucleotides (nt) 292–311 in humans (AGTCGCTTCGCCGACTACTT) (the underlined C is replaced by a T in mouse) and 277–296 in mouse upstream from the insertion; anti-sense oligonucleotide on2 corresponding to nt 568–550 in humans (CCAGCCCTTTCGGCATACA) and 553–535 in mouse (CCAGCCCTTTGGGCATGCA); and anti-sense oligonucleotide on3 corresponding to nt 813–794 in humans (GGTCACAGGAGTGTCTTCAC) and nt 724–704 in mouse (GCATGTGGTAGAGGGTCTGCA). Two siRNAs for R6IP1 were designed and produced by Qiagen. The target sequence of siRNA R6IP1-1 is AGATGCAGTAGGAATGCTA and the one of siRNA R6IP1-2 is CGAAAGGGCTGGCATTCAA. To knockdown Rab6A and A0 , we used the siRNA (siRab6A/A0 ) common for both Rab6 isoforms (23). Rab11A siRNA is described in (51). Luciferase siRNA (Proligo) was used as control. Reverse transcription–polymerase chain reaction Reverse transcription–polymerase chain reaction experiments were performed from mouse tissues (brain, leg muscle and small intestine), human cell lines (HeLa; HTB13, derived from astrocytoma, and IMR32, derived from neuroblastoma) and human tumor (Ewing tumors Erg and Fli, De Luca rhabdoid tumor and neuroblastoma). Human materials were a generous gift of Dr O. Delattre, Institut Curie, Paris. Total RNA and messenger RNA (mRNA) were prepared by utilizing the Nucleospin RNA II kit (MachereyNagel) or the Fast Track Mag mRNA kit (Invitrogen), respectively. Messenger RNAs were then reverse transcribed using reverse transcriptase (Finnzymes) and on3 oligonucleotide corresponding to human and mouse R6IP1 transcripts. Subsequently, single-stranded cDNAs were amplified using on1 and on2 oligonucleotides by PCR. The expected amplified cDNA is 204 nt for R6IP1A or 276 nt for R6IP1B. Control plasmids encoding either sequence were used as controls. Cell transfection Cells were trypsinized the day before transfection. For plasmid DNA transfection, the calcium phosphate precipitate technique was used on cells cultivated at about 70% confluency, using 2–4 mg total DNA in a well of a 24-well plate. For double-stranded siRNA transfection, transfection was performed using oligofectamine or HiPerFect reagent according to the manufacturer’s (Qiagen) procedure, except that the siRNA transfection mix was applied to cells in suspension. Two microliters of oligofectamine plus 0.2 mM siRNA in 250 mL total volume or 3 mL HiPerFect plus 10 nM siRNA in 600 mL total volume were used per well of 24-well plate. Oligofectamine mix was removed several hours before cells were trypsinized the next day (day 1) and plated as needed. For siRNA transfection followed by DNA transfection, 90% confluent, adherent cells were transfected with siRNA using Lipofectamine 2000 (day 0). Cells were trypsinized the next day (day 1) and transfected again with plasmid DNA using calcium phosphate precipitate (day 2). DNA transfection was usually stopped after 24 h. Live cell imaging Transfected cells were plated at day 1 in multiwell, uncoated glass dish, and time-lapse sequences recorded at 10-min intervals as previously described in Miserey-Lenkei et al. (26). For FLIM/FRET experiments, cells were plated on 32- or 40-mm glass coverslip before transfection. During observation, culture medium was replaced by Leibowitz medium (L15; Gibco) for CO2 compensation and to avoid additional fluorescence from the medium. Two-photon FRET–FLIM experiments The experimental setup is fully described elsewhere (52). Briefly, twophoton FRET/FLIM experiments were performed on a Leica SP2 AOBS confocal scanner coupled to a femtosecond pulsed laser (Mira 900F; Coherent Inc.). Lifetime measurements were performed using the timecorrelated single photon counting technique with Multi Channel Plate 1400 Photomultiplier Tube (MCP PMT) detector (R3809U-52; Hamamatsu) and SPC 730 card (Becker&Hickl, Berlin, Germany). This setup allowed us to measure the fluorescence decays for each pixel on an image. Images were acquired by continuously scanning the samples with the laser tuned at 830 nm for 300 seconds in order to get enough photons to build a fluorescence lifetime image. Lifetime images were analyzed with SPC IMAGE software (Becker&Hickl) by fitting data with either a single or a double exponential function after fluorescence intensity threshold in order to avoid pixels containing too few photons to achieve an adequate fitting. Mean lifetime for the whole image was calculated from the distribution of values corresponding to each pixel in the threshold image. For each experimental condition, the mean lifetime was measured on several cells, and the resulting distribution was filtered to decrease measurement noise: all values corresponding to an occurrence frequency lower than 5% of the maximum frequency were removed from the analyzed data. The fluorescence lifetime histograms show the data for n cells after this filtering and was calculated with bins of 100 picoseconds. Mean fluorescence lifetime values and FRET efficiencies ! SD were calculated by eliminating the cell population showing no FRET for cells expressing Rab6 (A or A0 ), R6IP1 and Rab11A constructs (most left three bars). FRET efficiencies were calculated from the fluorescence lifetimes using the expression E ¼ 1 # t/t0, where t0 is the value of fluorescence lifetime without acceptor. Glutathione S-transferase pull-down and immunoprecipitation experiments These experiments were performed as described in Monier et al. (24). We used, as a source of R6IP1, mouse brain PNS, R6IP1A from baculovirusinfected Sf9 cell lysates or purified R6IP1A from the same cells. Immunoprecipitation experiments were performed on HeLa cells transiently expressing various R6IP1 constructs and Rab proteins. Usually, Rab proteins were myc-tagged and R6IP1 were expressed in cells as GFP chimera. After cotransfection with equivalent amount of both plasmids (10 mg of each plasmid in a 10-cm-diameter dish), HeLa cells were lysed in a buffer containing 25 mM Tris pH 7.5, 150 mM NaCl, 10 mM MgCl2 and 0.5% Triton-X-100. Cell lysates were then centrifuged at 1000 $ g. 2.6 mg of goat anti-Myc (Santa Cruz) was added together with protein G– Sepharose. Immunoprecipitates were analyzed by SDS–PAGE, GFP fusion proteins detected using an anti-GFP antibody (Roche) and myc-tagged proteins detected using the goat anti-myc antibody. Transferrin recycling Cells transfected for 24 h with various GFP fusion proteins were incubated for 1 h in fetal calf serum-free medium. They were then harvested using Versen (Gibco BRL) at 1/25 and processed for Alexa 633–Tf binding and recycling as described by Perez et al. (52). Briefly, cells were incubated for 1 h at 378C in the presence of 5 mg/mL Alexa 633–Tf (Molecular Probes) in DMEM complemented with 10 mM HEPES pH 7.4 and 0.1% BSA. Cells were then diluted in ice-cold PBS, pelleted and resuspended in cold DMEM containing 10 mM HEPES pH 7.4, 10% BSA and 100 mg/mL unlabeled holotransferrin (Sigma). One aliquot (t0) was removed in duplicate and diluted in five volumes of ice-cold PBS containing 50 mg/mL unlabeled Tf and incubated for about 10 min on ice, and cells were pelleted and fixed in 1% paraformaldehyde in Versen on ice. The rest of the cell suspension was incubated at 378C with occasional stirring. Aliquots were removed at various time-points until 60 min, and processed as t0. Fluorescenceactivated cell sorter (FACS) analysis was performed using a FACScalibur as described previously (53). Electron microscopy Transfected HeLa cells were allowed to internalize Tf–HRP (US Biologicals) or Tf–Biotin conjugates (Invitrogen) for 45 min and fixed with 2% (w/v) paraformaldehyde with 0.2% (w/v) glutaraldehyde, in 0.1 M phosphate Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 buffer, pH 7.4. Cell pellets were processed for ultracryomicrotomy as previously described (54). Ultrathin cryosections were prepared with an ultracryomicrotome Ultracut FCS (Leica, Vienna, Austria), and single or double immunogold labeled with the indicated primary antibodies and using protein A conjugated to 10- or 15-nm gold [PAG10 and PAG15 (protein A-colloidal gold)]. Sections were analyzed under a Philips CM120 electron microscope (FEI, Eindhoven, the Netherlands), and digital acquisitions were made with a numeric camera Keen View (Soft Imaging System, Munster, Germany). Others R6IP1A was purified from Sf9 cells infected with baculovirus pFASTBac encoding the full-length mouse R6IP1A protein. Two rabbit antisera were raised against the protein (Eurogentech) and characterized by immunofluorescence and Western blotting. Immunofluorescence was performed as previously described in Miserey-Lenkei et al. (26). For experiments involving round metaphasic cells, coverslips were coated with 1 mg/mL fibronectin (Chemicon) plus 0.0005% collagen (Sigma) prior to cell plating. Endogenous Rab6 and Rab11 were revealed using previously described specific antibodies (13,17) or using commercially available anti-Rab6 (Santa Cruz) and anti-Rab11 (Zymed) antibodies. endosomes/lysosomes) and accessibility to internalized Tf–HRP (endosomes). Tubulovesicular membranes that were located at the trans side of the Golgi were considered as TGN. Endosomes were defined as electron-lucent vacuoles with no internal membranes or few internal vesicles and tubulovesicular elements closely apposed to the vacuoles. MVBs (Multivesicular bodies) were compartments delimited by a membrane with numerous internal vesicles. Electron-dense compartments with no or few internal membranes were classified as lysosomes. Other vesicular structures dispersed in the cytoplasm with no assigned origin were considered as ‘cytoplasmic vesicles’. Labeling of mitochondria with anti-GFP antibodies was less than 0.1%. LAMP, Lysosomal-Associated Membrane Protein; MVB, multivesicular bodies. Figure S4: Localization of GFP–R6IP1A and endogenous Rab6 and Rab11 in metaphase and cytokinesis. Cells were transfected for 24 h with a plasmid encoding GFP–R6IP1A and then processed for immunofluorescence. Cells were labeled with antibodies against Rab6, Rab11 and b-tubulin. In C, images were deconvolved for better visualization of Rab6positive structures in the intracellular bridge (arrowheads). Scale bars, 10 mm. Similar localization was found for GFP–R6IP1B (data not shown). Table S1: Yeast two-hybrid assays Acknowledgments We thank Florence Niedergang for her help with FACS analysis, Yohanns Bellaı̈che, Arnaud Echard and Joanne Young for critical reading of the manuscript. We also thank Jean-Baptiste Sibarita and Vincent Fraisier, members of the ‘RIO-Cell and Tissue Imaging Facility at Institut Curie’, for their help in image acquisition and restoration processing. This work was supported by the Institut Curie, the CNRS and the European Union (LSHMCT-2004-503228, Project acronyme ‘Signalling & Traffic’, http://www. signallingtraffic.com). Supplementary Materials Figure S1: R6IP1 is a cytosolic protein. Post-nuclear supernatants were prepared from mouse brain (lane a) and from HeLa cells (lane b). HeLa cell PNS (lane c) was then centrifuged at 100 000 $ g in order to sort membranes (mb, lane e) from the cytosol (lane d). Western blotting was performed using a polyclonal rabbit antiserum generated against recombinant R6IP1A purified from baculovirus-infected cells. The antibody recognized in mouse PNS (lane a) a major band migrating at an apparent molecular mass of 130 kDa, corresponding to the predicted size of the protein. In HeLa cell PNS (lane b), the antibody stained additional bands. They likely correspond to unrelated proteins because they were still present after silencing of R6IP1 by siRNAs (data not shown). Molecular weight markers are shown on the right. Figure S2: Localization of the Golgi apparatus in CFP–Rab11A and YFP–R6IP1A coexpressing cells. Cells transfected with CFP–Rab11A and YFP–R6IP1A (corresponding to Figure 4C) were stained for endogenous Rab6 (blue). Figure S3: Quantification of subcellular localization of GFP–R6IP1A/B. The relative distribution of R6IP1A and R6IP1B in transfected HeLa cells was evaluated by analyzing directly under the electron microscope selected cell profiles from two distinct grids. A total of 764 and 695 gold particles were counted for R6IP1A and R6IP1B, respectively, and assigned to the compartment over which they were located. The definition of the distinct compartments was based on their morphology and their previous characterization by immunogold labeling with organelle markers (TGN46 for the TGN and Lysosomal-Associated Membrane Protein-1 (LAMP-1) for late Traffic 2007; 8: 1385–1403 Table S2: Statistics of Tf recycling experiments (Figure 9; p of Student’s t-test) Movies: HeLa cells treated with control (Movie S1) or R6IP1 siRNA (movie S2) were analyzed for 72 h by time-lapse phase-contrast videomicroscopy. Some R6IP1-siRNA-treated cells (Movie S2, arrow) entered mitosis normally, rounded up, but were subsequently arrested in mitosis and died after several hours. Details of a R6IP1-siRNA-treated cell undergoing a metaphase arrest (Movie S3, arrow); note the alignment of chromosomes on the metaphase plate during the metaphase block (arrowhead). Details of a cell (arrow) undergoing cytokinesis defect after R6IP1 siRNA (Movie S4). The numbers correspond to the time in minutes after the beginning of the recording. Supplemental materials are available as part of the online article at http:// www.blackwell-synergy.com References 1. Grosshans BL, Ortiz D, Novick P. Rabs and their effectors: achieving specificity in membrane traffic. Proc Natl Acad Sci U S A 2006; 103:11821–11827. 2. Zerial M, McBride H. Rab proteins as membrane organizers. Nat Rev Mol Cell Biol 2001;2:107–117. 3. Sannerud R, Marie M, Nizak C, Dale HA, Pernet-Gallay K, Perez F, Goud B, Saraste J. Rab1 defines a novel pathway connecting the pre-Golgi intermediate compartment with the cell periphery. Mol Biol Cell 2006;17:1514–1526. 4. Vitale G, Rybin V, Christoforidis S, Thornqvist P, McCaffrey M, Stenmark H, Zerial M. Distinct Rab-binding domains mediate the interaction of Rabaptin-5 with GTP-bound Rab4 and Rab5. EMBO J 1998; 17:1941–1951. 5. Nagelkerken B, Van Anken E, Van Raak M, Gerez L, Mohrmann K, Van Uden N, Holthuizen J, Pelkmans L, Van Der Sluijs P. Rabaptin4, a novel effector of the small GTPase rab4a, is recruited to perinuclear recycling vesicles. Biochem J 2000;346:593–601. 6. de Renzis S, Sonnichsen B, Zerial M. Divalent rab effectors regulate the sub-compartmental organization and sorting of early endosomes. Nat Cell Biol 2002;4:124–133. 7. Fouraux MA, Deneka M, Ivan V, van der Heijden A, Raymackers J, van Suylekom D, van Venrooij WJ, van der Sluijs P, Pruijn GJ. Rabip4’ is an 1401 Miserey-Lenkei et al. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. effector of rab5 and rab4 and regulates transport through early endosomes. Mol Biol Cell 2004;15:611–624. Sonnichsen B, De Renzis S, Nielsen E, Rietdorf J, Zerial M. Distinct membrane domains on endosomes in the recycling pathway visualized by multicolor imaging of Rab4, Rab5, and Rab11. J Cell Biol 2000; 149:901–914. Lindsay AJ, Hendrick AG, Cantalupo G, Senic-Matuglia F, Goud B, Bucci C, McCaffrey MW. Rab coupling protein (RCP), a novel Rab4 and Rab11 effector protein. J Biol Chem 2002;277:12190–12199. Peden AA, Schonteich E, Chun J, Junutula JR, Scheller RH, Prekeris R. The RCP-Rab11 complex regulates endocytic protein sorting. Mol Biol Cell 2004;15:3530–3541. Ortiz D, Medkova M, Walch-Solimena C, Novick P. Ypt32 recruits the Sec4p guanine nucleotide exchange factor, Sec2p, to secretory vesicles; evidence for a Rab cascade in yeast. J Cell Biol 2002;157:1005–1015. Rink J, Ghigo E, Kalaidzidis Y, Zerial M. Rab conversion as a mechanism of progression from early to late endosomes. Cell 2005;122:735–749. Goud B, Yang C, Roa M, Martinez O, Mayau V. Study of rab6, a ras-like GTP-binding protein associated with the Golgi complex. Ann N Y Acad Sci 1994;710:192–195. Ullrich O, Reinsch S, Urbe S, Zerial M, Parton RG. Rab11 regulates recycling through the pericentriolar recycling endosome. J Cell Biol 1996;135:913–924. Casanova JE, Wang X, Kumar R, Bhartur SG, Navarre J, Woodrum JE, Altschuler Y, Ray GS, Goldenring JR. Association of Rab25 and Rab11a with the apical recycling system of polarized Madin-Darby canine kidney cells. Mol Biol Cell 1999;10:47–61. Hoekstra D, Tyteca D, van IJzendoorn SC. The subapical compartment: a traffic center in membrane polarity development. J Cell Sci 2004; 117:2183–2192. Wilcke M, Johannes L, Galli T, Mayau V, Goud B, Salamero J. Rab11 regulates the compartmentalization of early endosomes required for efficient transport from early endosomes to the trans-Golgi network. J Cell Biol 2000;151:1207–1220. Mallard F, Tang BL, Galli T, Tenza D, Saint-Pol A, Yue X, Antony C, Hong W, Goud B, Johannes L. Early/recycling endosomes-to-TGN transport involves two SNARE complexes and a Rab6 isoform. J Cell Biol 2002;156:653–664. Martinez O, Antony C, Pehau-Arnaudet G, Berger EG, Salamero J, Goud B. GTP-bound forms of rab6 induce the redistribution of Golgi proteins into the endoplasmic reticulum. Proc Natl Acad Sci U S A 1997;94:1828–1833. White J, Johannes L, Mallard F, Girod A, Grill S, Reinsch S, Keller P, Tzschaschel B, Echard A, Goud B, Stelzer EH. Rab6 coordinates a novel Golgi to ER retrograde transport pathway in live cells. J Cell Biol 1999;147:743–760. Girod A, Storrie B, Simpson JC, Johannes L, Goud B, Roberts LM, Lord JM, Nilsson T, Pepperkok R. Evidence for a COP-I-independent transport route from the Golgi complex to the endoplasmic reticulum. Nat Cell Biol 1999;1:423–430. Utskarpen A, Slagsvold HH, Iversen TG, Walchli S, Sandvig K. Transport of ricin from endosomes to the Golgi apparatus is regulated by Rab6A and Rab6A’. Traffic 2006;7:663–672. Del Nery E, Miserey-Lenkei S, Falguieres T, Nizak C, Johannes L, Perez F, Goud B. Rab6A and Rab6A’ GTPases play non-overlapping roles in membrane trafficking. Traffic 2006;7:394–407. Monier S, Jollivet F, Janoueix-Lerosey I, Johannes L, Goud B. Characterization of novel Rab6-interacting proteins involved in endosome-to-TGN transport. Traffic 2002;3:289–297. Wilson GM, Fielding AB, Simon GC, Yu X, Andrews PD, Peden AA, Gould GW, Prekeris R. The FIP3-Rab11 protein complex regulates recycling endosome targeting to the cleavage furrow during late cytokinesis. Mol Biol Cell 2004;16:849–860. 1402 26. Miserey-Lenkei S, Couedel-Courteille A, Del Nery E, Bardin S, Piel M, Racine V, Sibarita JB, Perez F, Bornens M, Goud B. A role for the Rab6A’ GTPase in the inactivation of the Mad2-spindle checkpoint. EMBO J 2006;25:278–289. 27. Janoueix-Lerosey I, Jollivet F, Camonis J, Marche PN, Goud B. Twohybrid system screen with the small GTP-binding protein Rab6. Identification of a novel mouse GDP dissociation inhibitor isoform and two other potential partners of Rab6. J Biol Chem 1995;270:14801–14808. 28. Callebaut I, de Gunzburg J, Goud B, Mornon JP. RUN domains: a new family of domains involved in Ras-like GTPase signaling. Trends Biochem Sci 2001;26:79–83. 29. Mari M, Macia E, Le Marchand-Brustel Y, Cormont M. Role of the FYVE finger and the RUN domain for the subcellular localization of Rabip4. J Biol Chem 2001;276:42501–42508. 30. Lan Z, Kurata WE, Martyn KD, Jin C, Lau AF. Novel rab GAP-like protein, CIP85, interacts with connexin43 and induces its degradation. Biochemistry 2005;44:2385–2396. 31. Bateman A, Sandford R. The PLAT domain: a new piece in the PKD1 puzzle. Curr Biol 1999;9:R588–R590. 32. Levivier E, Goud B, Souchet M, Calmels TP, Mornon JP, Callebaut I. uDENN, DENN, and dDENN: indissociable domains in Rab and MAP kinases signaling pathways. Biochem Biophys Res Commun 2001; 287:688–695. 33. Tramier M, Gautier I, Piolot T, Ravalet S, Kemnitz K, Coppey J, Durieux C, Mignotte V, Coppey-Moisan M. Picosecond-hetero-FRET microscopy to probe protein-protein interactions in live cells. Biophys J 2002; 83:3570–3577. 34. Lakowicz J. Principles of Fluorescence Spectroscopy. 2nd edn. New York: Plenum Press; 1999. 35. Ang A, Taguchi T, Francis S, Fölsch H, Murrells LJ, Pypaert M, Warren G, Mellman I. Recycling endosomes can serve as intermediates during transport from the Golgi to the plasma membrane of MDCK cells. J Cell Biol 2004;167:531–543. 36. Ren M, Xu G, Zeng J, De Lemos-Chiarandini C, Adesnik M, Sabatini DD. Hydrolysis of GTP on rab11 is required for the direct delivery of transferrin from the pericentriolar recycling compartment to the cell surface but not from sorting endosomes. Proc Natl Acad Sci U S A 1998;95:6187–6192. 37. Sheff D, Pelletier L, O’Connell CB, Warren G, Mellman I. Transferrin receptor recycling in the absence of perinuclear recycling endosomes. J Cell Biol 2002;156:797–804. 38. Hobdy-Henderson KC, Hales CM, Lapierre LA, Cheney RE, Goldenring JR. Dynamics of the apical plasma membrane recycling system during cell division. Traffic 2003;4:681–693. 39. Mallard F, Antony C, Tenza D, Salamero J, Goud B, Johannes L. Direct pathway from early/recycling endosomes to the Golgi apparatus revealed through the study of shiga toxin B-fragment transport. J Cell Biol 1998;143:973–990. 40. Ghosh RN, Mallet WG, Soe TT, McGraw TE, Maxfield FR. An endocytosed TGN38 chimeric protein is delivered to the TGN after trafficking through the endocytic recycling compartment in CHO cells. J Cell Biol 1998;142:923–936. 41. Wada M, Nakanishi H, Satoh A, Hirano H, Obaishi H, Matsuura Y, Takai Y. Isolation and characterization of a GDP/GTP exchange protein specific for the Rab3 subfamily small G proteins. J Biol Chem 1997; 272:3875–3878. 42. Urbe S, Huber LA, Zerial M, Tooze SA, Parton RG. Rab11, a small GTPase associated with both constitutive and regulated secretory pathways in PC12 cells. FEBS Lett 1993;334:175–182. 43. Wang T, Wong KK, Hong W. A unique region of RILP distinguishes it from its related proteins in its regulation of lysosomal morphology and interaction with Rab7 and Rab34. Mol Biol Cell 2004;15: 815–826. Traffic 2007; 8: 1385–1403 Functional Characterization of Rab6IP1 44. Matanis T, Akhmanova A, Wulf P, Del Nery E, Weide T, Stepanova T, Galjart N, Grosveld F, Goud B, De Zeeuw CI, Barnekow A, Hoogenraad CC. Bicaudal-D regulates COPI-independent Golgi-ER transport by recruit ing the dynein-dynactin motor complex. Nat Cell Biol 2002;4: 986–992. 45. Short B, Preisinger C, Schaletzky J, Kopajtich R, Barr FA. The Rab6 GTPase regulates recruitment of the dynactin complex to Golgi membranes. Curr Biol 2002;12:1792–1795. 46. Hickson GR, Matheson J, Riggs B, Maier VH, Fielding AB, Prekeris R, Sullivan W, Barr FA, Gould GW. Arfophilins are dual Arf/Rab 11 binding proteins that regulate recycling endosome distribution and are related to Drosophila nuclear fallout. Mol Biol Cell 2003;14: 2908–2920. 47. Provance DW Jr, Gourley CR, Silan CM, Cameron LC, Shokat KM, Goldenring JR, Shah K, Gillespie PG, Mercer JA. Chemical-genetic inhibition of a sensitized mutant myosin Vb demonstrates a role in peripheral-pericentriolar membrane traffic. Proc Natl Acad Sci U S A 2004;101:1868–1873. 48. Lapierre LA, Kumar R, Hales CM, Navarre J, Bhartur SG, Burnette JO, Provance DW Jr, Mercer JA, Bahler M, Goldenring JR. Myosin vb is associated with plasma membrane recycling systems. Mol Biol Cell 2001;12:1843–1857. Traffic 2007; 8: 1385–1403 49. Cuif MH, Possmayer F, Zander H, Bordes N, Jollivet F, CouedelCourteille A, Janoueix-Lerosey I, Langsley G, Bornens M, Goud B. Characterization of GAPCenA, a GTPase activating protein for Rab6, part of which associates with the centrosome. EMBO J 1999;18: 1772–1782. 50. Yu X, Prekeris R, Gould GW. Role of endosomal rab GTPases in cytokinesis. Eur J Cell Biol 2007;86:25–35. 51. Uzan-Gafsou S, Bausinger H, Proamer F, Monier S, Lipsker D, Cazenave JP, Goud B, de la Salle H, Hanau D, Salamero J. Rab11A controls the biogenesis of Birbeck Granules by regulating Langerin recycling and stability. Mol Biol Cell 2007. 52. Waharte F, Spriet C, Heliot L. Setup and characterization of a multiphoton FLIM instrument for protein-protein interaction measurements in living cells. Cytometry A 2006;69:299–306. 53. Perez F, Pernet-Gallay K, Nizak C, Goodson HV, Kreis TE, Goud B. CLIPR-59, a new trans-Golgi/TGN cytoplasmic linker protein belonging to the CLIP-170 family. J Cell Biol 2002;156:631–642. 54. Raposo G, Kleijmeer M, Posthuma G, Slot J, Geuze HJ. Immunogold labeling of ultrathin cryosections: application in immunology. In: Herzenberg LA, Weir D, Herzenberg LA, Blackwell C, editors. Handbook of Exp Immunol. Cambridge, MA: Blackwell Science, Inc.; 1997, pp. 1–11. 1403