Ciliary targeting motif VxPx directs assembly of a trafficking module through Arf4
2009; Springer Nature; Volume: 28; Issue: 3 Linguagem: Inglês
10.1038/emboj.2008.267
ISSN1460-2075
AutoresJana Mazelova, Lisa Astuto-Gribble, Hiroki Inoue, Beatrice M. Tam, Eric Schonteich, Rytis Prekeris, Orson L. Moritz, Paul A. Randazzo, Dusanka Deretic,
Tópico(s)Cellular transport and secretion
ResumoArticle15 January 2009free access Ciliary targeting motif VxPx directs assembly of a trafficking module through Arf4 Jana Mazelova Jana Mazelova Division of Ophthalmology, Department of Surgery, University of New Mexico, Albuquerque, NM, USA Search for more papers by this author Lisa Astuto-Gribble Lisa Astuto-Gribble Division of Ophthalmology, Department of Surgery, University of New Mexico, Albuquerque, NM, USA Search for more papers by this author Hiroki Inoue Hiroki Inoue Laboratory of Cellular and Molecular Biology, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA Search for more papers by this author Beatrice M Tam Beatrice M Tam Department of Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, British Columbia, Canada Search for more papers by this author Eric Schonteich Eric Schonteich Department of Cell and Developmental Biology, University of Colorado Health Sciences Center, Aurora, CO, USA Search for more papers by this author Rytis Prekeris Rytis Prekeris Department of Cell and Developmental Biology, University of Colorado Health Sciences Center, Aurora, CO, USA Search for more papers by this author Orson L Moritz Orson L Moritz Department of Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, British Columbia, Canada Search for more papers by this author Paul A Randazzo Paul A Randazzo Laboratory of Cellular and Molecular Biology, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA Search for more papers by this author Dusanka Deretic Corresponding Author Dusanka Deretic Division of Ophthalmology, Department of Surgery, University of New Mexico, Albuquerque, NM, USA Department of Cell Biology and Physiology, University of New Mexico, Albuquerque, NM, USA Search for more papers by this author Jana Mazelova Jana Mazelova Division of Ophthalmology, Department of Surgery, University of New Mexico, Albuquerque, NM, USA Search for more papers by this author Lisa Astuto-Gribble Lisa Astuto-Gribble Division of Ophthalmology, Department of Surgery, University of New Mexico, Albuquerque, NM, USA Search for more papers by this author Hiroki Inoue Hiroki Inoue Laboratory of Cellular and Molecular Biology, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA Search for more papers by this author Beatrice M Tam Beatrice M Tam Department of Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, British Columbia, Canada Search for more papers by this author Eric Schonteich Eric Schonteich Department of Cell and Developmental Biology, University of Colorado Health Sciences Center, Aurora, CO, USA Search for more papers by this author Rytis Prekeris Rytis Prekeris Department of Cell and Developmental Biology, University of Colorado Health Sciences Center, Aurora, CO, USA Search for more papers by this author Orson L Moritz Orson L Moritz Department of Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, British Columbia, Canada Search for more papers by this author Paul A Randazzo Paul A Randazzo Laboratory of Cellular and Molecular Biology, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA Search for more papers by this author Dusanka Deretic Corresponding Author Dusanka Deretic Division of Ophthalmology, Department of Surgery, University of New Mexico, Albuquerque, NM, USA Department of Cell Biology and Physiology, University of New Mexico, Albuquerque, NM, USA Search for more papers by this author Author Information Jana Mazelova1, Lisa Astuto-Gribble1, Hiroki Inoue2, Beatrice M Tam3, Eric Schonteich4, Rytis Prekeris4, Orson L Moritz3, Paul A Randazzo2 and Dusanka Deretic 1,5 1Division of Ophthalmology, Department of Surgery, University of New Mexico, Albuquerque, NM, USA 2Laboratory of Cellular and Molecular Biology, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA 3Department of Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, British Columbia, Canada 4Department of Cell and Developmental Biology, University of Colorado Health Sciences Center, Aurora, CO, USA 5Department of Cell Biology and Physiology, University of New Mexico, Albuquerque, NM, USA *Corresponding author. Division of Ophthalmology, Department of Surgery, University of New Mexico School of Medicine, Basic Medical Sciences Building, Room 377, 915 Camino de Salud, NE, Albuquerque, NM 87131, USA. Tel.: +1 505 272 4968; Fax: +1 505 272 6029; E-mail: [email protected] The EMBO Journal (2009)28:183-192https://doi.org/10.1038/emboj.2008.267 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Dysfunctions of primary cilia and cilia-derived sensory organelles underlie a multitude of human disorders, including retinal degeneration, yet membrane targeting to the cilium remains poorly understood. Here, we show that the newly identified ciliary targeting VxPx motif present in rhodopsin binds the small GTPase Arf4 and regulates its association with the trans-Golgi network (TGN), which is the site of assembly and function of a ciliary targeting complex. This complex is comprised of two small GTPases, Arf4 and Rab11, the Rab11/Arf effector FIP3, and the Arf GTPase-activating protein ASAP1. ASAP1 mediates GTP hydrolysis on Arf4 and functions as an Arf4 effector that regulates budding of post-TGN carriers, along with FIP3 and Rab11. The Arf4 mutant I46D, impaired in ASAP1-mediated GTP hydrolysis, causes aberrant rhodopsin trafficking and cytoskeletal and morphological defects resulting in retinal degeneration in transgenic animals. As the VxPx motif is present in other ciliary membrane proteins, the Arf4-based targeting complex is most likely a part of conserved machinery involved in the selection and packaging of the cargo destined for delivery to the cilium. Introduction Primary cilia are specialized organelles in vertebrate cells, which house arrays of signal-transduction modules that sense a wide variety of signals. Despite ciliary involvement in a broad range of human diseases (ciliopathies), including retinal degeneration, polycystic kidney disease, Bardet–Biedl syndrome and neural tube defects (Rosenbaum and Witman, 2002; Singla and Reiter, 2006; Christensen et al, 2007; Fliegauf et al, 2007), the molecular basis for selection and packaging of sensory receptors and membrane cargo targeted to the primary cilia and cilia-derived sensory organelles remains elusive. Nonetheless, the directed delivery of ciliary cargo most likely involves vesicular transport regulated by the small GTPases of the Rab and Arf families, which have a central function in coordinating intracellular membrane trafficking (Gillingham and Munro, 2007; Leroux, 2007). Upon GTP binding, Rabs recruit a multitude of effectors that organize membrane domains involved in the tethering of membranes to other membranes and to cytoskeletal elements, thus conferring directionality to membrane traffic (Zerial and McBride, 2001; Grosshans et al, 2006; Markgraf et al, 2007). Rab8 (Wandinger-Ness and Deretic, 2008) has recently emerged as a central regulator of the biogenesis of primary cilia (Nachury et al, 2007; Yoshimura et al, 2007; Omori et al, 2008). In retinal rod photoreceptors, Rab8 regulates the biogenesis of cilia-derived sensory organelles, the rod outer segments (ROSs), by regulating trafficking of post-TGN (trans-Golgi network) rhodopsin transport carriers (RTCs) (Deretic et al, 1995; Moritz et al, 2001). RTCs travel vectorially through the cell body, the rod inner segment (RIS) (Figure 1A), to the base of the cilium and fuse with the specialized domain separating the ciliary membrane from the surrounding RIS plasma membrane (Deretic and Papermaster, 1991), thereby regulating the replenishment of light-sensitive ROS membranes, the major component of which is rhodopsin. Figure 1.Rhodopsin VxPx motif engages Arf4 at the TGN to regulate RTC budding. (A) Diagram of the photoreceptor cell. RTCs travel from the Golgi/TGN (arrow) to the base of the cilium (C). BB, basal body; M, mitochondria; N, nucleus; Sy, synapse; AJ, adherens junction. (B) adRP mutations in the rhodopsin VxPx targeting motif are indicated with asterisks. (C) Frog retinas were pulse-labelled for 60 min and retinal PNS was incubated for 30 min with 50 μM peptides, as indicated in the panel, prior to a 2 h cell-free chase. Radiolabelled membrane proteins from two retinas were fractionated into Golgi/TGN and RTCs and analysed by SDS–PAGE and autoradiography ([35S]Rh). (D) Immunoblots of a duplicate gel were probed successively with anti-Arf3 and anti-Arf4, which specifically recognizes an ∼20-kDa protein (left panel) and quantified (lower right panels). (E) Golgi/TGN and RTCs were separated on sucrose density gradients, and distribution of rhodopsin ([35S]Rh) was visualized as in (C). Duplicate gels were blotted and probed with antibodies, as indicated in the panel (partial loss of the Rab11 band in RTC fractions is due to a cut to allow Arf3 detection on the same immunoblot). The differential distribution of calnexin, galactosyltransferase (Gal T) and sialyltransferase (Sial T), markers for the ER, trans-Golgi and TGN, respectively, in membranes from identically run sucrose gradients, as reported (Deretic and Papermaster, 1991, 1993; Deretic et al, 2004). (F) Quantification of autoradiograms and immunoblots from (E). Download figure Download PowerPoint The continuous renewal of ROS membranes, which are removed through daily shedding and phagocytosis by retinal pigment epithelial cells, necessitates the hypertrophy of the Rab8/RTC-mediated pathway and renders photoreceptors vulnerable to mutations that affect membrane trafficking (Deretic, 2006). Among the mutations with the most severe phenotypes are those that alter the rhodopsin C-terminal VxPx targeting motif (Deretic et al, 1998), which causes rapid retinal degeneration and blindness in a group of diseases known as autosomal dominant retinitis pigmentosa (adRP) (Berson et al, 2002). Within the targeting motif, V345 and P347 are the primary sites of C-terminal adRP mutations involving single amino-acid substitutions (http://www.retina-international.com/sci-news/rhomut.htm). Abundant evidence points to the role of the VxPx motif in rhodopsin sorting into RTCs at the TGN, and delivery to the cilium and the ROS (Deretic et al, 1998; Green et al, 2000; Tam et al, 2000; Shi et al, 2004). Similar to rhodopsin, polycystin-2 (PC2), a protein affected in autosomal dominant polycystic kidney disease, and the cyclic nucleotide-gated (CNG) channel CNGB1b subunit in olfactory neurons, also possess a functional ciliary targeting VxPx motif (Geng et al, 2006; Jenkins et al, 2006). The VxPx targeting motif of rhodopsin binds directly and specifically to Arf4, suggesting a distinct role for this Arf in the generation of RTCs (Deretic et al, 2005). The Arf family members generally regulate membrane trafficking, lipid metabolism, organelle morphology and cytoskeleton dynamics (Nie et al, 2003; Volpicelli-Daley et al, 2005; D'Souza-Schorey and Chavrier, 2006; Kahn et al, 2006); however, the function of Arfs 4 and 5 remains poorly understood. The class II Golgi Arfs, Arf4 and Arf5, bind Arfophilins/FIPs, the effectors of the small GTPase Rab11 (Shin et al, 1999; Meyers and Prekeris, 2002). Arfophilin 1/FIP3 coordinates Arf- and Rab11-regulated trafficking pathways (Hickson et al, 2003; Fielding et al, 2005; Eathiraj et al, 2006). Rab11 has been localized to the photoreceptor TGN and the RTCs (Deretic et al, 1996), and has been implicated in the regulation of plasma membrane trafficking through the recycling endosomes (Prekeris et al, 2000; Pasqualato et al, 2004) or through the TGN (Urbe et al, 1993; Chen et al, 1998; de Graaf et al, 2004). Cargo recruitment through Arf4 most likely depends on GTP hydrolysis mediated by Arf GTPase-activating proteins (GAPs) (Inoue and Randazzo, 2007). Arf GAPs are essential for coupling the proof-reading of cargo incorporation to carrier budding, and are often incorporated into protein coats (Lee et al, 2004). ASAPs are Arf GAPs containing ankyrin repeats, pleckstrin-homology (PH), proline-rich and SH3 domains that belong to a family of multifunctional scaffold proteins regulating membrane trafficking and actin remodelling (Randazzo and Hirsch, 2004; Nie and Randazzo, 2006). The BAR (Bin/amphiphysin/Rvs) domain (Peter et al, 2004), at the N terminus of ASAP1, mediates homodimerization and, in concert with the GAP domain, functions as a coincidence detector sensing the membrane phospholipid composition and the presence of GTP–Arf and responding by increasing membrane curvature (Nie et al, 2006). Thus, ASAP1 functions both as an Arf GAP and an Arf effector that may regulate carrier budding (Luo et al, 2007). In the present study, we identify components of the molecular machinery that assembles at the TGN upon binding of Arf4 to the conserved rhodopsin C-terminal VxPx motif and examine the possibility that this machinery regulates ciliary targeting, a specialized form of the TGN-to-plasma membrane trafficking. We report that Arf4, ASAP1, FIP3 and Rab11 interact to form a complex that regulates budding of post-TGN carriers targeted to the primary cilia. Results The association of Arf4 with TGN membranes is dependent on its access to the VxPx targeting motif of rhodopsin To gain further insight into interactions of the rhodopsin VxPx targeting motif (Figure 1B) with Arf4 at the TGN, we employed our established cell-free assay, which reconstitutes RTC budding in the retinal post-nuclear supernatant (PNS), enriched in biosynthetic membranes that mediate rhodopsin trafficking, in an ATP-, GTP-, cytosol-, Rab- and Arf4-dependent manner (Deretic et al, 1996, 1998, 2005). Consistent with our published reports, the peptide corresponding to the α helix 3 unique to Arf4, and the rhodopsin C-terminal peptide inhibited RTC budding in vitro, whereas the peptide bearing the adRP mutation P347S in the VxPx motif had no effect (Figure 1C). Remarkably, both inhibitory peptides nearly completely displaced Arf4 from the Golgi/TGN membranes (Figure 1D), whereas the P347S peptide had no effect on Arf4 membrane association. The peptide effect was specific for Arf4, as Arf3 was unaffected (Figure 1D). Thus, association of Arf4 with TGN membranes is dependent on its access to the VxPx sorting motif of rhodopsin. Arf3 and Arf4 are differentially distributed among membrane fractions separated on sucrose density gradients (Figure 1E and F). Arf4 co-fractionated with Rab11 and the sialyltransferase-enriched TGN membranes, but was absent from RTCs, which contained Rab11, as reported (Deretic et al, 1996). The absence of Arf4 from RTCs suggests that it most likely hydrolyses GTP and dissociates from the membrane during RTC budding from the TGN. ASAP1 functions as a GAP for Arf4 and regulates RTC budding To identify the candidate GAP for Arf4, we examined the retinal expression of Arf GAPs ARAP1 and 2, and ASAP1 and 2. We determined that ASAP1 was abundant in photoreceptor Golgi/TGN and cytosolic fractions, and was also present on RTCs (Figure 2A). ASAPs display phosphatidylinositol 4,5-bisphosphate (PIP2)-dependent activity on Arf1 and Arf5 (Brown et al, 1998), but the activity of ASAP1 towards Arf4 has not been tested. To test it, we employed ASAP1 BAR-PZA, a recombinant protein containing the BAR, PH, GAP and ankyrin repeat domain (Figure 2B), which stimulates membrane tubulation and regulates trafficking in a BAR- and GAP-dependent manner (Nie et al, 2006). We determined that ASAP1 BAR-PZA has a comparable GAP activity towards Arf4, Arf1 and Arf5 (Figure 2C), but not Arf6, as reported (Brown et al, 1998). Therefore, ASAP1 is a good candidate for an Arf4 GAP. Figure 2.ASAP1 is a GAP for Arf4 and regulates RTC budding from the TGN. (A) Membrane proteins from one retina were separated by SDS–PAGE and immunoblotted with anti-ASAP1, which recognized an ∼120-kDa protein. Upon fractionation (lower panel), ASAP1 was detected in the G/TGN and the cytosol, with a small portion associated with RTCs. (B) Schematic representation of the ASAP1 BAR-PZA construct. (C) ASAP1 BAR-PZA was incubated with recombinant Arfs 1, 4, 5 or 6, and the hydrolysis of [α32P]GTP bound to Arf was measured. The data are presented as the means±s.d. of three separate experiments. (D) Radiolabelled retinal PNS was separated into membrane and cytosolic fractions, which were reconstituted as indicated in the panel. ASAP1-BAR-PZA (1 μM) was added to the assay in limited (25%) cytosol during the cell-free chase, and incorporation of rhodopsin into RTCs was determined ([35S]Rh). The quantification data are presented as the means±s.e. of three separate experiments (*P=0.01). (E–G) Negatively stained in vitro budded RTCs (∼250 nm) examined by EM. (H) Following in vitro budding, Golgi and RTCs were immunoblotted with anti-GM130. (I) Cytosol was immunodepleted with anti-ASAP1 and protein A beads, reducing ASAP1 to ∼10% of control. (J) After in vitro budding in ASAP1-immunodepleted cytosol, membrane proteins from two retinas were separated by SDS–PAGE and autoradiographed ([35S]rhodopsin, left panels). Duplicate gels were immunoblotted with ASAP1 (right panels). Download figure Download PowerPoint To examine the function of ASAP1 in rhodopsin trafficking, we tested the ability of ASAP1 BAR-PZA to stimulate RTC budding in our established cell-free assay. The full cytosol present in the PNS supports maximum RTC budding in vitro, equivalent to the maximum RTC budding in vivo in the same time period (Deretic et al, 1996). To test the activity of ASAP1 BAR-PZA, we reduced the amount of cytosol present in the assay. We separated membranes from the cytosol, followed by the add-back of the limited (25%) or the full cytosol complement. Under limited cytosol conditions, ASAP1 BAR-PZA stimulated RTC budding by ∼50%, almost to the level observed with full cytosol (Figure 2D). EM analysis of the budded carriers (Figure 2E–G) showed that RTCs formed with ASAP1 BAR-PZA were identical to the ones formed in full cytosol, and that these structures were undetectable when the cytosol was omitted from the budding assay. Moreover, similar to the control post-TGN carriers, RTCs budded in the presence of ASAP1 BAR-PZA were devoid of the Golgi marker GM130 (Figure 2H), indicating that the observed increase in budding was not due to nonspecific Golgi fragmentation. We next prepared ASAP1-immunodepleted cytosol (Figure 2I), and assayed its activity in RTC budding (Figure 2J). RTC budding was diminished (<50%), but was not completely inhibited due to the remaining cytosolic and TGN-associated ASAP1 (Figure 2I and J). Because of high retinal tissue requirements, the effect on RTC budding was established using a single concentration of ASAP1-immunodepleted cytosol. Importantly, ASAP1 was detected on budded RTCs in quantities proportional to the amount of radiolabelled rhodopsin (Figure 2J). Taken together, these data indicate that ASAP1 regulates budding and is incorporated into budded RTCs. ASAP1 associates with Golgi/TGN-derived buds and with microfilaments As ASAP1 regulates both membrane trafficking and actin remodelling, we examined its association with photoreceptor biosynthetic membranes and with the actin cytoskeleton by confocal microscopy (Figure 3; Supplementary Figure S1). Photoreceptor actin filaments, which are anchored at the adherens junctions (AJs), encircle the RIS (Figure 3A) and extend to the calycal processes (CPs), the villous structures that evaginate from the RIS and surround the ROS. ASAP1 was detected in the structures resembling actin filaments (arrowheads, Figure 3B; Supplementary Figure S1A), in the cytoplasm, as well as in punctate structures within the RIS area occupied by the Golgi (G, arrow, Figure 3B; Supplementary Figure S1A). ASAP1 localized along phalloidin-labelled microfilaments (arrowheads, Figure 3C; Supplementary Figure S1B and C), from the AJs to the CPs. We also examined ASAP1 association with the Golgi/TGN in further detail. ASAP1 localized to the punctate structures tightly juxtaposed to the trans-Golgi labelled by Rab6 (arrows, Supplementary Figure S1D, E and H), and to RTCs (arrowhead, Supplementary Figure S1D). In retinas treated with latrunculin or nocodazole to disrupt photoreceptor cytoskeleton, ASAP1-positive puncta remained associated with the Golgi (arrows, Supplementary Figure S1F and G). However, upon treatment with brefeldin A (BFA), which we showed leads to Golgi disruption and the inhibition of rhodopsin trafficking (Deretic and Papermaster, 1991), ASAP1 was distributed diffusely in the cytosol (Supplementary Figure S1K–M). Conversely, treatment with propranolol (Deretic et al, 2004) to enlarge the TGN and RTCs (Supplementary Figure S1O) resulted in uniform distribution of ASAP1 in these enlarged membrane structures (arrows, Supplementary Figure S1P and Q). Taken together, the Arf-dependent localization at the photoreceptor TGN and its presence on RTCs are consistent with a role for ASAP1 as a regulator of Arf4 in RTC budding. Figure 3.ASAP1 is associated with both the Golgi/TGN and actin cytoskeleton, and colocalizes with Rab11 and FIP3 on nascent RTCs at the TGN. (A) A diagram representing the microfilaments (red) and calycal processes (CPs) that evaginate from the RIS and surround the base of the ROS (modified from Deretic et al, 1995). (B) A confocal optical section (0.7 μm) of frog retina. ASAP1 (green) is detected in the RIS cytoplasm, in punctate structures around the Golgi (G, arrow), and in axially aligned structures (arrowheads), resembling microfilaments. N, nucleus (TO-PRO-3, blue). (C) ASAP1 (green) and phalloidin (red), colocalize (yellow, arrowheads) along microfilaments, including the CPs (arrow). Microfilaments are anchored at the adherens junctions (AJs). (D) Rhodopsin C-terminal mAb 11D5 (red) labels the ROS and the Golgi (G), where ASAP1-positive puncta (yellow, arrows) line up with regular periodicity. ASAP1 is also detected in CPs. (E) Rab11-positive puncta (yellow, arrows) aligned with rhodopsin-laden Golgi. Rab11 is also present on RTCs (arrowheads). (F) ASAP1 (blue) and Rab11 (red) colocalize in the bud-like profiles at the tips of the trans-Golgi (Rab6, green) (boxed area magnified in G). (G) Magnified trans-Golgi area from (F), with ASAP1- and Rab11-positive buds (arrows). (H) Rab11 (red) and ASAP1 (blue) are shown separately. (I) FIP3 (green, rabbit Ab; Wilson et al, 2005), overlaps with Rab11 (red) and ASAP1 (blue) in the punctate structures within the Golgi region (boxed area magnified in J). (J) FIP3, Rab11 and ASAP1 are shown in separate channels to better visualize their colocalization (arrows). (K) An EM image showing a greatly enlarged nascent bud at the TGN (arrowhead) and enlarged RTCs (arrows) in propranolol-treated retina (modified from Deretic et al, 2004). (L) Rab11 (red) and ASAP1 (green) colocalize (arrows, yellow) in propranolol-enlarged RTCs and buds lining Rab6-postive trans-Golgi (blue). (M, N) Rab11 and ASAP1 shown individually. Insets demonstrate identical distribution of Rab11 and ASAP1. (O) Goat anti-FIP3 detects an ∼80-kDa protein on isolated RTCs. (P) Following in vitro budding and subcellular fractionation, the distribution of rhodopsin was determined by autoradiography ([35S]Rh). Duplicate gels were immunoblotted as indicated in the panel. (Q) Quantification of autoradiograms and immunoblots. Bar, 4 μm in (B, C), 3 μm in (D–F, I), 2 μm in (L–N), 1 μm in (H, J), 0.7 μm in (G, K) and insets in (L–N). Download figure Download PowerPoint TGN-derived nascent RTCs contain ASAP1, Rab11 and FIP3 By confocal microscopy, ASAP1 and rhodopsin colocalized at the Golgi/TGN in photoreceptor cells (Figure 3D, arrows). Surprisingly, Rab11 was similarly detected in Golgi-associated puncta, as well as in RTCs (Figure 3E). This prompted us to examine whether ASAP1 and Rab11 reside in the same TGN-associated bud-like structures. We first performed double labelling of Rab6 and Rab11 (Supplementary Figure S2A), which revealed Rab6 in the trans-Golgi cisternae and Rab11 in punctate structures at the TGN in control retinas (arrows, Supplementary Figure S2B–G) and in propranolol-treated retinas (Supplementary Figure S2H and I). Next, we examined the colocalization of Rab11 and ASAP1 by triple labelling with Rab6 (Figure 3F–H). Buds containing both Rab11 and ASAP1 localized to the tips of the Rab6-positive trans-Golgi cisternae (arrows, Figure 3G and H; Supplementary Figure S3). Propranolol treatment produced greatly enlarged buds and RTCs (Figure 3K), in which Rab11 and ASAP1 completely overlapped (arrows, Figure 3L–N). The localization of Rab11 at the sites of Arf4 function led us to probe these structures for the presence of FIP3, a member of the family of Arf- and Rab11-interacting proteins (Hales et al, 2001; Junutula et al, 2004). By confocal microscopy, FIP3 was present on the ASAP1/Rab11-positive buds (Figure 3I and J). By immunoblotting, FIP3 was detected on purified RTCs (Figure 3O). On sucrose density gradients, Rab11, ASAP1 and FIP3 co-fractionated with the Golgi/TGN-enriched fractions and with RTCs (Figure 3P and Q), which segregated into two major fractions. Although higher buoyant density RTCs contained all three proteins, ASAP1 and FIP3 were nearly absent from the lower density RTCs (Figure 3P and Q). Because the fusion regulator Rab8 predominantly associates with the low-density membrane carriers (Deretic et al, 1995), we surmise that the higher density RTCs correspond to the nascent post-TGN carriers, which mature, possibly by loosing protein coats, and give rise to the low-density Rab8-positive periciliary RTCs. Arf4, ASAP1, Rab11 and FIP3 form a complex Colocalization and co-fractionation of ASAP1, Rab11, FIP3 and Arf4 at the TGN prompted us to ask whether they form a complex at the RTC budding sites. In co-immunoprecipitation experiments we determined that anti-Arf4 immunoprecipitated Rab11 and ASAP1 from retinal PNS (Figure 4A). Moreover, Rab11 and ASAP1 were mutually co-immunoprecipitated, and anti-FIP3 (Wilson et al, 2005) precipitated ASAP1, along with Rab11, a known FIP3-binding protein (Figure 4A). A control Golgi-associated protein G20 was not co-immunoprecipitated (Figure 4A). These data suggest that Arf4, ASAP1, Rab11 and FIP3 not only colocalize at the TGN but also interact to form at least one complex. Figure 4.Arf4, Rab11, FIP3 and ASAP1 form a complex and regulate RTC budding. (A) Specific antibodies were incubated with retinal PNS and proteins bound to protein-A beads were separated by SDS–PAGE and analysed by immunoblotting, as indicated. Left panels, pAb IPs; right panels, mAb IPs. (B) Arf4 and Arf6 were expressed as GST fusion proteins and incubated with 6His–FIP3. Proteins bound to glutathione beads were separated by SDS–PAGE and analysed by immunoblotting with anti-FIP3. (C) FIP3-F1 C-terminal fragment (aa 441–756) (1 μM) was added to the assay in limited (25%) cytosol during the cell-free chase, and rhodopsin incorporation into RTCs was determined ([35S]rhodopsin). Autoradiograms and immunoblots of duplicate gels probed for Rab11 and ASAP1 were quantified (right panel). (D) Cytosol was immunodepleted with anti-Rab11, reducing it to nearly undetectable levels. (E) Following in vitro budding in Rab11-immunodepleted cytosol, membrane proteins from two retinas were separated by SDS–PAGE and autoradiographed ([35S]rhodopsin, left panels). Rab11 immunoblots of duplicate gels (right panels). Download figure Download PowerPoint To determine whether Arf4 and Rab11 associate through their interacting protein FIP3, we tested the ability of Arf4 to bind FIP3 directly. FIP3 was shown to have significantly higher affinity for Rab11 and Arf6 than for Arf5, a class II Arf (Fielding et al, 2005). Our pull-down assay confirmed strong binding of recombinant FIP3 to GST–Arf6, and showed a direct, albeit weak binding of FIP3 to GST–Arf4. Retinal cytosol greatly enhanced FIP3 binding to GST–Arf4 (Figure 4B), but did not affect its binding to GST–Arf6 (not shown). ASAP1- or Rab11-immunodepleted cytosol did not facilitate the binding of GST–Arf4 to FIP3 (Figure 4B). These data suggest that Rab11 and ASAP1 stabilize the Arf4–FIP3 interaction and that these proteins associate to form a single complex, most likely at the TGN. A FIP3 fragment containing Arf and Rab11 interaction domains, along with Rab11, stimulates RTC budding To further define the roles of the individual components of the complex, we tested the effect on RTC budding of the FIP3-F1 C-terminal fragment, encompassing the Arf- and Rab11-binding sites (Schonteich et al, 2007). In low cytosol, the addition of the FIP3-F1 fragment nearly doubled in vitro RTC budding measured by the radiolabelled rhodopsin content (Figure 4C). The budded RTCs displayed a proportional increase in membrane-associated Rab11 and ASAP1 (Figure 4C). As Rab11 is predominantly membrane-associated in photoreceptors (Supplementary Figure S2J and K), a greater fraction of total Rab11 was bound to RTCs than the fraction of total ASAP1, which exists in equilibrium between the membranes and the cytosol (see Figure 2A). Budded carriers were devoid of GM130 (not shown), suggesting that sorting was not compromised in the presence of the FIP3-F1 fragment. Surprisingly, we found that the cytosolic pool of Rab11 increases when RTC budding is inhibited (Supplementary Figure S2J and K). We thus tested the capacity of Rab11-immunodepleted cytosol (Figure 4D) to support RTC budding. In a repr
Referência(s)