Skip to main content
PMC home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2005 Feb 22;102(9):3301–3306. doi: 10.1073/pnas.0500095102

Rhodopsin C terminus, the site of mutations causing retinal disease, regulates trafficking by binding to ADP-ribosylation factor 4 (ARF4)

Dusanka Deretic *,†,‡, Andrew H Williams †, Nancy Ransom *, Valerie Morel §,, Paul A Hargrave ∥, Anatol Arendt ∥
PMCID: PMC552909  PMID: 15728366

Abstract

The maintenance of photoreceptor cell polarity is compromised by the rhodopsin mutations causing the human disease autosomal dominant retinitis pigmentosa. The severe form mutations occur in the C-terminal sorting signal of rhodopsin, VXPX-COOH. Here, we report that this sorting motif binds specifically to the small GTPase ARF4, a member of the ARF family of membrane budding and protein sorting regulators. The effects of blocking ARF4 action were functionally equivalent to the effects of blocking the rhodopsin C-terminal sorting signal. ARF4 was essential for the generation of post-Golgi carriers targeted to the rod outer segments of retinal photoreceptors. Thus, the severe retinitis pigmentosa alleles that affect the rhodopsin sorting signal interfere with interactions between ARF4 and rhodopsin, leading to aberrant trafficking and initiation of retinal degeneration.

Keywords: ADP-ribosylation factor GTPase, membrane trafficking, retinitis pigmetosa, rhodopsin

Rhodopsin is the transmembrane protein that, upon photoexcitation, initiates the visual transduction cascade taking place in photoreceptor rod outer segments (ROS, Fig. 1A). Mutations found in the rhodopsin gene represent the most common known cause of the human disease retinitis pigmentosa (RP). Over 100 different mutations in rhodopsin are known to cause the autosomal dominant form of the disease [autosomal dominant RP (adRP)] (RetNet, www.sph.uth.tmc.edu/RetNet/disease.htm). Particularly severe adRP is caused by point mutations clustered within the highly conserved five C-terminal amino acids and by the mutation Q344Ter, which removes these five amino acids (1, 2). The integrity of the rhodopsin C terminus and its accessibility for molecular interactions appear crucial to rod photoreceptor survival. This is evidenced by one of the most severe adRP phenotypes ever observed in a family with a rhodopsin mutation, Ter349Glu, which is predicted to add an additional segment of 51 amino acids to the C terminus of rhodopsin, making it dysfunctional while containing all of the normal 348 amino acid residues (3).

Fig. 1.

Fig. 1.

Open in a new tab

The rhodopsin C-terminal peptide and the GST fusion protein containing the C terminus of rhodopsin specifically bind a 20-kDa protein, a member of the ARF family of small GTPases. (A) RTCs move vectorially (dashed arrow) from the Golgi (G) and the TGN, through the ellipsoid region filled with mitochondria (M), to the connecting cilium (C), where they fuse with the plasma membrane of the RIS. Newly synthesized rhodopsin is then delivered to the ROS. N, nucleus; Sy, synapse. (B) Sequence of the frog rhodopsin C-terminal peptide (peptide C) folded as predicted from the crystal structure of rhodopsin (29). The truncated peptide (peptide CΔ5) models the adRP mutation Q344ter. The UV-activatable crosslinker Br-DAP was attached to the N-terminally biotinylated peptide C, and CΔ5, as indicated. (C) Retinal PNS was incubated with the biotinylated Br-DAP peptides. After UV illumination, proteins crosslinked to the peptides were either (i) separated by SDS/PAGE, blotted and probed with SA-HRP followed by the chemiluminescent immunodetection system (Left)or(ii) bound to SA-Sepharose and eluted (SA-Seph. B.), detected as above (Right). A 20-kDa protein (*) specifically crosslinked to peptide C. The band detected in the sample with peptide C and CΔ5 after SA-Sepharose binding (♦) is SA in complex with biotin (biotin-SA; see below). (D) Sequence analysis by microcapillary reverse-phase HPLC nanoelectrospray tandem MS of the 20-kDa band (*), excised from the silver-stained gel of the sample crosslinked to peptide C, bound, and eluted from SA-Sepharose, showed the presence of ARF4 (peptide 1) and potentially another GTPase that belongs to the class I ARFs (peptide 2). Peptide 3 is highly conserved among all ARFs. Sequence analysis of the band present in samples with peptide C and CΔ5(♦) revealed three tryptic fragments originating from SA in complex with biotin (biotin-SA), which was likely released from the SA-Sepharose column during elution. (E) GST fusion proteins were separated by SDS/PAGE and stained with SilverStain (Left). The GST fusion protein containing the rhodopsin C-terminal sequence (C) is a 30-kDa protein, the truncated protein lacking the regulatory sequence (CΔ5) migrates at 29 kDa and GST alone (-) at 28 kDa. The 26-kDa band present in all lanes is a breakdown product of GST. Immunoblotting of GST-fusion proteins with anti-rhodopsin C-terminal mAb 11D5 (Center) reveals that the fusion protein C contains the mAb 11D5 antigenic site, whereas CΔ5 does not. In GST pull-down assays (Right), GST-C specifically bound a 20-kDa protein (*). The higher-Mr bands are GST-fusion proteins and their breakdown products, as in Left.

The mutations in the C-terminal region of rhodopsin that cause adRP interfere with rhodopsin trafficking both in in vitro assays and in animal models (4–9). Using a retinal cell-free system, we have established that the rhodopsin C-terminal amino acid sequence QVS(A)PA represents a sorting motif regulating the budding of rhodopsin transport carriers (RTCs) from the trans-Golgi network (TGN) (4). Rhodopsin is crucial for the morphogenesis of the light-sensitive organelle ROS (10). Expression of normal rhodopsin restores retinal morphology in rhodopsin knockout mice lacking ROS, but expression of C-terminally truncated rhodopsin does not (11, 12). However, transgenic expression of the cone color pigment S-opsin can restore ROS generation in rhodopsin knockout mice (13). Because the sequence of the C terminus of S-opsin is KVGPH (14), these findings suggest that the essential targeting sequence, VXPX-COOH, also governs the onset of ROS morphogenesis (15).

Small GTPases, so called because of their monomeric form and their low molecular mass, 20–25 kDa, are important regulators of intracellular trafficking. ARFs, or ADP-ribosylation factors, are ubiquitous regulators of membrane traffic in organisms as diverse as yeast and mammals and include six family members [class I, ARFs 1–3; class II, ARFs 4 and 5; and class III, ARF6] (16). ARF GTPases regulate recruitment of coat proteins that facilitate formation of transport carriers, alter membrane lipid composition through the activation of phospholipase D (PLD), and regulate the assembly of the actin cytoskeleton (17–19). Activation of PLD by class I ARFs leads to the formation of microdomains enriched in phosphatidylinositol-4,5-bisphosphate [PI(4,5)P2] and phosphatidic acid in the Golgi and the TGN (20). These lipid microdomains recruit regulatory proteins and cause local rearrangement of the actin-based Golgi membrane skeleton, which is believed to facilitate post-Golgi membrane budding (21). ARF guanine nucleotide exchange factors that regulate post-Golgi trafficking are sensitive to the fungal toxin Brefeldin A (22), which profoundly affects photoreceptor morphology and inhibits rhodopsin trafficking (23). Virtually nothing is known about the function of the class II ARFs, ARF4 and -5 (24, 25), but their interactions with the effectors of other small GTPases (17, 19) suggest their potential role in coordinating membrane traffic in conjunction with additional small G proteins.

In this study, we demonstrate a role for ARF4 in the initial steps of the post-Golgi trafficking of rhodopsin. We report that ARF4 specifically recognizes the C-terminal sorting signal of rhodopsin and regulates its incorporation into RTCs, the specialized membrane carriers targeted to the ROS.

Materials and Methods

Southern leopard frogs, Rana berlandieri, were used because of their highly hypertrophied rhodopsin biosynthesis, which offers a unique advantage for the study of the basic mechanisms underlying photoreceptor membrane biogenesis. The C-terminal sequence of rhodopsin is highly conserved among different species, and the amino acids that are mutated in adRP are conserved in the frog sequence (26).

Pulse–Chase Labeling and Retinal Subcellular Fractionation. Frogs were dark-adapted for 2 h before the experiment. Isolated retinas were incubated for 1 h at 22°C in oxygenated medium with [35S]-Express protein labeling mixture (25 μCi/retina; 1 Ci = 37 GBq). Retinal fractionation and preparation of postnuclear supernatant (PNS) enriched in photoreceptor biosynthetic membranes were performed as described (23).

Cell-Free Budding of RTCs in Photoreceptor-Enriched PNS. The standard assay for cell-free budding of RTCs was as described (4, 27). Radiolabeled PNS (1 ml) was preincubated with antibodies (0.5 mM) or synthetic peptides (50 μM, dissolved in 10 mM Hepes·KOH, pH 7.0) for 30 min on ice. The assay was initiated by the addition of 50 μl of an ATP regenerating system and by transfer to 22°C. The samples containing N-bromoacetyl-N-(3-diazopyruvoyl)-1,3-phenylenediamine (Br-DAP) peptides were illuminated for 10 min at 366 nm. In some experiments, 1 mM sulfosuccinimidyl-3-[(4-azidophenyl) dithio] propionate, a cleavable photoactivatable heterobifunctional membrane-impermeable crosslinker, was added to radiolabeled retinal PNS for 20 min before crosslinking was initiated by illumination with a short-wavelength UV lamp (265–275 nm, 2 min on ice).

Generation of the GST Fusion Proteins GST-C and GST-CΔ5 and the GST Pull-Down Assay. Oligonucleotides corresponding to the frog rhodopsin C-terminal sequence AA 327–354 (C) or AA 327–349 (CΔ5) (26) were inserted in the EcoRI-HindIII sites of the pGSTtag polylinker. Fusion proteins were expressed in the Escherichia coli HA63 protease-deficient strain, because GST-C was degraded in strains DH-5-α and BL21-RIL. Expression of fusion proteins was induced by the addition of 0.1 mM isopropyl-β-d-thiogalactoside, and GST-C and GST-CΔ5 were purified from cell lysates by using glutathione–Sepharose beads (Amersham Pharmacia Biotech). GST fusion proteins were incubated either with frog retinal PNS during the cell-free chase or with the cytosol or membrane fraction. After the chase, membranes were solubilized with 1% Triton X-100 100 for 1 h at 4°C. Insoluble aggregates were sedimented, and solubilized proteins were incubated with glutathione–Sepharose beads overnight at 4°C. After extensive washing, beads were solubilized in SDS sample buffer, and bound proteins were separated by SDS/PAGE.

Synthesis of Peptides C and CΔ5 Containing a Photoactivatable Crosslinker Br-DAP. The frog rhodopsin C-terminal peptide (AA 330–354, C) and a truncated peptide (AA 330–349, CΔ5) were synthesized with an automatic peptide synthesizer (Applied Biosystems 431A) by using Fmoc-Ala-Wang resin and Fmoc amino acids. Peptides on resin with a free N-terminal amino group were biotinylated by using d-biotin-N-hydroxysuccinimide ester overnight. Resin with peptide was washed, peptide-cleaved, deprotected, and purified by reverse-phase preparative HPLC.

Direct acylation of Lys-340 using the photoactivatable crosslinker Br-DAP was unsuccessful. A new peptide with Ala-336 replaced with Cys was synthesized and then used for Br-DAP attachment. Selective reaction of the cysteine SH group with Br-DAP was carried out in acetate buffer, pH 8.0, in the dark, according to Taylor et al. (28). Ala-336 was chosen for its likely proximity to the QVSPA motif, given the crystal structure of bovine rhodopsin and the structural analysis of the bovine rhodopsin C-terminal peptide (29, 30). The free Lys-340 also increased the solubility of the peptide. Peptides were purified by preparative HPLC and were demonstrated to have the correct mass by MS.

Purification of the Proteins Crosslinked to the Biotinylated Br-DAP Peptides with Streptavidin (SA)–Sepharose Beads. SA-Sepharose beads (Amersham Pharmacia Biotech) were spun for 1 min in an Eppendorf microcentrifuge, washed three times with TBS (20 mM Tris, pH 7.5/500 mM NaCl), and incubated with retinal cytosolic proteins crosslinked to the biotinylated Br-DAP peptides for 2 h at 4°C. Unbound proteins were collected and beads washed four times with TBS. Bound proteins were eluted with 2 mM biotin in TBS. In some experiments, bound proteins were eluted by boiling the beads in SDS sample buffer. In both preparations, the 20-kDa protein was solely eluted from the sample crosslinked to peptide C.

Protein Identification by Tandem MS. The 20-kDa protein specifically crosslinked to peptide C and bound to SA-Sepharose was excised from a silver-stained gel and subjected to sequence analysis. Protein identification was performed at the Harvard Microchemistry Facility (Boston) by microcapillary reverse-phase HPLC nanoelectrospray tandem MS, as described (31). In addition to a tryptic fragment of frog rhodopsin C-terminal peptide (26), sequence analyses revealed one peptide unique to Xenopus ARF4 (A. L. Boman, National Center for Biotechnology Information, direct submission), accession no. AAA74951 (peptide 1 in Fig. 1D), and two peptides consistent with human ARF1 (32), accession no. 1065361 (peptides 2 and 3). However, peptide 2 is conserved among all class I ARFs, and peptide 3 is conserved among all ARFs. The 18-kDa protein showed the tryptic fragment of the rhodopsin C-terminal peptide and three tryptic peptides originating from SA in complex with biotin, released from the SA-Sepharose column during elution.

Preparation of Polyclonal Anti-ARF4 Antibody. The ARF4 peptide IQEAAEELQKMLQ was synthesized and conjugated to keyhole limpet hemocyanin by SynPep (Dublin, CA). New Zealand white rabbits were immunized with the conjugated peptide, and anti-ARF4 antibody was prepared by Covance Research Products (Denver, PA). Anti-ARF4 serum was purified on the ARF4 peptide affinity column.

SDS/PAGE and Immunoblotting. Radiolabeled proteins were analyzed by SDS/PAGE, with or without 2-mercaptoethanol (βME). Gels were stained by Phast Gel Blue R (Amersham Pharmacia Biotech) or with the silver-staining kit (SilverQuest, Invitrogen). Imaging of stained gels was performed with a Model GS-700 Imaging Densitometer (Bio-Rad). The images of the radiolabeled proteins were generated by autoradiography at -85°C by using Kodak BioMax MR film. Dried SDS gels were subjected to quantitative analysis of [35S]-rhodopsin in a PhosphorImager (Molecular Dynamics). Gels were blotted onto Immobilon-P membranes, and blots were probed with the following antibodies: anti-rhodopsin C-terminal mAb 11D5 (23), affinity-purified anti-ARF4 antibody, or monoclonal antibody against glutathione–protein complexes (Advanced ImmunoChemical, Long Beach, CA), followed by the secondary antibodies conjugated to horseradish peroxidase (HRP). Proteins crosslinked to the biotinylated Br-DAP peptides were detected with SA-HRP. Bound antibodies, or SA-HRP, were detected by using a chemiluminescent Western Lightning immunodetection system (Perkin–Elmer Life Sciences). The distribution of detected antigens was quantified by using multianalyst software (Bio-Rad).

Confocal Microscopy. Confocal microscopy was performed on dark-adapted frog retinas, as described (31). Sections were labeled with monoclonal anti-rhodopsin mAb 11D5 (1:100), followed by Cy3 (red) goat anti-mouse IgG (1:200; Jackson ImmunoResearch), or with affinity purified anti-ARF4 (1:100), followed by Cy2 (green) goat anti-mouse IgG (1:200). All sections were counterstained with the nuclear stain TO-PRO-3 (blue) (1:1,000; Molecular Probes). Confocal optical sections were obtained on a Zeiss 510 Laser Scanning Confocal Microscope by using a 488-nm argon ion laser for Cy2 and 543- and 633-nm HeNe lasers for Cy3 and TO-PRO-3 excitation, respectively. Digital images were prepared by using adobe photoshop CS (Adobe Systems, San Jose, CA).

Electron Microscopy. Frog retinas were fixed in 4% formaldehyde and 1% glutaraldehyde in 0.12 M cacodylate buffer, pH 7.5, for 1 h at 22°C, postfixed in OsO4, and embedded in Epon. Thin sections were stained with uranyl acetate and lead citrate and examined in a Philips (Eindhoven, The Netherlands) CM-100 transmission electron microscope.

Results

The Rhodopsin C-Terminal Peptide and the GST Fusion Protein Containing the C Terminus of Rhodopsin Specifically Bind a 20-kDa Protein, the Small GTPse ARF4. To identify proteins that recognize and bind to the sorting signal of rhodopsin, we used the well-established cell-free assay and the synthetic peptide corresponding to the rhodopsin C terminus, previously shown to inhibit rhodopsin trafficking in vitro (4). The peptide was N-terminally biotinylated and modified by the covalent attachment of the UV-activatable crosslinker Br-DAP (Fig. 1B, peptide C). A truncated peptide modeling the Q344Ter adRP mutation (Fig. 1B, peptide CΔ5), lacking the sorting signal, and consequently the inhibitory activity in the retinal cell-fee system (4), was synthesized as a control. Biotinylated peptides were added to the frog retinal cell-free system that reconstitutes RTC budding in the photoreceptor-enriched PNS (4, 33). Proteins crosslinked to the peptides were detected with SA-HRP. Peptide C specifically crosslinked to a 20-kDa cytosolic protein, which was consequently retained on the SA-Sepharose column (Fig. 1C).

To further confirm the interaction of the 20-kDa cytosolic protein with the C terminus of rhodopsin, we used a GST fusion protein containing the rhodopsin C-terminal sequence (GST-C) and a truncated protein lacking the regulatory sequence (GST-CΔ5). These GST-fusion proteins allowed differential interactions with the rhodopsin C-terminal sorting signal, because GST-C bound the inhibitory anti-rhodopsin mAb 11D5 specific for the targeting sequence (23, 33), but GST-CΔ5 did not (Fig. 1E). In pull-down assays, GST-C specifically bound a cytosolic 20-kDa protein (Fig. 1E).

Sequence analysis, by tandem MS, of the 20-kDa protein crosslinked to the rhodopsin C-terminal peptide revealed one tryptic peptide of ARF4 (peptide 1, Fig. 1D), one with a sequence conserved among class I ARFs (ARFs 1–3) (peptide 2, Fig. 1D) and one tryptic peptide with a sequence highly conserved among all ARF GTPases (peptide 3, Fig. 1D). The rhodopsin C-terminal peptide was also identified in the crosslinked product by its tryptic fragment (Fig. 1D). In Xenopus ARF4, the sequence equivalent to the class I ARF peptide (peptide 2) has one amino acid substitution. Because the 20-kDa protein was isolated from the frog R. berlandieri, it is possible that all three peptides were derived from Rana ARF4. Alternatively, more than one ARF may crosslink to the rhodopsin C-terminal peptide. ARF3, which is present in the photoreceptor Golgi (31, 34), is a potential candidate for this interaction.

Because one of the peptides was unique to ARF4, we made a polyclonal antibody against this peptide to examine the role of ARF4 in rhodopsin trafficking. Anti-ARF4 antibody specifically recognized a ≈20-kDa cytosolic protein (Fig. 2A) and confirmed that the protein crosslinked to peptide C was ARF4 (Fig. 2B). UV illumination was essential for the crosslinking of peptide C to ARF4 but had no effect on the peptide CΔ5 (Fig. 2B). Crosslinking of ARF4 to peptide C caused it to specifically bind to SA-Sepharose (Fig. 2C). Therefore, the interaction of ARF4 with the C terminus of rhodopsin is specific and depends strictly on the presence of the last five amino acids, which constitute its sorting signal.

Fig. 2.

Fig. 2.

Open in a new tab

Interaction of the rhodopsin C terminus with the small GTPase ARF4. (A) Affinity-purified anti-ARF4 antibody recognizes a 20-kDa protein (*) inthe retinal cytosol. (B) Retinal cytosolic proteins crosslinked to the biotinylated Br-DAP peptides separated by SDS/PAGE and immunoblotted with anti-ARF4 or with SA-HRP. Anti-ARF4 detects a 20-kDa protein (*), indicating that ARF4 is present in all cytosolic samples (Left). A duplicate blot probed with SA-HRP (Right) reveals that, in UV-illuminated samples (+UV), the 20-kDa band (*) also contains biotinylated Br-DAP peptide C. Br-DAP peptide CΔ5 does not crosslink to ARF4. (C) Retinal cytosolic proteins crosslinked to the biotinylated Br-DAP peptides, purified with SA-Sepharose, immunoblotted with anti-ARF4 or SA-HRP. ARF4 is readily detectable with anti-ARF4 in the starting material (S) and the unbound fraction (UB) in the samples containing both peptides. However, ARF4 is detected exclusively, albeit in a small quantity, in the bound fraction containing peptide C (peptide C, lane B) but not peptide CΔ5. SA-HRP detects biotinylated Br-DAP peptide C in all three fractions (S, UB, B) but only a ≈20-kDa (♦) and lower-Mr proteins in the bound fraction containing peptide CΔ5 (lane B). Because peptide CΔ5 does not specifically crosslink to any cytosolic proteins in the 20-kDa range (see Fig. 1C), there are no crosslinked products detectable with SA-HRP in the starting or unbound material (peptide CΔ5, lanes S and UB) that could be the source of this material. This protein is biotin-SA and its breakdown products released in abundance from SA-Sepharose during elution (see also Fig. 1 C and D). Biotin-SA eluted with biotinylated Br-DAP peptide C (♦) interferes with the mobility of the specific crosslinked protein containing ARF4 and peptide C (*), which appears to be migrating more slowly in the bound fraction than in the starting material (compare lanes S and B, probed either with anti-ARF4 or with SA-HRP).

Newly Synthesized Rhodopsin Is Crosslinked to ARF4. To determine whether ARF4 interacts with membrane-bound rhodopsin, a cleavable crosslinker sulfosuccinimidyl-3-[(4-azidophenyl) dithio] propionate was added to the frog retinal cell-free system, and crosslinked proteins were separated on nonreducing gels. The crosslinked 50-kDa product detectable by short-term autoradiography, which reveals mostly newly synthesized rhodopsin (23), also contained ARF4 (Fig. 3A, -βME). Anti-ARF4 antibody specifically recognized a ≈20-kDa doublet on nonreducing gels of control PNS (Fig. 3A). Upon crosslinking, a lower-molecular-weight band of the ARF4 doublet was nearly undetectable. Because ARF4 is susceptible to monoglutathionation (35), we sought to determine whether the partial glutathionation is responsible for the differential crosslinking to rhodopsin. A monoclonal antibody that detects protein–glutathione adducts recognized the ARF4 doublet (Fig. 3A). The lower-molecular-weight band recognized by both antibodies appeared to preferentially crosslink to rhodopsin. Thus, in addition to glutathionation, another modification of ARF4 may regulate its binding to rhodopsin.

Fig. 3.

Fig. 3.

Open in a new tab

Newly synthesized rhodopsin is crosslinked to ARF4. (A) Sulfosuccinimidyl-3-[(4-azidophenyl) dithio] propionate (sSADP) was added (+) or not (-) to radiolabeled retinal PNS for 20 min before crosslinking was initiated by UV illumination. Crosslinked proteins were separated on nonreducing (-βME) 8% SDS/PAGE. The autoradiogram of the Commassie stained gel (Left Center), detecting mostly [35S]-rhodopsin, revealed, in addition to the rhodopsin monomer (Rh), a 50-kDa crosslinked product (*) containing newly synthesized rhodopsin and a ≈20-kDa protein. Anti-ARF4 (Right Center) recognized a doublet in the 20-kDa range and a 50-kDa crosslinked protein (*) on the nonreducing gels. The ARF4 doublet was also recognized by anti-glutathione–protein complex antibody (Right), suggesting that both bands contain glutathionated ARF4. Because the 50-kDa crosslinked product is detected in the [35S]-rhodopsin autoradiogram and is recognized by anti-ARF4, the protein crosslinked to rhodopsin is identified as ARF4. (B) The 50-kDa crosslinked product was excised from a nonreducing gel (-βME) and separated on reducing (+βME) 12% SDS/PAGE (Left). The autoradiogram of this gel (Center) shows that cleavage by βME released radiolabeled rhodopsin (Rh), whereas the anti-ARF4 immunoblot (Right) shows that βME also released ARF4 (*) from the 50-kDa crosslinked product.

The 50-kDa crosslinked band was excised and treated with βME, releasing newly synthesized rhodopsin (Fig. 3B). Most importantly, a 20-kDa protein highly enriched in the crosslinked sample was recognized by the anti-ARF4 antibody (Fig. 3B). By contrast, the amount of ARF4 detected in the 50-kDa band on nonreducing gels was very low, possibly due to anti-ARF4 epitope masking in the crosslinked product (Fig. 3A).

ARF4 Is Concentrated in the Vicinity of the Golgi/TGN, Where It Regulates Budding of Post-Golgi RTCs. Although ARF4 is predominantly a cytosolic protein, confocal microscopy of frog retinal sections revealed high immunoreactivity in the vicinity of the Golgi region (arrows in Fig. 4A). The exclusion of ARF4 from the ROS (Fig. 4 A and D), where mature rhodopsin is localized (Fig. 4B), rules out a role for ARF4 in the visual transduction cascade. Instead, the intracellular distribution of ARF4 is consistent with its potential interaction with rhodopsin during RTC membrane budding from the Golgi/TGN. The high immunoreactivity of rhodopsin in the same region (arrows in Fig. 4B, see also Fig. 4 C and F) further supports the notion that the Golgi/TGN may be the site of the interaction of newly synthesized rhodopsin and ARF4.

Fig. 4.

Fig. 4.

Open in a new tab

ARF4 is concentrated in the vicinity of the Golgi/TGN, where it regulates budding of post-Golgi RTCs. (A and D) ARF4 (green) is a retinal cytosolic protein. It is highly concentrated around the photoreceptor Golgi (G) in the RIS (arrows in A) but completely absent from the ROS. Dashed lines outline the RIS and the base of the ROS of a single photoreceptor. M, mitochondria; N, nucleus (blue); retinal layers: ONL, outer nuclear; OPL, outer plexiform; INL, inner nuclear; IPL, inner plexiform. (B) Rhodopsin (red) is abundant in the ROS. In the RIS it is detectable, with mAb 11D5, mostly in the Golgi (arrows), and the RTCs (star). (C and F) EM images detailing photoreceptor structure and the localization of biosynthetic organelles. RPE, retinal pigment epithelium; COS, cone outer segments; asterisk, oil droplet. (Bar, 5 μm in A–C; 10 μm in D; and 0.5 μm in F.) (E) Anti-ARF4 and the ARF4 peptide were added to the in vitro trafficking assay at concentrations inhibitory for mAb 11D5 and the rhodopsin C-terminal peptide but not for the control antibody or peptide (4, 33). (Upper) A representative autoradiogram. (Lower) Budding of RTCs in the control +ATP was set at 100%. The data are presented as the means ± SE of three separate experiments. Like mAb 1D5, anti-ARF4 and the ARF4 peptide profoundly inhibited RTC budding (**, P < 0.005).

To determine whether binding of ARF4 to the C terminus of rhodopsin regulates RTC budding, we added anti-ARF4 antibody to the in vitro assay and compared its effect on RTC budding with that of anti-rhodopsin C-terminal mAb 11D5. As shown in Fig. 4E, at similar concentrations, anti-ARF4 and mAb 11D5 nearly completely inhibited RTC budding, strongly suggesting that both antibodies inhibit the same step in rhodopsin trafficking. Furthermore, the peptide corresponding to the α helix 3 of ARF4, which was used to generate the anti-ARF4 antibody, inhibited RTC budding (Fig. 4E) to the same extent as the rhodopsin C-terminal peptide (4). Thus, the effects of blocking ARF4 action were functionally equivalent to the effects of blocking the rhodopsin C-terminal sorting signal.

Discussion

In this study, we demonstrate that the small GTPase ARF4 specifically recognizes the C-terminal sorting signal of rhodopsin and regulates its incorporation into specialized post-Golgi RTCs. This is supported by chemical and peptide crosslinking, GST pull-downs, confocal microscopy, and by in vitro budding assays. Members of the rhodopsin family of G protein-coupled receptors can interact with ARF1, a class I ARF, to activate PLD (36, 37). This interaction, which involves the highly conserved NPXXY region in transmembrane helix 7, as well as the entire C-terminal domain, is likely to be involved in receptor signaling. By contrast, we find that the activity of ARF4 is necessary for RTC budding from the TGN and, therefore, the specific binding of ARF4 to rhodopsin regulates its trafficking.

Our data suggest that ARF4 and a class I ARF, possibly ARF1 or ARF3, may interact with rhodopsin sequentially, at different stages of transport. Rhodopsin may recruit ARFs through both the NPXXY and VXPX-COOH motifs. The NPXXY region alone is not sufficient for rhodopsin targeting or for ROS morphogenesis (8, 11, 13), but it may contribute to the class I ARF-binding site and participate in PLD activation. Therefore, by recruiting class I ARFs to the Golgi, rhodopsin may activate PLD, which regulates phospholipid biosynthesis and membrane delivery to the ROS (31). The budding and fission reactions involved in the generation of RTCs, which are essential for rhodopsin targeting to the ROS, probably have additional distinct requirements that are best fulfilled by ARF4, so the C-terminal targeting sequence VXPX-COOH plays a crucial role in the recruitment of this ARF. Very little is known about the function of ARF4 and ARF5, but their interaction with arfophilins and arfaptin (17, 19) suggests their potential role in mediating crosstalk between small G protein families. Of particular interest is the possibility that ARF4 represents the link between rhodopsin and the small GTPase rac1, which is essential for RTC fusion (31), and thus may be involved in ROS morphogenesis. In rhodopsin-null Drosophila mutants lacking rhabdomere, constitutively activated rac1 can bypass the requirement for rhodopsin and restore morphogenesis of this light-detecting organelle (38).

The data presented here raise the possibility that the α helix 3 of ARF4 is involved in an interaction with the rhodopsin C terminus. In ARF1, α helix 3 may be the site of interaction with ARF GAP1 (39). Thus, the rhodopsin C terminus may regulate the access of an ARF-GAP to ARF4. Rhodopsin, possibly in conjunction with the TGN coat/adaptor proteins (40), may confer spatially restricted kinetic control of the ARF4-GTPase reaction, thereby coordinating budding and fission reactions until all requirements are met for the generation of mature RTCs that are properly targeted and competent for fusion with the plasma membrane. Alternatively, or additionally, rhodopsin may represent a membrane receptor for ARF4. Although ARF GTPases are activated by phospholipid binding directly and through their nucleotide exchange factors (41), their targeting to distinct organelles requires membrane receptors. Recently, A. Honda, O. S. Al-Awar, J. C. Hay, and J. G. Donaldson (personal communication) have determined that the α helix 3 of ARF1 contains the Golgi targeting signal, which is recognized by the ER/Golgi fusion regulator, membrin. Given the diverse array of cellular activities regulated by ARF GTPases, it is likely that rhodopsin selectively recruits ARF4 to the TGN because of its unique features, such as the recruitment of specific regulatory and/or coat proteins that facilitate RTC budding. Potential regulation of ARF4 activity by glutathionation may also be highly relevant in the retina, which is exposed to near-arterial levels of oxygen and yet is highly susceptible to the damage caused by oxidants and free radicals (42).

Although the precise mechanism of interaction remains to be elucidated, we document direct and specific functional association between ARF4 and the C terminus of rhodopsin, which is essential for its sorting into RTCs and its polarized trafficking. Others have reported that the rhodopsin-targeting sequence binds to the light chain of the microtubule-dependent motor, cytoplasmic dynein (Tctex-1) (43). However, the biological relevance of this potential interaction remains controversial (15), particularly because the delivery of rhodopsin to the ROS appears unperturbed by microtubule depolymerizing agents (44).

The interaction between rhodopsin and ARF4 may help explain how the lack of targeting of rhodopsin C-terminal mutants results in the pathology of adRP. Defective in their capacity to bind ARF4 and hold on to the structural “glue” provided by ARF activity (19) at the sites of RTC budding, rhodopsin C-terminal mutants are likely to distribute through the TGN environment and incorporate indiscriminately by “bulk flow” into post-Golgi carriers destined for different subcellular domains, not only into RTCs. Mislocalization of even a fraction of rhodopsin, which represents 90% of newly synthesized protein in photoreceptor cells, is likely detrimental to synaptic transmission, which is essential for the growth factor regulation and survival of these specialized neurons. Indeed, mistargeting and accumulation of mutant rhodopsin in the rod inner segment (RIS) plasma membrane and the synapse in vivo precede photoreceptor cell death (5, 7). Because ARF4 is a ubiquitous protein, it is unlikely that mutations in ARF4 are a cause of human adRP. However, rhodopsin-ARF4-dependent pathways identified in this study are likely to include other components, which, when mutated, can cause the pathological changes and retinal degeneration seen in patients with adRP.

Acknowledgments

We thank J. Rodman, V. Traverso, and J. Tellier for help with the experiments. This work was supported by National Institutes of Health Grant EY-12421 (to D.D.). Confocal images were generated in the Fluorescence Microscopy Facility supported by National Center for Research Resources, National Science Foundation, National Cancer Institute, and the University of New Mexico Cancer Center. P.A.H. was supported by National Institutes of Health Grant EY-6225 and by an unrestricted departmental award from Research to Prevent Blindness.

Author contributions: D.D. designed research; D.D., A.H.W., N.R., V.M., and A.A. performed research; D.D., A.H.W., N.R., V.M., and P.A.H. analyzed data; V.M. and A.A. contributed new reagents/analytic tools; and D.D. wrote the paper.

Abbreviations: adRP, autosomal dominant retinitis pigmentosa; ARF, ADP-ribosylation factor; PLD, phospholipase D; PNS, postnuclear supernatant; RIS, rod inner segment(s); ROS, rod outer segment(s); RTC(s), rhodopsin transport carrier(s); SA, streptavidin; HRP, horseradish peroxidase; TGN, trans-Golgi network; βME, 2-mercaptoethanol; Br-DAP, N-bromoacetyl-N-(3-diazopyruvoyl)-1,3-phenylenediamine.

References

  • 1.Berson, E. L., Rosner, B., Weigel-DiFranco, C., Dryja, T. P. & Sandberg, M. A. (2002) Invest. Ophthalmol. Visual Sci. 43, 3027-3036. [PubMed] [Google Scholar]
  • 2.Jacobson, S. G., Kemp, C. M., Sung, C. H. & Nathans, J. (1991) Am. J. Ophthalmol. 112, 256-271. [DOI] [PubMed] [Google Scholar]
  • 3.Bessant, D. A., Khaliq, S., Hameed, A., Anwar, K., Payne, A. M., Mehdi, S. Q. & Bhattacharya, S. S. (1999) Hum. Mutat. 13, 83. [DOI] [PubMed] [Google Scholar]
  • 4.Deretic, D., Schmerl, S., Hargrave, P. A., Arendt, A. & McDowell, J. H. (1998) Proc. Natl. Acad. Sci. USA 95, 10620-10625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Green, E. S., Menz, M. D., LaVail, M. M. & Flannery, J. G. (2000) Invest. Ophthalmol. Visual Sci. 41, 1546-1553. [PubMed] [Google Scholar]
  • 6.Li, T., Snyder, W. K., Olsson, J. E. & Dryja, T. P. (1996) Proc. Natl. Acad. Sci. USA 93, 14176-14181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Li, Z. Y., Wong, F., Chang, J. H., Possin, D. E., Hao, Y., Petters, R. M. & Milam, A. H. (1998) Invest. Ophthalmol. Visual Sci. 39, 808-819. [PubMed] [Google Scholar]
  • 8.Tam, B. M., Moritz, O. L., Hurd, L. B. & Papermaster, D. S. (2000) J. Cell Biol. 151, 1369-1380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Sung, C. H., Makino, C., Baylor, D. & Nathans, J. (1994) J. Neurosci. 14, 5818-5833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Humphries, M. M., Rancourt, D., Farrar, G. J., Kenna, P., Hazel, M., Bush, R. A., Sieving, P. A., Sheils, D. M., McNally, N., Creighton, P., et al. (1997) Nat. Genet. 15, 216-219. [DOI] [PubMed] [Google Scholar]
  • 11.Concepcion, F., Mendez, A. & Chen, J. (2002) Vision Res. 42, 417-426. [DOI] [PubMed] [Google Scholar]
  • 12.McNally, N., Kenna, P., Humphries, M. M., Hobson, A. H., Khan, N. W., Bush, R. A., Sieving, P. A., Humphries, P. & Farrar, G. J. (1999) Hum. Mol. Genet. 8, 1309-1312. [DOI] [PubMed] [Google Scholar]
  • 13.Shi, G., Concepcion, F. A. & Chen, J. (2004) in Photoreceptor Cell Biology and Inherited Retinal Degenerations, ed. Williams, D. S. (World Scientific, Singapore), pp. 93-109.
  • 14.Chiu, M. I., Zack, D. J., Wang, Y. & Nathans, J. (1994) Genomics 21, 440-443. [DOI] [PubMed] [Google Scholar]
  • 15.Deretic, D. (2004) in Photoreceptor Cell Biology and Inherited Retinal Degenerations, ed. Williams, D. S. (World Scientific, Singapore), pp. 29-65.
  • 16.Moss, J. & Vaughan, M. (1998) J. Biol. Chem. 273, 21431-21434. [DOI] [PubMed] [Google Scholar]
  • 17.Donaldson, J. G. & Jackson, C. L. (2000) Curr. Opin. Cell Biol. 12, 475-482. [DOI] [PubMed] [Google Scholar]
  • 18.Jackson, C. L. & Casanova, J. E. (2000) Trends Cell Biol. 10, 60-67. [DOI] [PubMed] [Google Scholar]
  • 19.Nie, Z., Hirsch, D. S. & Randazzo, P. A. (2003) Curr. Opin. Cell Biol. 15, 396-404. [DOI] [PubMed] [Google Scholar]
  • 20.Roth, M. G. (1999) Trends Cell Biol. 9, 174-179. [DOI] [PubMed] [Google Scholar]
  • 21.De Matteis, M., Godi, A. & Corda, D. (2002) Curr. Opin. Cell Biol. 14, 434-447. [DOI] [PubMed] [Google Scholar]
  • 22.Peyroche, A., Antonny, B., Robineau, S., Acker, J., Cherfils, J. & Jackson, C. L. (1999) Mol. Cell 3, 275-285. [DOI] [PubMed] [Google Scholar]
  • 23.Deretic, D. & Papermaster, D. S. (1991) J. Cell Biol. 113, 1281-1293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kim, S. W., Hayashi, M., Lo, J. F., Yang, Y., Yoo, J. S. & Lee, J. D. (2003) J. Biol. Chem. 278, 2661-2668. [DOI] [PubMed] [Google Scholar]
  • 25.Hill, K., Li, Y., Bennett, M., McKay, M., Zhu, X., Shern, J., Torre, E., Lah, J. J., Levey, A. I. & Kahn, R. A. (2003) J. Biol. Chem. 278, 36032-36040. [DOI] [PubMed] [Google Scholar]
  • 26.Pittler, S. J., Fliesler, S. J. & Baehr, W. (1992) FEBS Lett 313, 103-108. [DOI] [PubMed] [Google Scholar]
  • 27.Deretic, D. (2000) Methods Enzymol. 315, 77-88. [DOI] [PubMed] [Google Scholar]
  • 28.Taylor, J. M., Jacob-Mosier, G. G., Lawton, R. G. & Neubig, R. R. (1994) Peptides 15, 829-834. [DOI] [PubMed] [Google Scholar]
  • 29.Palczewski, K., Kumasaka, T., Hori, T., Behnke, C. A., Motoshima, H., Fox, B. A., Le Trong, I., Teller, D. C., Okada, T., Stenkamp, R. E., et al. (2000) Science 289, 739-745. [DOI] [PubMed] [Google Scholar]
  • 30.Yeagle, P. L., Alderfer, J. L. & Albert, A. D. (1995) Nat. Struct. Biol. 2, 832-834. [DOI] [PubMed] [Google Scholar]
  • 31.Deretic, D., Traverso, V., Parkins, N., Jackson, F., Rodriguez De Turco, E. B. & Ransom, N. (2004) Mol. Biol. Cell 15, 359-370. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Amor, J. C., Harrison, D. H., Kahn, R. A. & Ringe, D. (1994) Nature 372, 704-708. [DOI] [PubMed] [Google Scholar]
  • 33.Deretic, D., Puleo Scheppke, B. & Trippe, C. (1996) J. Biol. Chem. 271, 2279-2286. [DOI] [PubMed] [Google Scholar]
  • 34.Morel, V., Poschet, R., Traverso, V. & Deretic, D. (2000) Electrophoresis 21, 3460-3469. [DOI] [PubMed] [Google Scholar]
  • 35.Berger, S. J., Claude, A. C. & Melancon, P. (1998) Anal. Biochem. 264, 53-65. [DOI] [PubMed] [Google Scholar]
  • 36.Mitchell, R., McCulloch, D., Lutz, E., Johnson, M., MacKenzie, C., Fennell, M., Fink, G., Zhou, W. & Sealfon, S. C. (1998) Nature 392, 411-414. [DOI] [PubMed] [Google Scholar]
  • 37.Robertson, D. N., Johnson, M. S., Moggach, L. O., Holland, P. J., Lutz, E. M. & Mitchell, R. (2003) Mol. Pharmacol. 64, 1239-1250. [DOI] [PubMed] [Google Scholar]
  • 38.Chang, H. Y. & Ready, D. F. (2000) Science 290, 1978-1980. [DOI] [PubMed] [Google Scholar]
  • 39.Goldberg, J. (1999) Cell 96, 893-902. [DOI] [PubMed] [Google Scholar]
  • 40.Godi, A., Di Campli, A., Konstantakopoulos, A., Di Tullio, G., Alessi, D. R., Kular, G. S., Daniele, T., Marra, P., Lucocq, J. M. & De Matteis, M. A. (2004) Nat. Cell Biol. 6, 393-404. [DOI] [PubMed] [Google Scholar]
  • 41.Kahn, R. A., Terui, T. & Randazzo, P. A. (1996) J. Lipid Mediat. Cell Signal. 14, 209-214. [DOI] [PubMed] [Google Scholar]
  • 42.Winkler, B. S., Boulton, M. E., Gottsch, J. D. & Sternberg, P. (1999) Mol. Vis. 5, 32. [PMC free article] [PubMed] [Google Scholar]
  • 43.Tai, A. W., Chuang, J. Z., Bode, C., Wolfrum, U. & Sung, C. H. (1999) Cell 97, 877-887. [DOI] [PubMed] [Google Scholar]
  • 44.Vaughan, D. K., Fisher, S. K., Bernstein, S. A., Hale, I. L., Linberg, K. A. & Matsumoto, B. (1989) J. Cell Biol. 109, 3053-3062. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES