|Home | About | Journals | Submit | Contact Us | Français|
Mutations in the retinitis pigmentosa GTPase regulator (RPGR) gene account for almost 20% of patients with retinitis pigmentosa. Most mutations are detected in alternatively-spliced RPGR-ORF15 isoform(s), which are primarily but not exclusively expressed in the retina. We show that, in addition to the axoneme, the RPGR-ORF15 protein is localized to the basal bodies of photoreceptor connecting cilium and to the tip and axoneme of sperm flagella. Mass spectrometric analysis of proteins that were immunoprecipitated from the retinal axoneme-enriched fraction using an anti-ORF15 antibody identified two chromosome-associated proteins, Structural Maintenance of Chromosomes (SMC) 1 and SMC3. Using pulldown assays, we demonstrate that the interaction of RPGR with SMC1 and SMC3 is mediated, at least in part, by the RCC1-like domain (RLD) of RPGR. This interaction was not observed with phosphorylation-deficient mutants of SMC1. Both SMC1 and SMC3 localized to the cilia of retinal photoreceptors and MDCK cells, suggesting a broader physiological relevance of this interaction. Additional immunoprecipitation studies revealed the association of RPGR-ORF15 isoform(s) with the intraflagellar transport polypeptide IFT88 as well as microtubule motor proteins, including KIF3A, p150Glued and p50-dynamitin. Inhibition of dynein function by over-expressing p50 abrogated the localization of RPGR-ORF15 to basal bodies. Taken together, these results provide novel evidence for the possible involvement of RPGR-ORF15 in microtubule organization and regulation of transport in primary cilia.
X-linked retinitis pigmentosa (XLRP; MIM 312610) is a relatively severe and genetically heterogeneous inherited retinal degeneration. RP3 is the major subtype of XLRP accounting for over 70% of affected families (1), (2). The RP3 gene, called Retinitis Pigmentosa GTPase Regulator (RPGR), encodes several distinct alternatively-spliced transcripts that are widely expressed (3–5). Mutations in the constitutive RPGR protein of 815 amino acids are detected in approximately 20% of XLRP (6). Subsequent studies revealed an unusual exon, ORF15 (immediately following exon 15) encoding a Gly- and Glu-rich carboxyl-terminal domain of 567 amino acids; mutations in ORF15 accounted for additional 50% of XLRP patients and 25% of RP males with no family history (7–9). Several distinct RPGR isoforms that include complete or part of ORF15 (RPGR-ORF15) are detected preferentially, but not exclusively, in the retina (7,10) and localized to the connecting cilium and/or outer segments of photoreceptors (11–13).
The amino-terminal region of RPGR (termed RCC1-like domain, RLD) shows homology to RCC1, a guanine nucleotide exchange factor (GEF) for Ran, a GTPase involved in nucleocytoplasmic transport (14). Hence, RPGR was predicted to be a GEF for a small GTP-binding protein. However, no such activity or interaction has yet been demonstrated. Yeast two-hybrid analysis using RPGR-RLD had previously identified two proteins, RPGRIP1 (RPGR-interacting protein 1) and PDE6D (delta subunit of rod cyclic GMP phosphodiesterase) (15–18). RPGRIP1 has been localized to the connecting cilium of mouse retina and shown to be mutated in some patients with Leber congenital amaurosis (17,19). PDE6D, on the other hand, is a prenyl-binding protein involved in the solubilization of PDE from rod outer segment disc membrane during phototransduction (20,21). More recent studies have demonstrated that RPGR-ORF15 isoform(s) also interact with the centrosomal protein, nucleophosmin (22), and ciliary IQ-domain protein, nephrocystin-5 (NPHP5 or IQCB1) that is mutated in Senior-Loken syndrome (13).
A majority of RPGR mutations in humans result in early-onset photoreceptor disease (2,7–9). Mutations in RPGR-ORF15 have also been identified in two canine models of retinal degeneration; however, the severity of disease appears to depend upon the type of alteration (23). Mutations leading to complete loss of Rpgr function have not been reported in mouse as yet; nevertheless, the deletion of internal Rpgr exons encoding a part of RLD is shown to cause mild retinal phenotype with late-onset cone-rod degeneration (24). In this mouse retina, mislocalization of opsin-containing vesicles was observed, suggesting a role for RPGR in intracellular trafficking. Despite these studies, the precise role(s) of RPGR-ORF15 in ciliary transport are poorly understood.
To gain insight into RPGR-ORF15 function and to delineate mechanisms of RPGR-associated disease pathogenesis, we performed immunolocalization and immunoprecipitation studies of RPGR-ORF15. We demonstrate that, in addition to the axoneme of photoreceptor connecting cilium, RPGR-ORF15 isoform(s) are localized to the basal bodies in mammalian photoreceptors and to the tip and axoneme of sperm flagella. Further, we describe the interaction of RPGR-ORF15 with two chromosome-associated proteins, SMC1 and SMC3, and their localization to primary cilia of photoreceptors and MDCK cells. Based on these and additional interactions with IFT88 and components of the microtubule-associated molecular motors, we propose that RPGR-ORF15 is involved in regulating ciliary transport assemblies.
Details of RPGR antibodies, ORF15CP (ORF15-specific) and GR-P1 (raised against a peptide from exon 2), have been described (13,22,25). CT-15 antibodies were raised against a previously reported carboxyl-terminal peptide of human RPGR-ORF15, called MCW27-28 (11). Antibodies against γ-tubulin, 14-3-3, p50-dynamitin, dynein heavy chain (DHC), dynein intermediate chain (DIC), SMC1, and SMC3 were purchased from Chemicon (Temeculla, CA). Mouse anti-p150Glued antibody was obtained from BD Transduction Labs (San Jose, CA), and anti-KIF3A and KAP3 antibodies were from Sigma Chemical Co. (St Louis, MO). Anti-RP1 and anti-IFT88 antibodies were generously provided by Dr. Eric A. Pierce (University of Pennsylvania School of Medicine, Philadelphia, PA; (26)) and Dr. Bradley K. Yoder (University of Alabama at Birmingham, Birmingham, AL), respectively.
A mouse cDNA encoding the RPGR protein including RLD and a part of ORF15 (mRPGR-C1) was cloned into the pcDNA-4C vector (Invitrogen, Carlsbad, CA). The mammalian expression constructs encoding full-length human SMC3, SMC1, and its variants at the serine phosphorylation sites (SMC1 S957A, SMC1 S966A, and double mutant SMC1 S957A:S966A (SMC1-DM)) were a generous gift of Dr. Michael B. Kastan (St. Jude Children’s Research Hospital Memphis, TN; (27)).
The ORF15CP, SMC1 and SMC3 antibodies were used for immunogold electron microscopy (EM) of human and mouse retina, as described previously (13,28). The procedures for immunostaining of mouse sperm and MDCK cells have been published (29,30).
Photoreceptor axoneme extract was prepared from frozen bovine retina according to a published procedure (17). Although we did not use sucrose-gradient centrifugation to isolate axonemal proteins, enrichment of γ- and acetylated α-tubulin validated the purity of the axoneme preparation. IP using the ORF15CP antibody was carried out as described elsewhere (13), (22). The proteins were analyzed by SDS-PAGE, followed by immunoblotting and/or staining with Coomassie Blue. In some instances, the protein bands were excised from the gel and subjected to tandem mass spectrometry.
MDCK type II cells were transfected using the Polyfect reagent (Qiagen, Valencia, CA). For IP, cells were lysed in 1X PBS containing 0.1% Triton X-100 and complete protease inhibitor cocktail (Roche, Palo Alto, CA) and incubated with the primary antibody overnight. Immune-complexes were collected using Protein-A or -G Sepharose beads (Invitrogen, Carlsbad, CA), washed with 1X PBS containing 1% Triton X-100, and analyzed by SDS-PAGE followed by immunoblotting.
A fragment of the human RPGR cDNA (encoding residues 33–392, which are part of RLD) was cloned into pGEX4T-2 (Amersham Biosciences, Piscataway, NJ) in-frame with GST. The GST-RLD fusion protein and native GST were purified to homogeneity per manufacturer’s instructions. The pulldown assays were performed using 5 μg of GST or GST-RLD fusion protein with bovine retinal axoneme extract (250 μg), as described (31).
The proteins were synthesized in vitro from pcDNA plasmid constructs using TNT-T7 quick-coupled rabbit reticulocyte translation system (Promega, Madison, WI), in the presence or absence of 35S-labeled Methionine (Amersham Biosciences) and used for co-immunoprecipitation, as described (32).
mIMCD-3 (American Type Culture Collection, Manassas, VA; ATCC # CRL-2123) or ARPE-19 (ATCC # CRL-2302) cells were grown on coverslips in six-well plates and transfected with myc-tagged p50-dynamitin expression vector (kindly provided by Dr. R. Vallee, Columbia University, NY). After incubation for 48 h, cells were washed in PBS, fixed with ice-cold methanol, blocked with 2% BSA in PBS, and incubated with the primary antibody. After washing in PBS, cells were blocked again and incubated with Texas Red or FITC-conjugated secondary antibodies. Cells were mounted in Vectashield (Vector Laboratories Ltd) containing DAPI. Images were captured using an Axioplan fluorescence microscope and analyzed using IPLab software.
With a goal to determine the function of RPGR-ORF15 in primary cilia, and specifically in photoreceptors, we performed immunolocalization studies using the ORF15CP antibody. Previous studies have reported the localization of RPGR-ORF15 to photoreceptor axoneme. However, we observed by immunoEM that the basal bodies of both human and mouse photoreceptor cells were also labeled (Fig. 1A). Consistent with the observation that XLRP patients may exhibit abnormal sperm tails and axoneme (33), ORF15CP co-localized with acetylated-α-tubulin to the tip and tail axoneme of mouse sperm flagella (Fig. 1B). We also detected similar co-localization of RPGR-ORF15 with acetylated-α-tubulin to the primary cilia of MDCK cells with a punctate staining pattern (Fig. 1B), as observed with another RPGR-interacting ciliary protein IQCB1 (13).
To identify proteins that exist in complex(es) with RPGR-ORF15 in photoreceptor cilia, we immunoprecipitated retinal axoneme-enriched fraction using the ORF15CP antibody. As predicted, the axoneme fraction was enriched in γ- and acetylated α-tubulin and RPGR-ORF15 isoforms (Fig. 2A). Mass spectrometry analysis of Coomassie blue-stained protein bands of the immunoprecipitate identified multiple peptides for two ATP-binding proteins, SMC1 and SMC3 (5 out of 6 peptides for SMC1, and 8 out of 10 for SMC3) that are involved in maintaining chromosome dynamics during cell cycle (34). Both SMC1 and SMC3 were detected in the axoneme fraction and in the detergent soluble fraction, which includes nuclear proteins (Fig. 2A). Detection of SMC1 and SMC3 upon immunoblot analysis of the ORF15CP-immunoprecipitate (Fig. 2B) provided further evidence in support of their existence in RPGR-ORF15 complex(es). Similar results were independently obtained with another set of antibodies against RPGR-ORF15, SMC1 and SMC3 (data not shown). Reverse IP with anti-SMC1 or SMC3 antibody did not reveal RPGR-ORF15, probably due to the low abundance of SMC1 and SMC3 in the axoneme preparation. Normal rabbit IgG did not immunoprecipitate any specific protein (Fig. 2B).
Given that RPGR-RLD interacts with RPGRIP1 and PDE6D (16,17), we examined whether this domain is involved in interaction with SMC1 and/or SMC3. In pull-down assays, the GST-RLD fusion protein but not GST was able to associate with endogenous SMC1 and SMC3 in retinal axoneme extracts (Fig. 3A). To further investigate this interaction, we transfected MDCK cells with a construct encoding the 90 kDa Xpress-tagged mRPGR-C1 protein that included the intact RLD but only a truncated ORF15. In these experiments, IP using anti-Xpress antibody could pull-down endogenous SMC1 and SMC3 (Fig. 3B). Cells transfected with the vector alone did not pulldown either SMC1 or SMC3. In reciprocal experiments, anti-SMC1 or anti-SMC3 antibodies could immunoprecipitate Xpress-tagged mRPGR-C1 (Fig. 3B).
It has been demonstrated that phosphorylations at S957 and S966 residues are critical for SMC1 function (27). To examine the effect of SMC1 phosphorylation on its interaction with RPGR, we used in vitro translated 35S-labeled wild-type and mutant SMC1 proteins. IP using anti-Xpress (Xp) antibody followed by autoradiography revealed the interaction of RPGR with the wild-type SMC3, SMC1, and the single mutant SMC1 S957A, but not with SMC1 S966A and SMC1-DM (Fig. 3C). Similar results were obtained in reverse experiments using 35S-labeled mRPGR-C1 and unlabeled mutant SMC1 proteins (Fig. 3C).
SMC1 and SMC3 have been shown to be associated with chromosomes and mitotic spindle (34). To evaluate the physiological relevance of RPGR’s interaction with SMCs, we performed EM immunogold studies. Antibodies against SMC3 labeled the entire length of the cilium in mouse photoreceptors, including the basal bodies (Fig. 4A). Antibodies against SMC1 also significantly labeled the photoreceptor cilium, although this labeling was much less robust than that with the SMC3 antibodies (Fig. 4A). Additional immunogold labeling with SMC1 antibodies was observed in the photoreceptor inner segments and the ribbon synapse (data not shown). The cilium labeling was confirmed by co-localization of SMC1 and SMC3 staining with acetylated α-tubulin in the primary cilia of dissociated mouse rod photoreceptors (data not shown). Furthermore, SMC1 and SMC3 co-localized with acetylated α-tubulin along the entire length of the cilia in MDCK cells in a punctate pattern similar to RPGR-ORF15 (Fig. 4B).
The studies described above prompted us to investigate the interaction (whether direct or indirect) of RPGR-ORF15 with other basal body and microtubule-associated proteins. Immunoblot analysis revealed that basal body proteins IFT88, γ-tubulin and 14-3-3 could be co-immunoprecipitated by ORF15CP from the retinal axoneme preparation; these proteins were not detected when normal IgG was used instead of the ORF15CP antibody (Fig. 5A). Reverse IP experiments showed that 14-3-3 was able to pulldown RPGR-ORF15 isoforms, but γ-tubulin did not (Fig. 5B). This may reflect the relative abundance of different proteins. Notably, two other basal body proteins, centrin and pericentrin, were not present in the RPGR-ORF15 immunoprecipitate (Fig. 5A and B).
We then examined the association of RPGR-ORF15 with microtubule-associated motor assemblies in the axoneme of the connecting cilium (35,36). Immunoblotting of the RPGR-ORF15 immunoprecipitate showed the presence of kinesin II subunits KIF3A and KAP3, dynein subunit DIC, as well as dynactin subunits p150Glued and p50-dynamitin (Fig. 5A). The RP1 protein, a known axonemal component (26), was not detected in the ORF15CP immunoprecipitate. Although cytoplasmic DHC immunoreactive bands were not detected in the retinal axoneme fraction (data not shown), anti-DHC antibody was able to pulldown RPGR-ORF15 from axoneme extracts (Fig. 5B).
We then examined whether localization of RPGR-ORF15 to basal bodies is dependent upon the retrograde dynein-dynactin motor complex. For this purpose, the dynein activating complex of dynactin was disrupted by overexpressing p50-dynamitin subunit (37). In the transfected ARPE19 cells expressing p50-dynamitin, ORF15CP-specific RPGR signal was not evident (Fig. 6). As demonstrated previously, anti-ninein labeling was also not observed in p50-overexpressing cells (37), whereas γ-tubulin localization was unaltered (Fig. 6).
As we discuss later, our results indicate a role for RPGR-ORF15 in regulating ciliary transport. This is of particular interest with respect to the transport along photoreceptor cilium since human RPGR mutations result in relatively early-onset retinal degeneration. A major question is whether photoreceptor degeneration in the patients with RPGR mutations is caused by defects in protein trafficking through the cilia. An animal model is necessary to investigate the underlying biochemical mechanism(s). However, the only available Rpgr-knockout mouse exhibits a mild and late cone-rod degeneration with corresponding late-onset alterations in the transport of opsin to the photoreceptor outer segment (24). Since this model was generated by deleting Rpgr exons 4–6, we wanted to examine whether ORF15 transcripts or protein isoforms are expressed in the retina of this mouse. RT-PCR analysis using multiple primer sets revealed ORF15-containing Rpgr transcripts in the Rpgr-knockout retina (M. I. Othman and A. S., unpublished data). RPGR-ORF15 isoform(s), as identified by ORF15CP, CT-15 and GR-P1 antibodies, were still detectable in this mouse retina, whereas the constitutive Rpgr isoform of ~80 kDa (identified only by the amino-terminal GR-P1 antibody) was not observed (Fig. 7A). Further evidence of RPGR-ORF15 expression in the Rpgr knockout retina was provided by immunogold microscopy using the ORF15CP antibody, which revealed the basal body as well as axoneme staining (Fig. 7B). Hence, this genetic model is not useful for testing our hypothesis.
Vertebrate photoreceptors are highly polar neurons with distinct morphology and subcellular organization. The connecting cilium of a photoreceptor cell is a modified primary cilium that forms a bridge between the inner and the outer segment (36). It contains a microtubule-based axoneme, which initiates from the basal body in the inner segment and continues into the outer segment (26). The outer segment is comprised of an ordered array of stacked membrane discs; approximately 10% of disks are replenished each day and in a mouse photoreceptor ~70 opsin molecules per second are transported to the outer segment (38). In addition to the anterograde transport of opsin and other phototransduction proteins, bidirectional movement of arrestin and transducin has been demonstrated through the connecting cilium (39). The connecting cilium therefore represents a critical junction in the cell biology, and consequently, the viability of the photoreceptors. The present study develops our understanding of the role of RPGR in primary cilia, particularly the photoreceptor cilium. We have shown that the distribution of RPGR-ORF15 includes the ciliary basal bodies, which function as a gateway to the cilium (38). Moreover, we have identified binding partners of RPGR-ORF15, including microtubule motors and SMC proteins, which are involved in microtubule-based movement of chromosomes but whose ciliary function has not been realized. Thus, our findings suggest a significant role for RPGR in regulating transport along the photoreceptor cilium.
We have consistently observed multiple specific isoforms of RPGR-ORF15 that are generated, at least in part, by alternative splicing. Accumulating evidence indicates that distinct ORF15 isoforms may be localized to different subcellular compartments within the photoreceptors and perform specific functions (11–13,22,25). A common theme is now emerging regarding RPGR’s role in intraphotoreceptor transport. The interactions of RPGR with PDE6D (18), RPGRIP1 (17) and NPHP5 (13), the localization of one RPGR-ORF15 isoform to centrosomes of dividing cells and its association with nucleophosmin (22) are consistent with a role in microtubule dynamics. The studies described in this report strongly suggest specific function of RPGR-ORF15 in regulating the ciliary transport at the level of basal bodies.
Basal bodies of primary cilia are the docking sites for proteins involved in assembly, maintenance, and function of the cilia. There is a selective transport of cargo from the basal body to the axoneme, which is partly carried out by the IFT polypeptides and polarity proteins (30,40). IFT88 is required for assembly and maintenance of the photoreceptor outer segment and photoreceptor viability (41). Based on our observations of RPGR-ORF15, IFT88 and kinesin-2 proteins (KIF3A and KAP3) as part of a multi-protein complex in the retinal axoneme, we hypothesize that RPGR-ORF15 is involved in the selection of cargo, which is carried by kinesin-2 along the cilium. Our hypothesis is consistent with a previous report that IFT88 associates with kinesin-2 in the retina (42). Nevertheless, it should be noted that two of the potential cargo proteins, opsin and arrestin, were not detected as part of the RPGR-ORF15 complex(es) (data not shown).
The association of RPGR-ORF15 isoforms with both anterograde (kinesin-2) and retrograde (cytoplasmic dynein-dynactin complex) molecular motors is an interesting and significant finding. While the kinesin-2 complex has been shown to participate in compartmentalized ciliogenesis in Drosophila sensory cilia and inter-segmental transport in mouse retina (43,44), the function of dynein-dynactin complex in the retina is poorly understood. The dynactin subunits p50-dynamitin and p150Glued are responsible for tethering cargo to the dynein motor (45) and regulate transport of several microtubule-associated proteins (46). The dynein-dependent localization of RPGR-ORF15 to basal bodies, as observed for the BBS4 protein (47), provides further evidence in support of the functional relevance of RPGR-dynein association.
The interaction of SMC1 and SMC3 with RPGR-ORF15 and their localization to the photoreceptor axoneme suggest a broader role for SMC proteins in microtubule dynamics. SMC1 and SMC3 are large coiled-coil proteins associated with chromosomes, share structural similarity with the microtubule motor protein kinesin, and are involved in ATP-dependent chromosomal movement along spindle microtubules during cell division (34). The mechanism by which neurons establish their polarity is similar to spindle organization during mitosis (48). Our immunolocalization of SMC1 and SMC3 to primary cilia in the retina as well as in cultured mammalian cells demonstrates that these proteins are also associated with ciliary microtubules. SMC proteins, including 1 and 3, are listed as part of the sensory cilia in a recent genomic study (49), which supports our findings.
Abnormal sperm tails and instability of sperm axonemes have been observed in patients with XLRP (33). RPGR-ORF15 staining in the tip and the axoneme of mouse sperm flagella is consistent with these clinical findings. The flagellar tip is the site for axoneme turnover, a process similar to the turnover in photoreceptor outer segments (50). Notably, abnormal nasal ciliary axonemes and hearing defects are also detected in some patients with RPGR mutations (51,52). Taken together, it appears that mutations in RPGR lead to defects in microtubule-stability/maintenance but not cilia biogenesis. Consistent with this hypothesis, cilia formation is not compromised in the Rpgr−/− retina (24) or in XLRP patients (53).
In summary, we have demonstrated that RPGR-ORF15 isoform(s) are present in the axoneme and basal bodies of primary cilia and associated with proteins that are components of basal bodies and microtubule-based motor assemblies. These results suggest that RPGR-ORF15 functions in regulating transport along primary cilia, including the photoreceptor cilium. The photoreceptor degeneration (and sperm defects) observed in XLRP patients with RPGR mutations is therefore predicted to result from defects in transport assemblies in the photoreceptor cilia. A genetic test of this hypothesis must await an animal model in which the RPGR-ORF15 isoform is deleted or nonfunctional – unlike the present Rpgr knockout mouse (24).
We thank Dr. Tiansen Li for the Rpgr knockout mice and RPGRIP1 antibody; Robert Duerr, Michael Wade, and members of the Swaroop lab for constructive comments, and S. Ferrara for administrative support. The authors acknowledge the Michigan Proteome Consortium, the Michigan Economic Development Corporation, and Michigan Technology Tri-Corridor for mass-spectrometric analysis (Grant #085P1000818). This research was supported by grants from the National Institutes of Health (EY07961, EY07003, EY13408, EY12598, and DK069605), The Foundation Fighting Blindness and Research to Prevent Blindness (RPB). TWH was supported by polycystic kidney disease foundation grant (92a2f). AS is Harold F. Falls Collegiate Professor and RPB Senior Scientific Investigator. BM is an investigator of the Howard Hughes Medical Institute.