Developmental and Tissue Expression of Xenopus laevis RPGR

Abstract

purpose. The present study examined the developmental and tissue expression of the retinitis pigmentosa GTPase regulator (RPGR) gene in Xenopus laevis.

methods. The cDNA for X. laevis RPGR (XRPGR) was isolated from adult eye mRNA by reverse transcription–polymerase chain reaction (RT-PCR) and rapid amplification of cDNA ends. The deduced peptide sequence was aligned with RPGR orthologues. Gene expression was examined by whole-mount in situ hybridization and RT-PCR. The localization of XRPGR in X. laevis photoreceptor cells and XTC-2 cells was determined by immunostaining.

results. The XRPGR^ex1–19 isoform encodes a protein of 727 amino acids containing an RCC1 domain and a C-terminal isoprenylation anchorage motif. It shares 33% to 41% amino acid identity with human, mouse, and dog RPGR. The C-terminal exon of the alternatively spliced RPGR^ORF15 isoform is also conserved across species. XRPGR is expressed at the earliest stages of X. laevis development and persists into adulthood, where expression is highest in the eye. XRPGR is expressed in presumptive eye fields (stages 18 to 22), becoming largely restricted to the central retina (stages 28 to 40). XRPGR protein colocalizes with β-tubulin at the X. laevis ciliary axoneme and with γ-tubulin at centrosomes in XTC-2 cells.

conclusions. XRPGR is widely expressed throughout development but shows highest expression after the appearance of the eye primordium and persists in the eye into adulthood. The data are consistent with XRPGR expression in a single microtubular organelle—the centriole or basal body and associated ciliary transitional zone found in modified sensory cilia of photoreceptors and motile cilia.

Retinitis pigmentosa (RP) is a heterogeneous group of inherited retinal degenerations with a worldwide prevalence of 1 in 4000 in the general population. ¹ RP can be inherited as an autosomal dominant, autosomal recessive, X-linked, mitochondrial, or digenic trait. X-linked RP (XLRP) is one of the most severe forms of human retinal degeneration, as determined by age of onset and progression, and accounts for 6% to 20% of all RP cases. ^{2 3 4 5} Linkage analyses have identified at least four XLRP loci: RP2, RP3, RP23, and RP24. ^{6 7 8 9} RP3 accounts for the majority (70%–80%) of XLRP cases in most series. ^{10 11} The RPGR gene is responsible for the RP3 form of XLRP ^{7 12} and is mutated in 15% to 20% of all RP patients. ⁴ This gene was initially found to contain 19 exons (RPGR ^ex1–19 ), coding for a widely expressed transcript and a predicted protein of 90 kDa. ^{7 12} Mutations in this transcript accounted for only 15% to 20% of XLRP patients, but all known disease-causing mutations were subsequently found to be present within a single alternatively spliced transcript containing an unusual carboxyl-terminal exon called ORF15 (RPGR ^ORF15 ). ¹³ This exon encodes a glutamic acid– and glycine-rich domain and contains highly repetitive simple-sequence and purine-rich motifs that have been shown to harbor a high frequency of microdeletions and stop codons. ^{4 13 14 15 16 17 18 19} Mutations in RPGR ^ORF15 are found in a variety of other retinal dystrophies, including X-linked forms of human cone dystrophy, ^{13 17} cone-rod dystrophy, ¹⁸ atrophic macular dystrophy, ¹⁹ RP with sensorineural hearing loss, ²⁰ and progressive retinal atrophy in dogs. ²¹  

The function of RPGR is poorly understood, although the amino (N) terminal domain of RPGR (exons 1–11) shows homology to RCC1, a guanine nucleotide exchange factor for the small GTPase Ran. ^{7 12 22} This suggests that RPGR may play a role in the regulation of one or more small G-proteins in the retina or retinal pigment epithelium (RPE), but RPGR has not yet been shown to directly interact with a G-protein. RPGR is associated with the connecting cilium of rod and cone photoreceptors, but it remains unclear whether it is also expressed in the photoreceptor outer segment (OS). ^{23 24 25} Rpgr knockout mice develop a slowly progressive degeneration of cone and rod photoreceptors that is associated with abnormal basal discs and mislocalization of opsin-containing vesicles, suggesting that RPGR is required for opsin transport and the integrity of the photoreceptor OS. ²³  

Xenopus laevis has been widely used as a model to study development and to understand the mechanisms of retinal degeneration. ^{26 27 28} To use X. laevis as a model to investigate RPGR function, we examined the cloning and expression of X. laevis RPGR (XRPGR) during development, in adult tissues, and in the tadpole-derived XTC-2 cell line. 

Methods

Animals and Tissues

X. laevis embryos were dejellied in 2% cysteine-HCl (pH 8.0), grown in 0.1× modified Barth’s solution (MBS), and staged according to Nieuwkoop and Faber. ²⁹ Brain, eye, kidney, liver, and heart were dissected from adult X. laevis. Experimental procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Total RNA Extraction and RT-PCR

Total RNA was extracted from adult X. laevis brain, eye, kidney, liver and heart using an isolation system (RNAgents; Promega, Southampton, UK) according to the manufacturer’s instructions. X. laevis embryos at specific developmental stages were homogenized in an extraction buffer consisting of 0.3M NaCl, 1 mM EDTA, 20 mM Tris-HCl (pH 7.6), and 1% SDS. The homogenates were spun for 10 minutes at 13,000g, and supernatants were removed and extracted with phenol and chloroform, then precipitated with ethanol at −20°C. The resulting total RNA pellets were washed with 70% ethanol, dried, and resuspended in 50 μL nuclease-free water. The total RNA was then treated with 10 U amplification-grade DNase I (Roche, Lewes, UK) at 37°C for 15 minutes, extracted with phenol and chloroform, and precipitated with ethanol. After centrifugation and washing as described above, the pellets were resuspended in 30 μL nuclease-free water. 

Half of the resulting RNA sample (5 μg) was reverse transcribed using a first-strand cDNA synthesis kit (Invitrogen, Paisley, Scotland) with dT(15) primer, 400 U reverse transcriptase (SuperScript II; Invitrogen), and 40 U RNase inhibitor (RNasin; Promega) at 42°C in a final volume of 50 μL. The remaining RNA was used to prepare a control sample in which reverse transcriptase was omitted. Aliquots (1–2 μL) of each sample were used for PCR analysis in a 25-μL reaction volume using XRPGR gene-specific primers (Table 1)under standard conditions. 

Identification of X. laevis RPGR (XRPGR) cDNA Sequences

XRPGR gene-specific primers (XRPGR N1, XRPGR C1, XRPGR N2, and XRPGR C2; Table 1 ; Fig. 1B ) were designed based on two X. tropicalis expressed sequence tags (ESTs; Accession Nos. AL849961 and AL857113), which encompassed the RPGR RCC1-like domain. The initial reactions used X. tropicalis sequence-based primers XRPGR N1 (forward) and XRPGR C1 (reverse) from EST AL849961 and XRPGR N2 (forward) and XRPGR C2 (reverse) from EST AL857113) to perform PCR, using X. laevis eye cDNA as template (Fig. 1B) . A longer PCR spanning the two ESTs was carried out using primers XRPGR N1 and XRPGR C2. The resultant PCR products were resolved by electrophoresis in 1% (wt/vol) agarose gels, cloned into a plasmid vector (pGEM-T Easy; Promega), and sequenced using T7 and SP6 (Promega), XRseq N1, Xrseq C1, and Xrseq C2 primers (Table 1 ; Fig. 1B ). All sequences were obtained in forward and reverse orientations and verified by sequencing six independent clones. Control experiments using samples prepared without reverse transcriptase were performed to ensure that genomic DNA contamination did not contribute to the PCR amplification. These sequences were extended using 5′– and 3′–rapid amplification of cDNA ends (RACE) reactions in X. laevis cDNA, as described below. The ORF15 carboxyl (C)-terminal exon of XRPGR was PCR amplified from X. laevis cDNA using two primers, XRPGR N3 and XRPGR C3 (Table 1) , corresponding to the X. tropicalis genomic sequence scaffold_83 (http://genome.jgi-psf.org/), which gave a product of 1.77 kb, which was purified and sequenced in both directions. This sequence (Accession No. DQ176000) included a substantial part of C-terminal exon ORF15 (Fig. 2D) , but it was not possible to link this to upstream exons 13 to 15. 

5′- and 3′-RACE

The initial X. laevis cDNA sequences were extended with RACE reactions, using 5′- and 3′-RACE PCR kits (Invitrogen) according to the manufacturer’s instructions. Total mRNA from X. laevis eye was reverse transcribed and first-strand synthesis was carried out using primer XR5RaceR1, based on the initial X. laevis cDNA sequence, after which an oligo dC–containing 5′-RACE anchor primer (Invitrogen) was added to the 3′ end using terminal deoxynucleotide transferase. This was amplified by PCR, using nested XRPGR-specific primer XR5RaceR2 (Fig. 1B)and a deoxyinosine-containing 5′-RACE abridged anchor primer (Invitrogen). The product was electrophoresed on a 1.5% agarose gel, purified, and sequenced. 

For 3′-RACE, total X. laevis eye mRNA was similarly reverse transcribed using an oligo dT–containing 3′-RACE adapter primer (Invitrogen), and first-round PCR amplification was performed using an XRPGR-specific primer, XR3Race1 (Table 1) . A second round PCR amplification was performed with the XRPGR-specific primer XR3Race2 (Table 1)and the 3′-RACE abridged universal amplification primer, which is antisense to the unique 3′ end of the adapter primer (Invitrogen). Controls included the use of non-anchored first-strand template, reactions excluding reverse transcriptase, and the use of a single primer for amplification. The products were electrophoresed on 1.5% agarose gels, purified, and sequenced. All PCR amplifications were carried out in a final volume of 50 μL with 1.5 mM dNTP, 20 pmol each primer, and 2 U Taq polymerase using hot start. PCR conditions were 1 minute at 95°C for 1 cycle, then 94°C for 1minute, 60°C for 1 minute, and 72°C for 1 minute for 35 cycles. In this way, the full-length sequence of the exon-1 to -19 constitutive or default transcript ^{7 25} was identified (Accession No. DQ175998). 

Protein Sequence Alignment

XRPGR protein sequences were deduced from the cDNA sequence using a public system (Expert Protein Analysis System; available at www.expasy.ch), and aligned with RPGR orthologues using a public tool (ClustalW; available from the European Bioinformatics Institute at www.ebi.ac.uk). 

Whole-Mount In Situ Hybridization

In situ hybridization was carried out essentially as described previously. ³⁰ XRPGR probes (antisense and sense) corresponding to the 1.25-kb cloned sequence lying between primers XRPGR N1 and XRPGR C2, from amino acid residues 20 to 444, corresponding to most of the RCC1 domain (Figs. 1B 2A ; Table 1 ), were transcribed in vitro from linearized plasmid vectors (pGEM-T Easy; Promega) with XRPGR insert using 10 × digoxigenin (DIG) RNA labeling mix (Roche). Labeled probes were purified on microcolumns (ProbeQuant-G-50; Amersham Pharmacia Bioscience, Bucks, UK). X. laevis embryos were fixed in MEMFA buffer (0.1 M MOPS, 2 mM EDTA, 1 mM MgSO₄, 3.7% formaldehyde) for 2 hours at room temperature and stored in 100% methanol at −20°C until processing. Embryos were rehydrated through a series of graded methanol washes, then rinsed in PBS with 0.1% Tween 20. Embryos were bleached with 10% H₂O₂ in 5× SSC and prehybridized for 6 hours at 75°C in 50% formamide, 5 × SSC, 1 mg/mL yeast total RNA, 100 μg/mL heparin, and 0.1% Tween 20. The probe (500 ng/mL) was then added and hybridized overnight. Five 30-minute washes in 2 × SSC and 0.2 × SSC at 60°C were followed by similar washes at room temperature with maleic acid buffer (MAB; 0.1 M maleic acid, 0.15 M NaCl, 0.1% Tween [pH 7.8]) and finally blocking in 2% blocking reagent (Roche) at 4°C overnight. Embryos were then incubated with anti-DIG antibody (Roche) at a dilution of 1:2000 in a blocking solution (Roche) at 4°C overnight. After washing the embryos five times for 30 minutes each at room temperature in MAB, the antibody detection step was performed using a BM Purple precipitating alkaline phosphatase detection system (Roche). Embryos were cleared in Murray’s reagent (2:1 benzyl benzoate/benzyl alcohol) and photographed under a dissecting microscope with a digital camera (Coolpix; Nikon, Surrey, UK). In all cases, both sense and antisense probes were compared with controls for any background signal in the reactions. 

Some stained embryos (stage 40) were embedded in paraffin, and 5-μm-thick sections were cut to enable detailed analysis of structures where expression occurred. The sections were counterstained with eosin and analyzed using bright-field microscopy. 

Immunocytochemistry

Photoreceptor cells were dissociated from the retinas of adult X. laevis and processed for immunofluorescence using methods that retain protein immunoreactivity and microtubule stability, as previously described. ³¹ Anti-hORF15¹⁸⁷⁸ antibodies were raised against human RPGR^ORF15 as previously described. ³² Experimental slides were incubated with anti-human ORF15¹⁸⁷⁸ and/or monoclonal antibodies to α- or β-tubulin (Sigma, Dorset, UK). ³¹ Control slides were incubated with preimmune or normal sera or PBS. Experimental and control slides were then incubated with fluorochrome-conjugated secondary antibodies or PBS and examined as previously described. ³¹  

The X. laevis tadpole-derived XTC-2 cell line ³³ was cultured on coverslips in Leibovitz L-15 medium (Invitrogen) containing 10% FCS and 20% dH₂O. After two days in culture, the cells were fixed and permeabilized with −20°C methanol for 10 minutes. The cells were rehydrated with PBS for 10 minutes, then blocked in PBS with 2% BSA for 20 minutes. Cells were then incubated with the anti-bORF15^C2 or anti-hORF15¹⁸⁷⁸ antibodies (1:500) plus anti–γ-tubulin (1:1000, Sigma) in PBS with 2% BSA for 1 hour at room temperature. Cells were washed in PBS and blocked again with 2% goat serum in PBS with 2% BSA for 20 minutes. After washing with PBS, cells were incubated with FITC-conjugated and Texas Red–conjugated secondary antibodies (Jackson Laboratories, Soham, UK) for 1 hour at room temperature in darkness. Cells were then washed in PBS and mounted in mounting medium (Vectashield; Vector Laboratories, Ltd., Peterborough, UK) containing DAPI (1.0 μg/mL). Images were captured using a fluorescent microscope (Axioplan 2; Zeiss, Elstree, UK) and analyzed using commercial software (IPLab; Scanalytics, Rockville, MD). 

Immunoblotting

Retinas were dissected from four freshly enucleated X. laevis eyes and homogenized in an SDS–protein sample buffer (50 mM Tris-HCl [pH 6.8], 0.1% bromophenol blue, 10% glycerol, 1% SDS), and equal amounts of protein from retinal homogenates were subjected to 10% SDS-PAGE and transferred to a nitrocellulose membrane. Western blot analysis with anti-RPGR^ORF15 antibody was carried out as previously described. ³²  

Results

Cloning and Sequence Analysis of XRPGR

The full-length cDNA sequence of XRPGR ^ex1–19 was identified using the X. tropicalis EST clones AL849961 and AL857113, which contained the RCC1-like domain (Figs. 1A 1B) . The resulting PCR products were extended by 5′- and 3′-RACE reactions (Figs. 2A 2B) . RPGR ^ORF15 contains exons 1 to 15 of the RPGR ^ex1–19 transcript but has an alternative C-terminal exon, ORF15 (Fig. 1A) , ¹² which was identified using X. tropicalis genomic sequence to PCR amplify it from X. laevis eye cDNA. This identified X. laevis cDNA encoding 590 amino acids of exon ORF15, representing the major part of this exon, which ranges in size from 567 to 934 amino acids in various mammalian and fish species (Fig. 2C) . However, since it was not possible to link exon ORF15 with exon 15 by PCR, this isoform is incomplete. 

The XRPGR ^ex1–19 transcripts of X. laevis (727 amino acids) and X. tropicalis (729 amino acids) are shorter than their mammalian orthologues (815 to 1003 amino acids) (Figs. 2A 2B) . Alignment of the predicted XRPGR^ex1–19 protein sequence with orthologues shows that the deduced amino acid sequence is highly homologous to X. tropicalis (85% identity), and moderately homologous to human (41% identity), mouse, and dog (33% identity). The degree of homology is greater over the RCC1-like domain: identity was highest in X. tropicalis RPGR (90% identity), followed by human and dog (63%), and mouse (62%) (Fig. 1A) . The isoprenylation motif of human RPGR^ex1–19 was conserved in both Xenopus species (Fig. 2B) . The majority (18/22) of all disease-associated RPGR missense mutations in the RCC1-like domain reported to date was conserved across all five species, the exceptions being those occurring at residues D312, I289, and G436 (Fig. 2A) . The C-terminal region of XRPGR^ORF15, which is predicted to be a functional domain for interaction with the chaperone nucleophosmin, ³² was also strongly conserved across species (Fig. 2C) , supporting the view that this is a functionally important domain. 

Expression of XRPGR

The temporal and spatial expression pattern of XRPGR during embryogenesis was first examined by RT-PCR, using primers directed to the RCC1-like domain and therefore capable of detecting both XRPGR ^ex1–19 and XRPGR ^ORF15 . Total RNA was prepared from oocytes, fertilized eggs, and blastula-, gastrula-, neurula-, tailbud-, and tadpole-stage embryos. XRPGR mRNA was readily detected at the time of fertilization and in the early blastula stage (stage 6), and persisted during gastrulation and through the tailbud and tadpole stages (Fig. 3A) . 

XRPGR expression in adult tissues was examined in total RNA isolated from X. laevis heart, eye, kidney, brain, and liver by RT-PCR. XRPGR expression was readily detected in the eye, and was detectable in the liver, brain, heart, and kidney (Fig. 3B) . XRPGR protein expression in the retina was also identified by Western blotting, in which two anti-RPGR antibodies (1878 and C2) both recognized a single band of approximately 140 kDa, as predicted for the XRPGR^ORF15 isoform (Fig. 3C) . This indicated that both antibodies were suitable for analysis of X. laevis tissues. 

To determine the temporal and spatial expression pattern of XRPGR in the developing visual system, whole-mount in situ hybridization was carried out on X. laevis embryos at different developmental stages (Figs. 4A 4B) . At stage 18, XRPGR was detected in the whole anterior neural plate; by stage 22, XRPGR transcripts were expressed in the neural tube and developing optic vesicles. At stages 24 and 28, high-dorsal transcripts of XRPGR were detectable in the eye primordium and brain, and low-dorsal expression in the notochord. In the late-tailbud embryo (stage 32), XRPGR expression was more intense in the whole eye. At the tadpole stage, XRPGR expression persisted and intensified in the eye. To map XRPGR expression precisely at the tadpole stage, stage 40 embryos were sectioned after whole-mount in situ hybridization (Fig. 4B) . This showed XRPGR expression in the outer nuclear layer, inner nuclear layer, and ganglion cell layer. The apparent labeling of plexiform layers is thought to be due to diffusion of the visualization reagent from the cell bodies (Harris WA, personal communication, 2005). 

Localization of XRPGR Protein in X. laevis Photoreceptors

Immunofluorescent staining of XRPGR in photoreceptor cells on experimental slides is shown in Figure 5 . The rod photoreceptor OS showed continuous labeling along a streak resembling the connecting ciliary axoneme and punctate labeling along lines coinciding with the multiple incisures (Fig. 5A) . Double immunolabeling confirmed that the continuous streak of XRPGR fluorescence colocalized with α- or β-tubulin at the microtubules in the ciliary axonemes (which extended part of the way to the distal end of the rod OS but all the way to the tip of the cone OS), and that the punctate lines of XRPGR labeling colocalized with the microtubules at the rod OS incisures (Figs. 5A 5B) . XRPGR immunoreactivity within cone photoreceptors was also present in the inner segments (Fig. 5C) . Photoreceptor cells on control slides (not shown) showed no significant fluorescence, which, together with the specificity of the XRPGR^ORF15 antibodies, as shown by Western blot and immunolabeling of XTC-2 cells (see below), indicated that the fluorescence on experimental slides was specific for XRPGR and/or β-tubulin. 

Localization of XRPGR to Centrosomes in XTC-2 Cells

To identify the subcellular localization of native XRPGR, anti-bORF15^C2 and anti-hORF15¹⁸⁷⁸, which were raised against bovine and human RPGR^ORF15, respectively, ³² and which recognize a variety of other mammalian RPGRs, ^{25 32} were used in indirect immunocytofluorescence microscopy experiments in cultured XTC-2 cells, which are derived from X. laevis tadpoles. ³³ The cells were double-labeled with anti-bORF15^C2 or anti-hORF15¹⁸⁷⁸ and anti–γ-tubulin antibodies. XRPGR labeling was found to be coincident with the γ-tubulin signal at centrosomes at all stages of the cell cycle (Fig. 5) , as recently reported in a wide range of mammalian cell lines. ³²  

Discussion

The full-length protein sequence of the XRPGR^ex1–19 (previously referred to as the constitutive or default transcript ^{7 25} ) (727 amino acids) is shorter than its mammalian orthologues (815–1003 amino acids), but the RCC1-like domain is very similar (Fig. 1A) . The RCC1-like domain interacts with two rather different proteins: RPGRIP1, which colocalizes with RPGR at centrioles, basal bodies, and ciliary axonemes ^{32 34 35 36} ; and PDE6D, a 17-kDa prenyl-binding protein. ^{37 38} Eighteen (82%) of the 22 reported missense mutations in this domain, several of which have been shown to abolish or impair the interaction with RPGRIP1 or PDE6D, ^{34 37} were conserved across all five species examined (Fig. 2A) . The exceptions were D312Y and D312N, where an acidic residue (D,E) was present in all five species, suggesting that the change to a hydrophobic or neutral amino acid is deleterious; I289V, where either an isoleucine (three species) or a valine (two species) was present in all species, suggesting that this may not be disease-causing; and G436D, which was not conserved, but involved a major change of charge, potentially destabilizing the protein structure (Fig. 2A) . The predicted isoprenylation motif at the C-terminus of XRPGR^ex1–19 and the C-terminal exon of XRPGR^ORF15 were also conserved across species (Figs. 2B 2C) , supporting the view that these constitute major functional domains. 

RPGR expression has not previously been analyzed during development. RPGR is not essential either for retinal development or for photoreceptor differentiation in XLRP patients or in an Rpgr knockout mouse, ²³ in which photoreceptors show normal morphology and function at the completion of retinal development (∼30 postnatal days), and only later show structural abnormalities (such as abnormal basal discs) and functional abnormalities detectable by electroretinography. In the present study, XRPGR transcripts were detected by RT-PCR in oocytes and from the time of fertilization through the early developmental stages into the adult stage (Fig. 3) . XRPGR expression preceded the appearance of retinal cells, which differentiate at approximately stage 24. ³⁹ Similarly, in the adult, although XRPGR was strongly expressed in the eye, expression was also observed in the liver, brain, heart, and kidney (Fig. 3B) . These results suggest a more widespread role for XRPGR, such as in ciliary or axonemal function (see below). 

In contrast, the results using the less sensitive technique of whole-embryo in situ hybridization showed XRPGR expression first in the anterior neural plate (stage 18), which is the presumptive eye progenitor, then becoming increasingly concentrated in the eyes. In the differentiated retina, at the tadpole stage, XRPGR was expressed in the outer nuclear, inner nuclear, and ganglion cell layers (Fig. 4) . In both human and mouse retina, RPGR protein is most strongly localized to the connecting cilium of rod and cone photoreceptors, but in some species it has also been reported in the OS and in a restricted set of ganglion cells. ^{24 25 32} Consistent with this, XRPGR expression in adult X. laevis retina was strongest in photoreceptors: in rods it colocalized with microtubular proteins at ciliary axonemes and incisures, and in cones at ciliary axonemes (Fig. 5) . 

In the X. laevis XTC-2 cell line, XRPGR^ORF15 was detected in centrosomes at all stages of the cell cycle (Fig. 6) . Recent evidence has highlighted the localization of RPGR^ORF15 both to the basal bodies of ciliated cells and to the centriolar component of centrosomes in a wide range of nonciliated (actively dividing) mammalian cells. ³²  

These results are consistent with the proposal that XRPGR is important in the function of a single microtubular organelle, the paired centrioles or basal bodies, and the associated transitional zone of sensory and motile cilia. ³² The distinctive roles of the XRPGR^ex1–19 and XRPGR^ORF15 remain unclear: only RPGR^ex1–19 was detected in motile cilia of mouse trachea, but both isoforms were present in connecting cilia. ²⁵ The functions of primary (nonmotile) cilia are not fully elucidated, but there is increasing evidence that, as with sensory cilia, they also act as sensory organelles (with the exception of nodal cilia, which influence left-right asymmetry). ^{40 41} Primary cilia are present in virtually all cells, but are transiently lost (deciliated) when the centrioles separate during mitosis, to be reassembled in the G1 phase. ⁴¹ In early X. laevis development, XRPGR expression, detectable by RT-PCR, may reflect its association with such primary cilia, from fertilization through cleavage and subsequent embryonic stages. In whole-mount embryos, higher-level expression of XRPGR, first detectable in the anterior neural plate and developing optic vesicle, is increasingly localized to the eye. Its presence in the eye was associated with expression in all three neuronal layers of the retina, although in adult X. laevis retina, it was most evident in the rod and cone photoreceptors. Here, XRPGR localized to the axoneme of the connecting cilium, which, although similar to a primary cilium, appears to be an extended transitional zone. In motile cilia, the transitional zone lies between the distal axoneme (containing nine peripheral and one central microtubular doublets) and the basal body (containing nine microtubular triplets and no central doublet). ^{25 42} The transitional zone contains microtubules that are less stable than in the outer axonemal or basal body regions and is the site where deciliation normally occurs, by disorganization of microtubules and constriction of the ciliary membrane, ^{43 44} perhaps suggesting a vulnerability that may be important in RPGR-related retinal degenerations. This would be consistent with the observation that in XLRP, the connecting cilia of rods and cones are initially structurally normal, ⁴⁵ indicating a more specialized function than maintenance of ciliary structure. 

 Supported by the Medical Research Council, the British Retinitis Pigmentosa Society (AFW), and the Wellcome Trust (DCG, CS).

 Submitted for publication July 5, 2005; revised September 6, 2005; accepted November 18, 2005.

 Disclosure: X. Shu, None; Z. Zeng, None; M.S. Eckmiller, None; P. Gautier, None; D. Vlachantoni, None; F.D.C. Manson, None; B. Tulloch, None; C. Sharpe, None; D.C. Gorecki, None; A.F. Wright, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Primers Used for Cloning Xenopus laevis XRPGR

Primer Name	Oligonucleotide Sequence	Accession Number	Position (Nucleotide)
XRPGR N1	5′-GGAAAAAGTAAGTTTGCAGAAAAC-3′	AL849961	77–100
XRPGR C1	5′-GTCACAAAGGCTGAATGGTAATA-3′	AL849961	620–642
XRPGR N2	5′-GGACTTCTCTATACATTTGGAGAT-3′	AL857113	18–41
XRPGR C2	5′-GCTCATTGCATGTGTCATGTTTAA-3′	AL857113	603–626
XRPGR N3	5′-CATGCAAATCTGCATTCCATTCA-3′	X. tropicalis scaffold_83	23,150–23,172
XRPGR C3	5′-CTTTAAATTCAAGTAATGAGGTAA-3′	X. tropicalis scaffold_83	24,720–24,743
XR5RaceR1	5′-ACACAGGTTGGCTTGCTGATAG-3′	DQ175998	624–645
XR5RaceR2	5′-CACAGTGTGTTCATCCCCACATGA-3′	DQ175998	524–547
XR3RaceR1	5′-GAAATGTCTACATCATTGCAA-3′	DQ175998	1586–1606
XR3RaceR2	5′-AGCTTTTCTGCGAATGATACTGAT-3′	DQ175998	1862–1885
Xrseq N1	5′-CTTAAACATGACACATGCAATGAG-3′	DQ175998	1927–1950
Xrseq C1	5′-ATGCCTCCCATCTCCAAATGTATA-3′	DQ175998	1352–1375
Xrseq C2	5′-ATTTCTCAAAGGTGGTAACGTTCT-3′	DQ175998	1676–1699
ODC N1	5′-CAGCTAGCTGTGGTGTGG-3′	BC044004	657–674
ODC C1	5′-CAACATGGAAACTCACACC-3′	BC044004	866–884

Figure 1.

 

(A) Schematic representation of the domain structures of RPGR^ex1–19 and RPGR^ORF15 isoforms, showing their common RCC1-like domain (shaded); a C-terminal domain of unknown function unique to RPGR^ex1–19 (diagonal cross-hatch); a C-terminal isoprenylation site unique to RPGR^ex1–19; and the C-terminal ORF15 domain of RPGR^ORF15, containing a repetitive glycine-glutamic acid–rich region, which is a mutational hot spot. (B) Diagram of the method of cloning the X. laevis XRPGR gene, showing the XRPGR cDNA sequence of X. laevis (thick central line), with the start (ATG) and stop codons (TGA) and two X. tropicalis EST clones used for PCR amplification of XRPGR; and the resultant PCR products (thin lines) and primers (arrows), with the primers used in 5′- and 3′-RACE reactions to extend the cDNA in each direction. GE, glycine-glutamic acid.

Figure 2.

 

Conservation of the RPGR RCC1-like domain across human, dog, mouse, X. laevis (Xela), and X. tropicalis (Xetro) species. (A) The amino acid sequence of each species below residues 10 to 447 of human RPGR^ex1–19, corresponding to the RCC1-like domain, after alignment. Identical residues are boxed (black). Accession numbers for RPGR^ex1–19: human, NM_000328; dog, AF148801; mouse, NM_011285; Xela (current study), DQ175998; Xetro (current study), DQ17599. (B) Conservation of the C-terminal of RPGR^ex1–19 in five species. The sequences are aligned with the C-terminal 38 amino acids of human RPGR^ex1–19, showing the conserved isoprenylation site. (C) The C-terminal of Xenopus RPGR^ORF15 is also highly conserved in eight species. Accession numbers for RPGR^ORF15: Xenopus (Xela; current study), DQ176000; human, AF286472; mouse, AF286473; bovine, AF286474; dog, AF385629; pig, AY855167; sheep, AY855169; fugu, AF286475.

Figure 3.

 

(A, upper): Temporal expression of XRPGR by RT-PCR from total RNA extracted from oocytes and at different developmental stages (stages 1 to 40) of X. laevis development. Lower: RT-PCR with ornithine decarboxylase (ODC) primers (Table 1) , used as a control. Both: RNA samples without reverse transcriptase were used as negative controls (−) for PCR in each experiment; all were blank. Band size: RPGR, 567 nucleotides; ODC, 228 nucleotides. (B, upper): XRPGR expression in different X. laevis tissues by RT-PCR. Lower: ODC expression by RT-PCR as control. Band size: RPGR, 567 nucleotides; ODC, 228 nucleotides. (C) Analysis of XRPGR protein expression in X. laevis retina. Retinal extract was electrophoresed in 10% SDS-PAGE, transferred to nitrocellulose membrane, and probed with anti-RPGR^ORF15 (C2) antibody using preimmune IgG as a control. Band size: ∼140 kDa.

Figure 4.

 

XRPGR expression in the developing X. laevis embryo. (A) Whole-mount in situ hybridization of embryos at stages 18, 22, 24, 28, 32, 36, and 40, probed with the RCC1-like domain of XRPGR cDNA (antisense and sense RNA probes). (B) Sections of stage 40 embryos after in situ hybridization, showing XRPGR mRNA localized in the outer nuclear layer, inner nuclear layer, and ganglion cell layer using antisense probe. No signals were detected using sense probe. Magnification: left, ×20; right, ×40.

Figure 5.

 

XRPGR^ORF15 localization in dissociated X. laevis photoreceptor cells. (A) Individual rod outer segments (ROS) in bright field (left) and labeled with anti-hRPGR^ORF15 (1878) antibody ³² (right), showing strong labeling of the axoneme at the base of the ROS and weaker, punctate labeling of the incisures. (B) ROS in bright field (left) and labeled with anti-RPGR^ORF15 (center) and anti–β-tubulin antibodies (right), showing labeling of the connecting ciliary axoneme (arrows) and labeling along the multiple incisures. (C) Cone with the OS attached to the inner segment, in bright field (left) and labeled with anti-RPGR^ORF15 (center) and anti–β-tubulin antibodies (right), showing continuous labeling of the ciliary axoneme, which extends to the tip of the OS; RPGR immunoreactivity is also present in photoreceptor inner segments.

Figure 6.

 

XRPGR^ORF15 localization in X. laevis XTC-2 cells. Antibodies directed against RPGR^ORF15 (green) colocalize with antibodies to the centrosome-specific component, γ-tubulin (red). XRPGR^ORF15 is present at the centrosome at all stages of mitosis in XTC-2 cells.

The authors thank Bill Harris for advice and Sandy Bruce for the artwork.