Identification of an Intronic Single-Point Mutation in RP2 as the Cause of Semidominant X-linked Retinitis Pigmentosa

Abstract

Purpose.: A large family with 11 males and 2 females with X-linked retinitis pigmentosa (XLRP) was analyzed in search of pathologic mutations.

Methods.: Of the two major XLRP genes, RPGR was analyzed by SNP cosegregation and RP2 was directly screened for mutations. The pathogenicity of a new variant was assessed in silico, in vivo, and in vitro.

Results.: The results of cosegregation analysis with SNPs closely located to RPGR excluded this gene as the cause of the disease in this family. Sequencing of RP2 showed a putative pathogenic variant in intron 3 at the conserved polypyrimidine tract (c.1073-9T>A). This substitution cosegregated with the disease and was not found in 220 control chromosomes. In silico analyses using online resources indicated a decreased score of intron 3 acceptor splice site for the mutated sequence. Real-time RT-PCR analysis of the RP2 splicing pattern in blood samples of patients and carrier females showed skipping of exon 4, causing a frame shift that introduced a premature stop codon. Further verification of the pathogenicity of this point mutation was obtained by expression of a minigene RP2 construct in cultured cells.

Conclusions.: A transversion (T>A) at position −9 in intron 3 of RP2 causes XLRP by altering the splicing pattern and highlights the pathogenicity of intronic variants. The single point RP2 mutation leads to a wide range of phenotypic traits in carrier females, from completely normal to severe retinal degeneration, thus supporting that RP2 is also a candidate for semidominance in XLRP.

Retinitis pigmentosa (RP) is a hereditary retinal disorder characterized by night blindness, peripheral constriction of visual fields and pigment spicule deposits in the midperiphery of the retina. RP is the most prevalent genetic cause of blindness in adults, affecting 1:3000 to 4000 individuals worldwide. ^1,2 It is a highly heterogeneous disease at the clinical and genetic level. To date, more than 35 genes have been described, showing autosomal dominant, autosomal recessive, and X-linked (Retnet, http://www.sph.uth.tmc.edu/Retnet/ provided in the public domain by the University of Texas Houston Health Science Center, Houston, TX) inheritance patterns. The X-linked forms (XLRP) are associated with the most severe phenotypes and account for approximately 9% to 10% of all the RP cases. Six loci have been mapped to the X chromosome, and only two major genes have been identified: RPGR (70%–80%) and RP2 (7%–20% of the mutated alleles). ^3,4  

The RPGR gene stretches at least 23 exons and produces many proteins through a complex pattern of alternative splicing. ^5,6 This gene contains a mutational hotspot, ORF15, that presents a long purine-rich domain, full of low-complexity repeats. Most mutated alleles are insertions or deletions within these repeats. ^5,7  

The RP2 gene encompasses 5 exons, generates a 350 amino acid protein, and shows a wide pattern of expression. ⁸ Although its function has not been completely elucidated, recent reports associate RP2 with protein complexes involved in the photoreceptor ciliary transport, as well as in the interaction between plasma membrane and cytoskeleton. ^9,10  

We report the analysis of a large XLRP pedigree with a wide range of phenotypic variability in female carriers. We analyzed both RPGR and RP2 and identified a pathogenic point mutation in intron 3 of RP2, c.1073-9T>A. This transversion caused a shift in the polypyrimidine stretch of the acceptor splicing site, decreasing recognition by the splicing machinery and causing the skipping of exon 4. Real-time RT-PCR, used to study the splicing pattern of RP2 in blood samples, showed that the mutated allele favored the aberrant over the wild-type transcript. We further confirmed the pathogenicity of this allele by analyzing the splicing pattern in a minigene construct. Notably, the high levels of variation observed for RP2-spliced products in female carriers supports X chromosome–skewed inactivation as the cause of the disease in the affected females. 

Material and Methods

Clinical Assessment of Patients

Our study was based on a large Spanish pedigree with branches currently living in three different countries (Spain, Italy, and Switzerland; see Fig. 1). All affected members had received the RP diagnosis several years ago through ophthalmic examination at the Hospital Universitario Central de Asturias (Oviedo, Spain). In this study, detailed clinical evaluation was obtained for patients IV14, IV16, and V.8, as well as the nonaffected obligate female carriers III.16 and IV.11, which included best corrected visual acuity and slit lamp biomicroscopy, followed by indirect ophthalmoscopy and fundus photography after pupillary dilation. The size and the extent of the visual-field defects were assessed with static perimetry (Humphrey; Carl Zeiss Meditec, Oberkochen, Germany). Electroretinograms (ERGs) were recorded in accordance with the protocol of the International Society for Clinical Electrophysiology of Vision (ISCEV) at the Hospital San Agustín (Avilés, Spain) (Table 1, and see Fig. 2). 

Table 1.

 

Clinical Characteristics of Carriers III.16 and IV.11 and Patients IV.14, IV.16, and V.8

Individual and Sex	Age* (y)	RP Symptoms	Age of Onset	Progression	Visual Acuity (OD/OS)	Refraction (OD/OS)	Visual Field	Fundus	ERG
III.16/female†	60	No	—	—	1.0/1.0	90°−0.75/+2.00	Normal	Normal	Normal responses OU
						90°−0.75/+2.25
IV.11/female†	41	No	—	—	0.8/0.9	10°−1.50/+150	Normal	Normal	Mildly altered cone response OU
						155°−1.00/+1.50
IV.14/female‡	29	Yes	22	Moderate	0.3/0.2	−1.5/−2.5	Highly altered OU	RPE atrophy, vascular attenuation, bone spicules	Highly altered OU
IV.16/female‡	30	Yes	24	Mild	0.1/0.7	5°−2.50/−7.00	Highly altered OD/altered OS	RPE atrophy, some vascular attenuation	Highly altered OD/altered OS
						170°−3.75/−4.50
V.8/male	13	Yes	5	Severe	0.3/0.1	−3.50/−5.50	Highly altered OU	General RPE atrophy, vascular attenuation, papillary drusen, and macular atrophy OU	Highly altered OU

*  Age at the time of the study.

†  Obligate female carrier.

‡  Affected female.

DNA Purification from Blood Samples

DNA was obtained from blood samples of patients III.14, IV.7, IV.12, IV.14, IV.16, V.8, and V.9; obligate carrier females II.3, III.6, III.16, IV.11; and nonaffected males and females IV.6, V.7). Informed consent from all the family members was obtained, according to the tenets of the Declaration of Helsinki. 

DNA from 110 matched control individuals of Spanish population was obtained from whole blood (Wizard Genomic DNA purification kit; Promega, Madison, WI). 

SNP Cosegregation RPGR Analysis

Seven highly informative SNP markers (heterozygosity values higher than 0.3 according to HapMap, http://www.hapmap.org/ provided in the public domain by the International HapMap Project, National Institutes of Health, Bethesda, MD), closely flanking the RPGR gene at 5′ and 3′ (Table 2), were genotyped by direct sequencing (BigDye ver. 3.1 kit in the Prism 3730 DNA sequencer; Applied Biosystems, Inc., [ABI], Carlsbad, CA) after genomic DNA PCR amplification. Haplotypes for cosegregation studies were constructed manually. 

Table 2.

 

SNPs for Cosegregation Analysis of the RPGR Gene

SNP	Distance (kb)*	Chromosome Position†
rs5964274	−334.7	37,737,017
rs2236153	−166.9	37,904,807
rs2268790	−164.5	37,907,201
RPGR		38,071,732–38,013,367
rs5917608	345.6	38,359,027
rs12839987	398.4	38,411,793

*  Negative and positive values indicate, respectively, upstream and downstream locations referring to the RPGR gene.

†  According to the NCBI reference values at Map Viewer (http://www.ncbi.nlm.nih.gov/SNP/, National Center for Biotechnology Information, Bethesda, MD).

Direct Mutational Screening of RP2

Five pairs of primers (Table 3) allowed the PCR amplification of RP2 exons plus adjacent intronic sequences in the studied family members. All the fragments were sequenced (BigDye v3.1 kit; Prism 3730 DNA sequencer; ABI). 

Table 3.

 

Primers Used for the Amplification and Mutational Screening of RP2 Exons

Exon	Primer Sequence (5′→3′)
	Forward	Reverse
1	AGGGTTCACGCCACACTCTAGG	AGCTATCCGCGTTCAAGAGTG
2	TGCCTGGCAGCCAATAGTCC	ACTTGAGGGTTGCCTGTATTTCCA
3	AATCAGTGTGCTGTTGTTGCATTT	GTCTAGGGCTCCTTGAGTGATGTG
4	GATATGTCCCATCTGTCTGC	CACACCCCAAAAATTCCAAGC
5	GGAGAACATGGGCTTTGGC	AAATTCTATATTCACAAGTTGGGAAAGG

Bioinformatic Analysis

All the sequences were analyzed using sequence assembly software (Seqman; DNAStar, Madison, WI) software and aligned to the genomic wild-type RP2 sequence from the public database UCSC Genome Browser (http://genome.ucsc.edu/ provided in the public domain by UCSC Genome Bioinformatics, University of California at Santa Cruz, Santa Cruz, CA). The alignments were verified de visu. 

SNP databases consulted were: dbSNP as implemented in the UCSC Genome Browser dbSNP of the NCBI server (http://www.ncbi.nlm.nih.gov/ provided in the public domain by the National Center for Biotechnology Information, Bethesda, MDS) and the mutation/SNPs databases for retinal disease genes reported in Retina International (http://www.retina-international.org/scinews/database.htm/ provided in the public domain by Retinal International, Zurich, Switzerland) and Retinal Information Network (http://www.sph.uth.tmc.edu/Retnet/). 

Splicing site score values of the wild-type and variant sequences of RP2 were predicted online at SplicePort (http://spliceport.cs.umd.edu/ provided in the public domain by the University of Maryland, College Park, MD), Geneid (http://genome.imim.es/genepredictions/index.html/ provided in the public domain by the Genome Bioinformatics Research Lab, Center for Genomic Regulation, Spanish Bioinformatics Institute, Barcelona, Spain) and BDGP (http://www.fruitfly.org/ Provided in the public domain by the Berkeley Drosophila Genome Project, University of California at Berkeley) (Table 4). 

Table 4.

 

In Silico Prediction Score Values of Intron 3 Acceptor Splice Site of the RP2 Wild-Type and Mutant Alleles

Splicing Prediction Algorithm	WT	MUT	Threshold*
SplicePort	1.13	0.03	0
Geneid	4.07	−0.15	—
BDGP	0.95	0.76	0.4

*  The higher the value, the higher the probability of being a splicing site in vivo. Threshold values for each program are also provided. Negative values or those close to 0 mean that the acceptor splice site was not detectable.

RT-PCR Analyses of the RP2 Splicing Pattern

Total RNA from patients and carriers III.14, III.16, IV.11, IV.14, IV.16, V.8 and V.9 (Fig. 1) were obtained after 300 μL of blood was processed (previously stabilized with RNAlater; Ambion, Austin, TX) with a purification kit (RiboPure-Blood; Ambion), according to the manufacturer's instructions. First, cDNA chains were obtained by reverse transcription (Cells-to-cDNA kit; Ambion), with MMLV reverse transcriptase using 0.62 μM oligo d(T) and 1.25 μM random decamers, for 15 minutes at 37°C, 15 minutes at 39°C and finally, 45 minutes at 42°C. For RT-PCR of transfected cells, the first cDNA chain was obtained from total RNA after lysis of cells (Cells-to-cDNA kit; Ambion) and directly proceeding to perform the reverse transcription, as stated earlier. 

Specific amplification of transcripts from endogenous blood RP2, mini RP2 gene and GAPDH was performed using the corresponding primers: RP2f: 5′-GACAGAAGAGC AGCGATGAAT-3′ (in exon 2); RP2r: 5′-CATATTCCCATCTGTATATCAGC-3′ (in exon 5); GFPf: 5′-CGAGCTGTACAAGTCCGGCC-3′ (in the C terminus of GFP); GFPr: 5′-GGTTCAGGGGGAGGTGTGGG-3′ (in the 3′UTR of the pEGFP-C2 vector); and GAPDHf: 5′-TGAAGGTCGGTGTGAACGGATTTGG-3′ and GAPDHr: 5′-CATGTAGGCCATGAGGTCCACCAC-3′. 

The PCR was performed in a final volume of 25 μL (GoTaq Flexi DNA polymerase; Promega) and three different sets of conditions. For amplification of GAPDH, a two-step PCR was performed as follows: first denaturation for 2 minutes at 94°C, followed by 35 cycles of 20 seconds at 94°C and 2 minutes at 63°C. For amplification of endogenous RP2 from blood, we performed a 3-step PCR: denaturation for 2 minutes at 94°C, followed by 35 cycles of 20 seconds at 94°C, 30 seconds at 58°C, and 25 seconds at 72°C. For amplification of the EGFP-RP2 minigene, a 3-step PCR was performed: denaturation for 2 minutes at 94°C, followed by 28 cycles of 20 seconds at 94°C, 30 seconds at 60°C, and 30 seconds at 72°C. 

Real-time PCR was performed with SYBR green master mix (The SYBR Green I Master assay in a LightCycler 480; Roche, Indianapolis, IN). Primers that detected all types of RP2 transcripts were used as an internal control: 5′-GGGGTCAGAGACAGAAGAGC-3′ (forward, in exon 2) and 5′-CTTTGTCTGAACTAGGAAAAAGC-3′ (reverse, in exon 3). Specific primers for the detection of either the wild-type or the aberrant transcripts were also designed. For the wild-type transcript, the primers were 5′-CCAGAAAACTAATTGATGAGATGG-3′ (forward, in exon 2/exon 3 junction) and 5′- GCAATAACAGGACCTTTGTTCAG-3′ (reverse, in exon 3/exon 4 junction). Primers for the aberrant transcript were 5′-CCAGAAAACTAATTGATGAGATGG-3′ (forward, in exon 2/exon 3 junction) and 5′-TTTCAGATACAAACATCTTTGTTCAG-3′ (reverse, in exon 3/exon 5 junction). For each real-time PCR reaction, 1 μL of cDNA template, 1 μM of each primer, 1× master mix, and 2 μL of water were mixed in a final volume of 10 μL. The reaction was set at 50 cycles in a two-step program: 95°C for 10 seconds and 58°C (wild-type and aberrant transcript detection) or 62°C (internal control) for 45 seconds. All samples were run in triplicate, with their corresponding negative controls. The level of transcript expression was quantified by using the second-derivative maximum method (LightCycler 480 software, ver. 1.5.0; Roche). 

Construction of the RP2 Minigene

Taking into account that the genomic sequences relevant for splicing are contained within the 100 to 200 bases near the exon–intron junctions, exon 3 plus downstream intron 3 sequence, exon 4 plus its flanking intron 3 and 4 sequences, and exon 5 plus upstream intron 4 sequence were amplified from the genomic DNA of a control individual and an affected male of the family (Table 5, see Fig. 6A). These fragments were cloned in frame into the pEGFP-C2 vector (Clontech-BD Biosciences, Palo Alto, CA). The primers used contained restriction sites to allow the cloning of the PCR product and are shown in Table 5. The exons were sequentially cloned in the corresponding order and in-frame after the GFP sequence between the BglII and BamHI sites of the pEGFP-C2 vector (Clontech-BD). The forward primer of exon 3 contained a BglII site and the reverse primer (within intron 3) contained a HindIII site. Cloning of exon 4 was achieved through digestion of the amplified fragment with ScaI (within intron 3) and the EcoRI site introduced in the reverse primer (complementary to a sequence in intron 4). Finally, cloning of exon 5 was performed by digestion with an endogenous EcoRI (in intron 4) and the BamHI site introduced in the reverse primer. 

Table 5.

 

Primers Used to Amplify Exons Plus Flanking Intronic Sequences of the Wild-Type and Mutant RP2 Alleles

Exon	Product Size (bp)	Primer Sequence (5′→3′)*
		Forward	Reverse
3	440	GAAGATCTTGATGGTTGGTAAAGGCTTTTC	CCCAAGCTTATCACTACACTATAAAGGTAC
4	516	TGTGACACTAGTCCTATAATAG	GGAATTCCCAATTATTTATCAAAGGTTC
5	340	GACCTTAATTATATAAGTGCTTC	CCAAGTCCTGGATCCACATTGC

*  Restriction site targets introduced to allow sequential cloning of each fragment are underlined.

Cell Culture and Transfection

HEK293T cells were seeded on 24-well plates (2 × 10⁵ cells/well) and grown in DMEM (Invitrogen, Barcelona, Spain) supplemented with 10% of fetal bovine serum. After 12 hours, the cells were transiently transfected with constructs bearing either the pEGFP-wtRP2 minigene, the pEGFP-mutRP2 minigene (containing the c.1073-9T>A mutation), or the empty pEGFP vector (Clontech-BD), with transfection reagent (Lipofectamine 2000; Invitrogen), according to the manufacturer's instructions. Forty-eight hours after transfection, the cells were collected and lysed, and total mRNA was used for RT-PCR. Transfection efficiency was monitored by checking and counting GFP-fluorescent cells under an optic fluorescence microscope (DMIL; Leica, Wetzlar, Germany) and for all the constructs and replicates, the efficiency was comparable. 

Results

Clinical Assessment and Findings

The RP in the family showed an X-linked inheritance pattern (Fig. 1). The clinical evaluations performed for one male (V.8) and two female (IV.14 and IV.16) patients as well as two obligate female carriers (III.16 and IV.11) are shown in Table 1. Night blindness and visual acuity loss increased with age in the affected members. Reduction of visual fields was more prominent in the affected male (V.8) and in the most affected female (IV.14). When comparing the ERG recordings, the male (V.8) and the most affected female (IV.14) showed nondetectable rod and cone waves and no response to flicker light (Table 1). Her sister, also affected (IV.16), showed asymmetrical retinal alteration, much more pronounced in the right eye, compatible with the fundus and visual acuity data. The ERG of the latter showed rod waves with decreased amplitudes and normal latency, cone waves of decreased amplitude and mildly increased latency, and flicker response with decreased amplitude. 

Figure 1.

 

Spanish XLRP pedigree spanning five generations. Note that obligate carrier females are nonaffected except for females IV.14 and IV.16.

The fundus of the eyes in three patients showed extensive atrophy of the RPE, bone spicule pigment deposits and vascular attenuation (Figs. 2A–C), as apparent in the male (Fig. 2A) as in the female patients (particularly in IV.14; Fig. 2B), in clear contrast to the normal fundus of the nonaffected obligate female carrier IV.11 (Fig. 2D). In addition, the eye fundus of the male patient showed macular atrophy and pseudopapilledema owing to the existence of papillary drusen (Fig. 2A), traits that have also been associated with RP. ^11,12  

Figure 2.

 

Fundus eye photographs from several affected and nonaffected mutation c.1073-9T>A carriers. (A) Affected male V.8. (B) Affected heterozygous female IV.14; the view of the left eye fundus is peripheral, to show retinal alterations. (C) Affected heterozygous female IV.16. (D) Nonaffected heterozygous female IV.11 with a normal fundus. Note the RP hallmarks in the affected fundi: pigmented epithelium atrophy (e.g., B, C), peripheral bone spicule pigment deposits (particularly prominent in the left eye, B), and reduction in the number and caliber of blood vessels (vascular attenuation). Macular atrophy and papillary drusen were also visible in both eyes (A). LE, left eye; RE, right eye.

In accordance with the best corrected visual acuity, visual field, ERG, and fundus examination of the five individuals studied, we concluded that two obligate female carriers were nonaffected, whereas one male and two female patients showed variable degrees of affectation. According to the dates of clinical examination (Table 1), the most affected member was the male V.8, with very low visual acuity, highly altered ERG in both eyes, and general atrophy of the RPE and macula. Of the two female patients, the most affected was patient IV.14, who showed low visual acuity and nondetectable rod and cone waves in the ERG. Finally, female patient IV.16 presented asymmetrical atrophy of the RPE, being the right eye the most affected, with less visual acuity and more altered visual field and ERG. 

Results of Cosegregation Analysis of RPGR

In our search for the genetic cause of RP in the family, we focused in RPGR as the major gene for X-linked forms. Given the size and complexity of the RPGR coding sequence, cosegregation analysis was performed in three affected (III.14, IV7, and V.8) and six nonaffected (II.3, III.6, III.16, IV.6, IV.11 and V.7) members with seven highly informative SNP markers very close to the gene (Table 2). The haplotype analysis showed lack of cosegregation with the disease, thus ruling out RPGR as the causative gene (Fig. 3). 

Figure 3.

 

Haplotype cosegregation analysis of SNP markers excluded RPGR as the causative gene. Only a partial view of the pedigree is shown. Generations and individuals are numbered according to Figure 1. The position of the SNPs used is detailed in Table 2.

Identification of a New Mutation in Intron 3 of RP2

Once RPGR had been excluded, we focused on RP2. The five exons of this gene were sequenced in patients (males IV.7 and V.8). A single sequence variation was identified in intron 3, relatively close to the AG acceptor-splicing site (Fig. 4). This substitution caused a transversion, c.1073-9T>A, in the conserved polypyrimidine tract (PPT; 15–40 pyrimidines) required for the splicing process. ¹³ This variation cosegregated with the disease in this family, and in addition, an exhaustive search in SNP databases did not show any previous report assignment. 

Figure 4.

 

Identification of the mutation c.1073-9T>A. Chromatograms of the RP2 intron 3/exon 4 junction sequence of one control male individual and the affected male V.8, showing the transversion T>A in position −9 from the first nucleotide of exon 4.

To verify that this intronic variation was not a rare polymorphism, we analyzed 220 chromosomes of a matched control Spanish population. None of them contained this variant allele, suggesting that it was the cause of RP in this family. 

Effect of Mutation c.1073-9T>A on the Strength of the Intron 3 Acceptor Site

Pathogenic mutations in introns mainly affect splicing events, either by removing or altering the recognition sites at the junctions (dinucleotides, PPT, enhancers/silencers) or generating new cryptic exons. ^14,15 The requirement of a PPT for recognition of intronic acceptor sites has long been established. Although there is not a clear consensus sequence, the number and type of pyrimidines is relevant, with a strong preference for Us over Cs. ¹³ Considering the wild-type RP2 sequence, the 15 bases upstream of the intron 3–exon 4 junction yielded 53.3% of the pyrimidines, lower than the 62.3% average in the other RP2 junctions. The identified single-point mutation decreased the percentage of pyrimidines below 50% (46.7%), suggesting a negative effect on this acceptor site recognition. 

Using to online resources, we evaluated the presumptive effect of the mutation c.1073-9T>A on the genomic sequence environment for the 5′ and 3′ splice sites. The wild-type and mutant RP2 sequences (from 100 nt upstream the first exon to 100 nt downstream the last exon, approximately 45 kb of genomic sequence) were submitted to three different splicing prediction algorithms: SplicePort, ¹⁶ Geneid, ¹⁷ and BDGP. ¹⁸ Of interest, the variant sequence substantially decreased the recognition of the intron 3 acceptor site (Table 4). 

Aberrant RP2 Transcript in Affected and Carrier Family Members

After the in silico results, the most probable outcome of the RP2 mutation would be skipping of exon 4, generating a frame shift and a premature stop codon in the mature mRNA. Most mRNAs containing premature stop codons are degraded by the nonsense-mediated decay (NMD) mRNA surveillance mechanism. ¹⁹ However, in the present case, the mutated RP2 transcript would avoid NMD degradation as the premature STOP codon is located in the last coding exon (exon 5). 

Total RNA from the blood of several affected males (IV.12, V.8, V.9) and females (IV.14, IV.16) and obligate carrier females (III.16, IV.11) was used for RT-PCR analysis of RP2. GAPDH was tested as the control. When we used primers RP2f and RP2r (see the Material and Methods section) for the amplification of the RP2 transcript, the control male sample (WT) produced a single band of 370 bp (Fig. 5A), as would be expected from a correctly processed RP2 transcript. In contrast, affected males showed a faint band of the corresponding wild-type size, and another of 284 bp, very prominent and lacking exon 4. Concerning affected and nonaffected carrier females, different levels of wild-type and aberrantly processed transcripts were identified. Sequencing all the amplified products confirmed that the 370-bp band contained the wild-type product, including exon 4, whereas the 284-bp band corresponded to the directly fused 3 and 5 exons (Fig. 5B). We also verified that the faint band of a slightly higher size was the heteroduplex structure. 

Figure 5.

 

RT-PCR analysis of RP2 mRNAs in blood of affected males, females, and obligate female carriers. (A) RT-PCR of the RP2 spliced transcripts showed that affected individuals, all bearing the mutation c.1073-9T>A, produced high levels of an aberrantly spliced transcript (284-bp band), whereas the control wild-type produced only the expected 370-bp band. Numbering of the individuals is according to Figure 1. Note that carrier females (affected or nonaffected) showed variable levels of the wild-type and aberrant spliced products. The GAPDH RT-PCR was used as the control for normalization. (B) Sequencing of the two PCR bands showed unequivocal assignment of the 370-bp band to the wild-type product (fused exons 3 to 4–5) and the 284-bp band to the aberrant transcript that skipped exon 4. Faint higher molecular bands correspond to heteroduplexes. (C) Quantification of the two RP2 splicings (WT versus aberrant) in affected males and females and two nonaffected obligate female carriers by real-time RT-PCR. The wild-type male control produced only the correctly processed transcript, and it is considered 100%. Note that affected males produced less than 5% of the correct mRNA and more than 95% of the aberrantly processed transcript, whereas all carrier females (either affected or nonaffected) showed higher variation in levels of aberrant transcript expression, ranging from a mere 8% (III.16) to 90% (IV.16). m, male; f, female; NA, nonaffected; A, affected.

Quantification of each type of RP2 transcript by real-time RT-PCR in patients and female carriers showed that affected males (hemizygous for the mutated allele) produced ∼95% to 99% of the aberrant transcripts (Fig. 5C). Thus, the mutation strongly impaired recognition of the acceptor site at intron 3. Carrier females (heterozygous for the mutated allele) presented highly variable amounts of the aberrant RP2 transcript, ranging from barely 8% to nearly 90% (Fig. 5C, e.g., compare females III.16 and IV.16). 

In Vitro Splicing Assays with an RP2–Derived Minigene to Determine the Pathogenicity of the c.1073-9T>A Mutation

To assess whether the single-point mutation was sufficient for the skipping of exon 4, a minigene construct was generated and tested in transiently transfected cultured cells. HEK293T cells were transfected with the wild-type, the mutant allele, or the empty pEGFP-C2 vector (negative control). To avoid amplifying the endogenous RP2 gene, we designed the primers for the RT-PCR from the vector sequence and amplified exclusively the transcripts produced by the minigene. Three independent replicates were tested, and GAPDH was used as a control for RNA quantity and quality. The wild-type sequence produced only a 410-bp band (fusion of exons 3 to 4–5), whereas the mutated construct mainly produced the band expected for exon 4 skipping (fusion of exons 3–5) and only a very faint band of the wild-type splicing was detected under high exposure (Fig. 6B). Other faint bands are nonspecific. The identity of all the bands was verified by sequencing. 

Figure 6.

 

Construction and expression analysis of an RP2 minigene. (A) Diagram showing the wild-type and mutant RP2 minigenes fused to the GFP coding sequence. The position of the mutation is indicated in bold. (B) RT-PCR of cells transfected with either the wild-type or mutant RP2 minigenes showed that the mutation c.1073-9T>A was sufficient to produce a spliced transcript that skipped exon 4 (324-bp band), whereas the wild-type sequence produced a 410-bp band. *A nonspecific band. Lane 1: negative PCR control; lane 2: RT-PCR from nontransfected HEK293T cells; lane 3: negative RT-PCR control; lane 4: RT-PCR from HEK293T transfected with the pEGFP empty vector; lane 5: negative RT-PCR control from HEK293T cells transfected with the wild-type RP2 minigene; lanes 6, 7, and 8: three independent replicates of HEK293T cells transfected with the wild-type RP2 minigene; lane 9: negative RT-PCR control from HEK293T cells transfected with the mutant RP2 minigene; lanes 10, 11, and 12: three independent replicates of HEK293T cells transfected with the mutant RP2 minigene. GAPDH was used as the control for normalization.

Discussion

Two genes, RPGR and RP2, have been reported to cause XLRP. Most XLRP mutations are located in RPGR and mainly cluster at ORF15, an exon comprising 1.5 kb of very low complexity, purine-rich sequence and small repeats. The following priorities for XLRP diagnosis have been suggested: first, perform the mutational screening of ORF15; second, sequence RP2; and finally, analyze the remaining RPGR exons. ⁴ We propose to further refine this strategy by adding a prior step of cosegregation analysis based on SNP markers to either include or exclude RPGR as the candidate gene in a particular family. In the present case, this simple step allowed us to rule out RPGR as the causative gene, and therefore, focus on RP2. 

Although most reported mutations affect coding sequences, there is an increasing number of new pathogenic variants located in introns. The development of new algorithms combined with in vitro studies has facilitated the assessment of their pathogenicity. ²⁰ The more obvious mutations destroy the highly conserved the donor and acceptor junction sites (GT-AG), but attention is now being drawn to mutations deep within the introns that cause the inclusion of cryptic exons ^21,22 or in flanking intronic sequences that affect the recognition and binding of the splicing ribonucleoprotein (RNP) complexes. ^23–26 This is the case of the new mutation we report herein, which lowers below 50% the amount of pyrimidines in the required PPT, thus disrupting the recognition and binding of the U2-RNP complex. ²⁷  

The mutation c.1073-9T>A produces the skipping of exon 4 in the mature mRNA, causing a frame shift and a prematurely truncated protein (294+4 amino acids versus 350 aa of the wild-type protein). This aberrant transcript was abundantly found in male patients (more than 95%) as it escaped the NMD mRNA control mechanism. The truncated RP2 contains the N-terminal domain required for (1) interaction with Arl3 and the retinal effector HRG4, and (2) formation of a ternary complex crucial for the ciliary transport. ^9,10 However, this truncated RP2 lacks the C-terminal domain, which shares sequence homology with the nucleotide diphosphate kinases (NDPK), with a function that is as yet undetermined. ²⁸ Therefore, it is difficult to hypothesize a priori whether this mutation causes loss or gain of function. Very few correctly spliced transcripts are produced in patients (<5% in hemizygotes) and thus, the levels of wild-type RP2 protein are most probably insufficient in patients. 

A recent report stated that no RP2 point mutation had yet been identified with a semidominant inheritance pattern in carrier females, although semidominance has been claimed for the major XLRP gene, RPGR. ^4,29 However, in the present study female carriers of an RP2 mutation presented ample variation in the phenotypic range, from normal fundus and visual acuity (including visual field and ERG) to asymmetrical retinal alteration or even severely affected retinas bearing all the RP hallmarks. These discrepancies could be explained by differential X chromosome inactivation. Although the RT-PCRs were performed in blood and not in the retina, considering that (1) there was large splicing variation among different carrier females in the same family, (2) one of the affected females had asymmetrical retinal alteration, and (3) all hemizygous affected males from different branches of the family showed comparable levels of aberrant transcript (versus wild-type), we surmised that skewed X-inactivation rather than a polymorphic modifier X-linked gene was a more likely explanation to account for the wide phenotypic range in female carriers. Hence, according to our results RP2 joins RPGR as the cause of semidominant XLRP. 

Footnotes

 Supported by Fundaluce (2004), Grant BFU2006-04562 from the Ministerio de Educación y Ciencia, and Grant INTRA/07-08/718.1 from CIBERER (RG-D). EP was the recipient of an FPI fellowship from the Spanish Ministry of Education and Science (MEC) and is now under contract by CIBERER. MR was the recipient of a scholarship from Fundació Bidons Egara and is now the recipient of an FPU fellowship from MEC.

Footnotes

 Disclosure: E. Pomares, None; M. Riera, None; J. Castro-Navarro, None; Á. Andrés-Gutiérrez, None; R. Gonzàlez-Duarte, None; G. Marfany, None

Footnotes

 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.

The authors thank the family members for their participation in the study and Andrés Mayor for sample collection, helpful discussions, and constant support of our research.