De Novo Mutation in the RP1 Gene (Arg677ter) Associated with Retinitis Pigmentosa

Abstract

purpose. The Arg677ter mutation in the RP1 gene is one of the most common causes of autosomal dominant retinitis pigmentosa (RP). In the current study, a de novo Arg677ter RP1 gene mutation was identified in a patient with RP.

methods. RP1 gene mutation screening was performed in probands with simplex RP. In one proband with the RP1 mutation, paternity was established by analyzing 24 short tandem repeat polymorphisms. Additional candidate RP genes, including rhodopsin, RDS/peripherin, RP2, and RPGR, were also examined in this proband. Phenotype was characterized with psychophysics, electroretinography, and optical coherence tomography.

results. An RP1 (Arg677ter) mutation was identified in one of the patients with simplex RP, but the sequence change was not detected in his parents. Parentage was confirmed, and other candidate genes were negative for mutations. Retinal function and cross-sectional imaging studies in the patient indicated greater rod than cone dysfunction with a photoreceptor basis for the abnormalities.

conclusions. The de novo origin of an RP1 (Arg677ter) mutation in a patient with simplex RP suggests that this common autosomal dominant RP mutation can arise independently in the population and supports the hypothesis of a mutational hotspot in the RP1 gene.

The first well-characterized disease locus for the genetically heterogeneous retinal degenerations known as retinitis pigmentosa (RP) was RP1. ^{1 2 3 4} The formative work was primarily performed on a nine-generation pedigree with autosomal dominant (ad)RP. In this large kindred, RP1 was mapped to the pericentric region of chromosome 8. A nonsense mutation in exon 4 at codon 677 (Arg677ter) of the cloned RP1 gene was subsequently identified in two separate studies of this same pedigree. ^{5 6} Other adRP families with the Arg677ter mutation were also described at this time, as well as other RP1 mutations. ^{5 6 7}  

RP1 mutations are now thought to cause between 6% and 10% of adRP in populations of varying ethnic backgrounds. ^{8 9 10} The nonsense mutation, Arg677ter, accounts for approximately 2% of adRP, ⁸ making RP1 Arg677ter one of the most common genotypes in patients with adRP. The question of whether adRP families with RP1 Arg677ter were ancestrally related has been asked, and results of haplotype analysis have not suggested a founder effect. ^{6 8 9} Candidate gene screening of RP1 in cohorts of isolated or autosomal recessive RP, however, has not revealed de novo mutations. ^{8 11 12 13}  

We identified a proband with simplex RP who was heterozygous for the Arg677ter mutation in RP1. The asymptomatic parents did not show the mutation. Hence, this is the first report of a de novo RP1 Arg677ter mutation. The results lend support to a mutational hotspot theory for the high frequency of this RP1 mutation ^{8 9} and raise issues about the value of screening RP1 in simplex RP and the caution needed in genetic counseling of these patients. 

Methods

Patients

Probands with simplex RP (n = 138; 67 male and 71 female) were included in this study. Also included were three family members of a proband found to have a mutation in RP1. Informed consent was obtained from all subjects after explanation of the nature of the studies. The research procedures were in accordance with institutional guidelines and the Declaration of Helsinki. 

Molecular Analyses

Codons 657 to 896 of the RP1 gene were screened in all probands by single-strand conformational polymorphism (SSCP) analysis followed by automated bidirectional DNA sequence confirmation of the observed SSCP shifts. Details of the methodology have been reported. ¹¹ DNA from the proband with the RP1 mutation was also screened for mutations in the rhodopsin, RDS/peripherin, RP2, and RPGR genes. ¹⁴  

To analyze segregation of alleles and parental identity, short tandem repeat polymorphism (STRP) amplification products were denatured for 3 minutes at 94°C and electrophoresed on 0.4 mm denaturing gels (6% 19:1 acrylamide-bis, 7 M urea) with a running buffer of 1.0% TBE at 65 W for 3 hours at room temperature. After electrophoresis, gels were stained with a silver nitrate solution. For each marker, samples were labeled according to allele pair size and analyzed for segregation within the family. In addition to the eight markers shown in Figure 1A —(D1S518, D4S2367, D7S821, D11S956, D5S820, D8S1110, D7S2847, and D9S921)—parental identity studies were also performed with the following 16 markers: D9S302, D18S535, D7S1804, D4S2366, D5S1719, D3S1358, FGA, vWA, D2S1338, D19S433, D8S1179, D18S51, D16S539, CSF1PO, TH01, and D21S11. 

Phenotype Analyses

All RP probands were diagnosed by one of the authors (SGJ) on the basis of clinical examinations, psychophysical testing, and electroretinography. In the proband with the RP1 mutation, the following studies were performed to characterize phenotype: Goldmann kinetic perimetry (V-4e and I-4e test targets); static threshold perimetry in the dark-adapted (500- and 650-nm stimuli) and light-adapted (600-nm stimulus on 10 cd/m² white background) states using a modified automated perimeter (Humphrey Field Analyzer; Zeiss-Humphrey Systems, Dublin, CA); full-field electroretinography (ERG); and cross-sectional retinal imaging with optical coherence tomography (OCT; Zeiss-Humphrey Systems). The methods to define phenotype were those used in a previous study of other patients with RP1 mutations. ¹¹  

Results

An SSCP mobility shift was detected in one proband of the 138 patients with RP simplex screened for the sequence variation at codon 677 of RP1. Direct DNA sequencing revealed a heterozygous C→T transition at nucleotide 2029 resulting in the Arg677ter (CGA→TGA) nonsense mutation (Fig. 1B) . To exclude the possibility of sample error at any point, a second sample was independently obtained from the proband, and both SSCP mobility and Arg677ter sequence results were confirmed. Because the Arg677ter variation in RP1 has been shown to date to be associated only with adRP, ^{8 11 12 13} we examined other members of the family. This variation was not present in the clinically unaffected mother, father, and brother of the proband (Fig. 1B) . Specifically, the family members did not exhibit an SSCP shift on a gel containing the proband as a control, and DNA sequencing confirmed the normal sequence in this amplimer. 

To prove parental identity, DNA samples of the proband and his family members were examined with 24 polymorphic markers (8 of which are shown in Fig. 1A ). Other genes known to cause adRP and X-linked RP were also screened in this proband, and no coding sequence mutations were detected in rhodopsin, RDS/peripherin, RP2, and RPGR.  

The proband with the Arg677ter RP1 mutation reported night and peripheral vision disturbances beginning in the third decade of life. At age 30, his best corrected visual acuities were 20/30 in the right eye and 20/20 in the left eye. Ophthalmoscopy revealed mild peripheral pigmentary retinopathy without apparent regional predilection. Kinetic visual fields with a large target (V-4e) were symmetrical and full. With a small target (I-4e), there was only a central island with a relative paracentral scotoma (Fig. 2A) . Rod sensitivity was normal near fixation, markedly reduced in the midperiphery, but less abnormal in the far periphery. Long/middle wavelength (L/M) cone sensitivity was normal at fixation and at some loci in the far periphery. The midperiphery showed greater loss of sensitivity than other regions (Fig. 2A) . Rod–cone comparisons by these psychophysical methods indicated rod sensitivity loss was at least 2 log units greater than cone sensitivity loss. ¹¹  

Rod and cone ERGs were abnormal in the patient (Fig. 2B) . Rod ERG b-waves were markedly reduced in amplitude (31 μV; normal mean ± SD = 299 ± 52 μV). Both a-wave (40 μV; normal = 297 ± 65 μV) and b-wave (64 μV; normal = 497 ± 111 μV) of the mixed cone–rod ERG were reduced in amplitude (approximately 10% of mean normal). Cone ERGs amplitudes were also reduced but to a lesser extent (approximately 30% of mean normal), with 49 μV for 1 Hz (normal = 174 ± 32 μV) and 42 μV for the 29 Hz flicker (normal = 172 ± 35 μV). Flicker timing was delayed (45 ms; normal = 30 ± 1.2 ms). 

Cross-sectional images of the central retina in the left eye, obtained with OCT, revealed abnormal retinal architecture in the patient compared with normal subject (Fig. 2C) . The fovea appeared thickened, and there was evidence of cystoid macular edema, whereas the adjacent superior retina was thinner than normal. Longitudinal reflectivity profiles were analyzed at the fovea and parafovea. Although retinal thickness at the fovea was increased, the double-peaked outer retina-choroidal complex (ORCC) was preserved, consistent with normal visual acuity and normal cone sensitivity at fixation in this eye. The superior retinal locus showed reduced OCT thickness (from vitreoretinal interface to ORCC offset) and only a single-peaked ORCC. The patient’s signals, if cut and overlaid onto a normal template, suggest missing components, mainly the features believed to represent the outer nuclear layer and the inner and outer segments. ^{15 16} The visual sensitivity losses (Fig. 2A) may thus be attributable to anatomic abnormalities at the level of the photoreceptors. 

Discussion

The Arg677ter mutation is the most commonly observed RP1 mutation in individuals with adRP (for example, Refs. ^{8 9 10} ). The relatively high frequency of this mutation may signify a founder effect, such as occurs in other autosomal dominant retinal degenerations, including Pro23His in rhodopsin, ¹⁷ Ser181Cys in TIMP3, ¹⁸ Arg172Trp in RDS/peripherin, ¹⁹ Arg345Trp in EFEMP1, ²⁰ Ser50Thr in NRL, ²¹ and Asp226Asn in IMPDH1. ²² Alternatively, Arg677ter mutations may have arisen independently as a result of a mutational hotspot in the RP1 gene. The current observations, taken together with previous work, support the latter hypothesis. Haplotypes encompassing RP1 revealed that markers telomeric to the gene were not shared in 10 Arg677ter probands. ⁶ Genotyping of additional markers and mutation-carrying alleles also argued against a founder effect as an explanation for the high Arg677ter frequency. ^{5 8} Our finding of a de novo RP1 Arg677ter mutation in an individual with simplex RP provides direct evidence in support of a mutational hotspot theory. 

The Arg677ter mutation is a cytosine-to-thymine (C→T) transition at a hypermutable CpG dinucleotide. In the human genome, 60% to 90% of CpG dinucleotides are methylated. ²³ The propensity of 5-methylcytosine to undergo spontaneous deamination to thymine leads to a mutation rate much greater than the base mutation rate. ²⁴ This mutational mechanism could account for the recurrence of Arg677ter mutations. It is of interest that a second mutation at codon 677, Arg677del1bp, has been reported in one adRP family. ⁸  

The hypermutability of CpG dinucleotides accounts for recurrent disease-causing point mutations throughout the human genome (reviewed in Ref. ²⁵ ). The most remarkable example is achondroplasia mutations, all of which occur at a CpG dinucleotide in a single FGFR3 codon. Ninety percent of these mutations occur de novo and almost all are caused by C→T transitions in the antisense DNA strand. ^{26 27 28} Recurrent mutations at CpG dinucleotides are also known to occur in genes associated with retinal degenerations. The Pro347Leu rhodopsin mutation is not a founder mutation, as demonstrated by haplotype analysis, but is a recurrent mutation occurring at a CpG dinucleotide; a de novo Pro347Leu mutation has been reported. ²⁹ Other rhodopsin mutations have been detected in simplex RP (see for example, Refs. ^{12 30} ). Less frequent are reports of de novo mutations in other adRP genes, including RDS/peripherin ³¹ and NRL. ³²  

The practical implications of our study relate to molecular screening and genetic counseling. With evidence that some simplex RP cases may be due to de novo mutations in adRP genes, screening of such genes or specific recurrent mutations that occur at hotspots may be informative. The rhodopsin Pro347Leu mutation and the RP1 Arg677ter mutation are frequent enough causes of adRP that they may be worth screening in patients with simplex RP. If the patient presents at early disease stages, it may be possible to guide molecular screening by details of phenotype. For example, patients with rhodopsin Pro347Leu mutation have an early-onset retinally diffuse loss of rod function with retained but impaired cone function. ^{33 34} In contrast, mildly affected patients with the RP1 Arg677ter mutation can show regional retinal variation of disease and detectable rod function that is more impaired than cone function. ¹¹ Later disease stages in either genotype do not permit such distinctions to be made. The phenotype of the patient with simplex RP in the present study showed detectable rod function in the central and peripheral retina by psychophysics and a detectable rod ERG—findings that would be inconsistent with a rhodopsin Pro347Leu mutation but consistent with some of the moderately advanced phenotypes we have reported in adRP with the RP1 Arg677ter mutation. ¹¹  

Accurate genetic counseling in simplex RP is particularly complex, because modes of autosomal and X-linked inheritance must be considered, and risk provision varies in each case. Because most simplex RP is likely to be inherited as an autosomal recessive or, in the case of males, an X-linked condition, the tendency is to focus on risk of recurrence and transmittal in these inheritance patterns. Until de novo mutations in adRP genes are excluded, the possibility of a patient with simplex RP having a dominantly transmittable condition must be considered and a range of risk including 50% must be addressed. 

 Supported National Eye Institute Grants EY-13385, EY-13729, and EY-13203; Foundation Fighting Blindness, Macula Vision Research Foundation, Macular Disease Foundation, the F. M. Kirby Foundation, Research to Prevent Blindness, and the Mackall Trust. EMS is an investigator at the Howard Hughes Medical Institute.

 Submitted for publication February 11, 2003; accepted March 20, 2003.

 Disclosure: S.B. Schwartz, None; T.S. Aleman, None; A.V. Cideciyan, None; A. Swaroop, None; S.G. Jacobson, None; E.M. Stone, 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.

 

Figure 1.

 

(A) The simplex RP pedigree with genotype results showing eight polymorphic markers supporting the family relationships as indicated. The markers are: (A) D1S518, (B) D4S2367, (C) D7S821, (D) D11S956, (E) D5S820, (F) D8S1110, (G) D7S2847, and (H) D9S921. Squares, males; circles, females; filled square and arrow, proband. (B) Sequencing chromatograph results of the codon 677 region of the RP1 gene in this family. The heterozygous variation of C/T at codon 677 in the proband is in contrast to the lack of this variation in the unaffected sibling and parents.

Figure 2.

 

Retinal phenotype of the proband with the de novo RP1 (Arg677ter) mutation. (A) Kinetic (top) and static (middle, bottom) visual fields for the left eye. Paracentral shading in the kinetic field is a relative scotoma to I-4e. Static perimetry is shown as grayscale maps of rod and cone sensitivity loss; white is normal; black indicates more than 3 log units of loss. The physiological blindspot is shown as a black square at 12° temporal (T) field. N, nasal; S, superior; I, inferior. (B) Electroretinograms (ERGs) in a representative normal subject (age 31) and the patient (at age 30). Stimulus onset for the top three waveforms is at trace onset. Lines on the flicker ERG trace indicate the stimulus onset. Amplitude calibrations are to the right of the traces; timing calibrations are to the right and below. (C, left) OCT scans, vertically from 15° S to 5° I, in a normal subject and the patient. Images are displayed with logarithm of reflectivity mapped to a grayscale, allowing comparison with conventional pseudocolor displays (1, white; 2, red; 3, yellow; 4, green; 5, blue; 6, black). Arrow: region of cystoid macular edema. (C, right) Longitudinal reflectivity profiles (LRPs) at two retinal locations (I, II) in a representative normal subject (N, gray trace) and in the patient (P, black trace). For the paracentral locus (II), the patient’s LRP is shown twice: first, as original data, and, second, as split data to illustrate missing waveform components (retinal layers) secondary to disease. Depth calibration for LRPs is to the right.

The authors thank Leigh M. Gardner, Elaine B. DeCastro, Elaine E. Smilko, Jiancheng Huang, Michael J. Pianta, Alexander Sumaroka, Jessica M. Emmons, Adam Reddick, and Jean L. Andorf for help with the studies.