Molecular Vision 2013; 19:2487-2500
<http://www.molvis.org/molvis/v19/2487>
Received 01 July 2013 |
Accepted 06 December 2013 |
Published 08 December 2013
Béatrice Bocquet,1,2,3 Nour al Dain Marzouka,1,2,3 Maxime Hebrard,1,2,3 Gaël Manes,1,2,3 Audrey Sénéchal,1,2,3 Isabelle Meunier,1,2,3,4 Christian P. Hamel1,2,3,4
The first two authors contributed equally to this work.
1INSERM U. 1051, Institute for Neurosciences of Montpellier, Montpellier, France; 2Université Montpellier 1, Montpellier, France; 3Université Montpellier 2, Montpellier, France; 4Genetics of Sensory Diseases, CHRU, Montpellier, France
Correspondence to: Béatrice Bocquet, INSERM U. 1051, Institut des Neurosciences de Montpellier, Hôpital Saint-Eloi, BP 74103, 80, rue Augustin Fliche, 34091 Montpellier Cedex 5, France; Phone: (33) 467 336 974; FAX: (33) 467 330 280; email: beatrice.bocquet@inserm.fr
Purpose: Autosomal recessive retinitis pigmentosa (arRP) is a genetically heterogeneous disease resulting in progressive loss of photoreceptors that leads to blindness. To date, 36 genes are known to cause arRP, rendering the molecular diagnosis a challenge. The aim of this study was to use homozygosity mapping to identify the causative mutation in a series of inbred families with arRP.
Methods: arRP patients underwent standard ophthalmic examination, Goldman perimetry, fundus examination, retinal OCT, autofluorescence measurement, and full-field electroretinogram. Fifteen consanguineous families with arRP excluded for USH2A and EYS were genotyped on 250 K SNP arrays. Homozygous regions were listed, and known genes within these regions were PCR sequenced. Familial segregation and mutation analyzes were performed.
Results: We found ten mutations, seven of which were novel mutations in eight known genes, including RP1, IMPG2, NR2E3, PDE6A, PDE6B, RLBP1, CNGB1, and C2ORF71, in ten out of 15 families. The patients carrying RP1, C2ORF71, and IMPG2 mutations presented with severe RP, while those with PDE6A, PDE6B, and CNGB1 mutations were less severely affected. The five families without mutations in known genes could be a source of identification of novel genes.
Conclusions: Homozygosity mapping combined with systematic screening of known genes results in a positive molecular diagnosis in 66.7% of families.
Photoreceptor degeneration is the leading cause of inherited blindness [1]. This is partly explained by the extreme genetic heterogeneity of these conditions as more than 200 genes are currently registered in the RetNet database, reflecting the vast repertoire of genes necessary for photoreceptor or other retinal cell function. The most frequent clinical entity, nonsyndromic retinitis pigmentosa (RP; OMIM 268000), is also the most genetically heterogeneous with more than 50 disease-causing genes currently associated with this condition. Among these, 36 are known to be responsible for autosomal recessive (ar) inheritance, accounting for 50–60% of all arRP cases [2]. Two major genes are responsible for ar inheritance, USH2A [3,4] and EYS [5-8].
Homozygosity mapping in inbred multiplex families or isolated cases with presumed ar inheritance has proven successful for finding novel genes [9-19] and identifying mutations in previously described genes [20-35]. Homozygosity mapping saves time as it readily highlights regions containing already known disease-causing genes or new genes/loci. This strategy has also been successful in a variable proportion of cases from outbred families who carry a homozygous mutation due to a high level of inbreeding encountered in some populations [21,22,36]. In this study, we applied this strategy to a series of 15 families with consanguineous parents and found that two-thirds of the families carried a mutation in a known arRP gene.
Consanguineous arRP families were selected from 423 families with arRP. Informed written consent and peripheral blood samples were obtained for genetic analysis from all family members according to approved protocols of the Montpellier University Hospital, in agreement with the Declaration of Helsinki.
Patients underwent standard ophthalmologic examination (refractometry, visual acuity, slit-lamp examination, applanation tonometry, funduscopy). Kinetic visual fields were determined with a Goldman perimeter with targets V4e, III4e, and I4e. OCT measurement of the macula was performed using an OCT-3 system (Stratus model 3000; Carl Zeiss Meditec, Dublin, CA) with software version 3.0. Autofluorescence measurements were obtained with the HRA2 Heidelberg retinal confocal angiograph (Heidelberg Engineering, Dossenheim, Germany), and fundus pictures were taken. Full-field ERG was recorded using a Ganzfeld apparatus (Metrovision, Pérenchies, France) with a bipolar contact lens electrode on maximally dilated pupils according to the ISCEV protocol [37].
Genomic DNA was isolated from leucocytes using a proteinase K digestion, followed by an ethanol precipitation [38]. DNA samples were quantified by a spectrophotometer, aliquoted and stored at +4 °C and -20 °C until use. From 31 consanguineous families (26 multiplex, five sporadic), 15 families that were not homozygous for EYS and USH2A microsatellite markers were selected and genotyped with 262,270 single nucleotide polymorphisms (SNPs; GeneChip Mapping 250 K Nsp Array; Affymetrix, Santa Clara, CA) at DNAVision, Charleroi, Belgium. Results were analyzed using the common homozygosity regions test of the transmitted allele search engine (TASE) [39]. TASE was designed to screen for common homozygous genotypes in all affected individuals that are heterozygous or wild type in unaffected individuals. Candidate chromosomal regions of homozygosity larger than 2 Mb were compared to the position of known genes and loci for retinal inherited diseases according to the RetNet database. All exons and exon–intron boundaries of the candidate genes were then sequenced. Each PCR was performed in a 25-µl reaction mix containing 50 ng of genomic DNA, 2 mM MgCl2, 200 µM deoxyribo nucleotide triphosphate (dNTPs), 0.2 µM of each primer (designed with Primer 3 software), and 1 U of AmpliTaq Gold DNA Polymerase (Applied Biosystems, Foster City, CA) in a buffer composed of 100 mM Tris-HCl, pH 8.3 and 500 mM KCl. Following the first denaturation at 95 °C for 9 min, amplification was carried out for 35 cycles at 95 °C for 30 s, at the melting temperature (Tm) of the primers (56 °C-60 °C) for 30 sec and at 72 °C for 1 min, ending with a final extension step at 72 °C for 10 min. PCR products were purified with ExoStar 1-step clean up (GE Healthcare, Little Chalfont, UK) and sequenced using the BigDye Terminator cycle sequencing ready reaction kit V3.1 on an Applied Biosystems 3130xL genetic analyzer, following the manufacturer’s instructions. Sequencing results were analyzed with Sequencing Analysis v5.2 software (Applied Biosystems).
Mutations identified were validated by reading the existing literature, performing familial segregation whenever possible, consulting the Human Gene Mutation Database (HGMD), and interrogating databases with mutation frequencies (1000 genomes, Exome Variant Server [EVS], dbSNP). Missense variations were systematically analyzed using Polyphen-2 and SIFT softwares.
From the 15 consanguineous families, eight families were simplex and seven were multiplex (Figure 1). An average of two SNP arrays per family was performed (range one to three arrays). The quality control (QC) call rate for all samples was always above 90%.
For each family, homozygous regions were classified as a function of SNP coverage (number of SNPs/Mb; Table 1). There were on average 7.7 homozygous regions (range 1–18 regions) with a mean size of 15.3 Mb (range 2.2–53 Mb) per family. The homozygous regions were compared with the position of the genes listed in the RetNet database, and the corresponding genes were systematically sequenced, with priority given to known nonsyndromic arRP genes. A total of 32 genes were sequenced for the 15 families. For eight of these families, only one gene was sequenced and this revealed a causative mutation.
We found that ten out of 15 families had a causative homozygous mutation in one of the genes screened (Table 2). In eight of the ten families, segregation analysis could be performed and the results obtained were in accordance with the autosomal recessive inheritance of the disease (Figure 1). Among the ten mutations, three were previously described (one mutation was only described in the dbSNP database). Seven mutations were either nonsense, frameshift, or large deletions and were presumed to lead to loss of protein function.
Three mutations were amino acid changes. The c.364C>T (p.Arg122Cys) in NR2E3 is not found in the HGMD database nor in the EVS. It is located two amino acids downstream of the DNA-binding domain of the protein and is considered to be probably damaging with a score of 1.000 by Polyphen-2. The c.2284C>T (p.Arg762Cys) in exon 23 of CNGB1 affects arginine 762 located in the extracellular domain, which is found in two isoforms of the protein but is missing in the third isoform (named GARP2, for glutamic acid-rich protein-2). This amino acid is conserved in all species in the Polyphen-2 software. It is not described in the EVS, 1000 genomes, and HGMD databases. It is predicted to be probably damaging by Polyphen-2 and damaging by SIFT. The c.1568T>G (p.Met523Arg) in exon 12 of PDE6B is located between the cGMP binding domain and the catalytic domain where many mutations known to cause RP are found. This mutation is not described in the EVS, 1000 genomes, and HGMD databases and is predicted to be possibly damaging by Polyphen-2 and damaging by SIFT.
In total, we found that ten out of 15 families, i.e., 66.7% of the sequenced families, had a mutation in a known gene. For the remaining five families (33.3%), all RP genes in homozygous regions were sequenced but no mutation was detected, suggesting that their proband could carry a mutation in a novel gene. We cannot, however, exclude that mutations in noncoding genomic sequences have been missed. These negative families will ultimately undergo exome sequencing.
Clinical description was available for 12 patients from the ten families in which a mutation was found (Appendix 1). The clinical features were variable depending on the gene involved, although all patients had night blindness (not known for II:4 from PB15), retinal vessel attenuation and retinal atrophy in fundus, and strongly decreased ERG responses.
The 19- and 17-year-old II:1 and II:2 sisters from RP1361 had the same clinical presentation, revealing severe RP, which was consistent with the homozygous PDE6A null mutation that they carried as the produced protein is expected to be unstable and degraded by nonsense-mediated decay [40]. They showed few pigment deposits in the fundus (Figure 2A, B) and few atrophic spots in peripheral retina (Figure 2C). The retinal arterioles, however, were already narrow and the macular area showed a typical ring of autofluorescence best seen in patient II:2 (Figure 2D). The sisters also had an important bilateral macular edema (Figure 2E, F) with decreased visual acuity between 0.5 and 0.7. They had no scotopic ERG responses but both still had minimal photopic responses. In comparison, 46-year-old patient II:1 from PB74, who had a missense mutation in PDE6B, retained relatively good visual acuity (0.8 in both eyes). The 44- and 57-year-old II:8 and II:3 brothers from RP1324 also had severe RP due to a missense homozygous mutation in CNGB1. The younger brother had ocular trauma on the left eye in infancy with no light perception, while the contralateral eye had decreased visual acuity at 0.3 (in part due to a cataract as shown by the blurred fundus image [Figure 2G]), bone spicule pigment deposits in the retinal periphery, and narrowing of retinal vessels without atrophy of the optic disc. The IS/OS line was still present in the foveal area. He had tunnel vision at 20–30°, and ERG responses were absent. The elder brother had advanced RP with bare light perception in both eyes, atrophy of both peripheral retina and macula, and large pigment deposits distributed throughout the retina (Figure 2H).
The 29-year-old III:2 patient from RP1013 carried a null mutation in C2ORF71, was myopic, and had severe RP; visual acuity was decreased at 0.3 OD and 0.4 OS. Fundus examination showed bilateral, round, foveal atrophy with narrowed retinal vessels and atrophic optic discs (Figure 2I). Retinal autofluorescence testing revealed small atrophic spots grouped in the foveal area (Figure 2J). The visual field was tubular at 30°, and ERG photopic responses were still recordable although very low.
Both patients with a homozygous RP1 mutation also had severe RP, were myopic, and showed bilateral macular involvement. The younger 10-year-old patient II:2 from RP1625 had decreased visual acuity at 0.6 on both eyes. The fundus showed an abnormal foveal reflex, a dark perifoveal area, narrowed retinal arterioles, atrophy of the peripheral retina, and a few small pigment deposits (Figure 2K). Retinal autofluorescence testing revealed many atrophic spots in the peripheral retina and a slightly increased autofluorescence around the fovea (Figure 2L). The outer nuclear layer and the IS/OS were absent except in the foveola where they remained only partly preserved (Figure 2M, N). The patient had tunnel visual field (10–20°), and ERG responses were absent. The older 37-year-old patient III:2 from RP854 had hand motion in both eyes. Fundus examination showed a bilateral round atrophy of the macula, narrowed retinal vessels, atrophic optic discs, and many bone spicule pigment deposits in the retinal periphery (Figure 2O, P). The visual field was undetectable, and ERG responses were absent. The 13-year-old patient from PB15 with a null mutation in IMPG2 also had severe RP with myopia, macular involvement, decreased visual acuity at 0.2 in both eyes, and no ERG response.
Both patients with homozygous RLBP1 mutations had less severe RP than the ten other patients and both had early onset night blindness. The 32-year-old patient from RP517 family had decreased visual acuity in accordance with foveal thinning, but the retinal vessels were moderately narrowed and the optic discs were not atrophic (Figure 2Q). The fundus had a whitish aspect and the retinal periphery showed rare clumps of pigment deposits. When this patient was examined at the age of 40, larger atrophic spots were visible (Figure 2R, S). The 58-year-old II:3 from RP1682 had a similar presentation although with more advanced disease. Visual acuity was still at 0.4 OD and 0.3 OS. The fundus showed large scallop-shaped spots of atrophy covering the mid-periphery of the retina (Figure 2T). There were semicircular atrophic spots around the fovea in both eyes that were secondary to previous laser treatment of macular edema.
With the advent of clinical trials for inherited retinal dystrophies, the causative gene needs to be identified. Molecular identification permits the diagnosis of the RP subtype, improved patient follow up, and prediction of disease course. Gene identification is also necessary for gene therapy and to organize patient series for clinical trials. However, molecular diagnosis in arRP, the most genetically heterogeneous form of inherited retinal disease, currently requires screening 36 genes, a process which has never been completed by any research group by Sanger sequencing because it is time and money consuming.
As a preliminary approach to exome sequencing, we used SNP genotyping for homozygosity mapping of consanguineous families and found that two-thirds of the 15 families carried a homozygous causative mutation in a known gene. At the time of our study, it was more economical to perform homozygosity mapping with SNP 250 K arrays to select for families negative for known genes. Presently, the cost of targeted sequencing of RP genes or exome sequencing using next generation sequencing (NGS) is equivalent to that of mapping with SNP arrays, therefore making this approach more affordable.
Given that an average of 13% of arRP cases (range 5% to 18%) may have a mutation in EYS [5-8] and that 16% (range 12% to 20%) may have a mutation in USH2A [3,4], then 76% of patients with arRP are estimated to have a mutation in a known gene, meaning that about a quarter of the arRP patients would have mutations in yet undiscovered genes. In this series, we did not find any homozygous or heterozygous mutations in candidate genes within homozygous regions for the five negative families. These regions represent good candidates to find novel genes by whole exome sequencing. Future analyzes will then be shortened by directly exploring the homozygous regions. Our results showed a high percentage (70%) of novel mutations, indicating that there is considerable allelic heterogeneity in arRP. Similar results were found in a recent study in which 63% of novel mutations were found by targeted sequencing using NGS in a Chinese patient cohort with arRP [41].
The analysis of patient phenotypes showed some variations in disease severity. We found that the three patients with mutations in a connecting cilium gene, i.e., RP1 and C2ORF71, had severe RP with early macular degeneration, while the patients with mutations in a phototransduction gene, i.e., PDE6A, PD6B, and CNGB1, had less severe RP [40,42-45]. Indeed, patients with arRP due to RP1 mutations were frequently reported with legal blindness by their twenties and thirties. Previous case reports described onset in childhood, flat ERG by 18 years, macular involvement before 20 years, or even total blindness before 20 [46-53]. Macular involvement is found earlier in RP due to cilia-associated genes, such as Bardet–Biedl syndrome genes and the recently described ARL2BP gene [54]. Conversely, PDE6A and PDE6B phenotypes show great variation in the severity of disease and frequent macular edema [55]. Yet, it remains difficult to preselect genes for screening based only on macular involvement and severity of the disease.
It is of note that the two patients with a mutation in RLBP1 were not diagnosed as having retinitis punctata albescens (RPA). Apart from typical RPA, RLBP1 mutations have been reported in two subclinical forms of RPA, Bothnia retinal dystrophy [56] and Newfoundland rod–cone dystrophy [57], as well as in rare cases of arRP [58]. In our patients, the specific, small, white, dot-like deposits usually observed on the fundus were not present [59]. It is possible that the dots were present at early stages and had progressively vanished in the course of the disease, thus preventing the correct diagnosis, as previously reported [60]. It is also important to mention that the two patients with this RPA form of arRP had the least severe phenotype among the 12 patients examined. It is known that signs of retinal degeneration (retinal vessel attenuation, optic disc pallor) progress more slowly in RPA than in typical arRP [61].
We thank the patients and their families and Vasiliki Kalatzis for critical review of the manuscript. Funding: This work was supported by private foundations (Fondation des Aveugles et Handicapés Visuels de France, Formicoeur, Information Recherche sur la Rétinite Pigmentaire, Retina France, SOS Rétinite, and UNADEV).