Molecular Vision 2020; 26:445-458
Received 20 August 2019 | Accepted 17 June 2020 | Published 19 June 2020
1Ophthalmology Department, Jordan University Hospital, The University of Jordan, Amman, Jordan; 2Medical Department, The University of Leipzig, Germany; 3Department of Pathology, Microbiology and Forensic Medicine, School of Medicine, The University of Jordan, Amman, Jordan; 4Cell Therapy Center, The University of Jordan, Amman, Jordan; 5Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA; 6Department of Ophthalmology St Thomas' Hospital, London, UK; 7Department of Human and Molecular Genetics, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA; 8Prevention Genetics, Marshfield, WI; 9Department of Medicine and Hematology, Jordan University Hospital, The University of Jordan, Amman, Jordan
Correspondence to: Belal Azab, Prevention Genetics, 3800 S Business Park Ave, Marshfield, 54449 WI, Phone: (804) 263-2950; email: email@example.com
Purpose: The aim of this study is to identify disease-causing variants in five consanguineous Jordanian families with a history of autosomal recessive retinitis pigmentosa (RP), and to investigate the clinical variability across the affected individuals.
Methods: Exome sequencing (ES) and ophthalmic examinations were performed to classify the underlying RP-causative variants and their pathogenic consequences. The candidate variants in the affected and unaffected family members underwent segregation analyses with Sanger sequencing.
Results: We described four variants in the RP1 and RLBP1 genes as disease-causing across the five families, including novel (c.398delC; p.Pro133GlnfsTer126) and recurrent (c.79delA; p.Thr27ProfsTer26) variants in RLBP1 and two previously reported variants in RP1 ((c.1126C>T; p.Arg376Ter) and (c.607G>A; p.Gly203Arg)). The consequent clinical manifestations were thoroughly investigated using a battery of ophthalmic tests, including electroretinography (ERG), optical coherence tomography (OCT), visual acuity (VA), and fundus examination. The phenotypes indicated clinical heterogeneity, typical RP for variants in RP1, and retinitis punctata albescens (RPA) for variants in RLBP1.
Conclusions: This study extends the pathogenic variant spectrum for the RP1 and RLBP1 genes. The study also revealed the consequent clinical progression, severity, and presentation of RP. Furthermore, we confirm that ES is an efficient molecular diagnostic approach for RP.
Inherited retinal dystrophies (IRDs) include clinically and genetically complex diseases such as retinitis pigmentosa (RP) . RP is the most common degenerative peripheral retinal dystrophy caused by progressive deterioration of photoreceptors [2,3]. The prevalence of RP depends on geographic location and ranges between 1:9,000  and 1:750 . Prevalence increases in ethnic groups that share common ancestors, such as the Jordanian population . During RP preclinical stages, rod degeneration is not often observed through standard clinical examination. Thorough retinal assessment, such as electroretinography (ERG), is required to investigate the disease, especially in its early stages . The first symptoms are night blindness followed by progressive visual field scotomas, decreased visual acuity, and abnormal color vision. RP later advances to cone degeneration, leading to total blindness . The clinical manifestations of deteriorated retinas include attenuated vessels, intraretinal pigment deposits, and waxy optic disc pallor [9,10]. The clinical presentation varies depending on the penetrance and expressivity of disease-causing genes . RP exhibits phenotypic heterogeneity with a wide range of severity and variable age of onset, ranging from early childhood to late adulthood . It is a complex disease with genetic and allelic heterogeneity , and can arise from different modes of inheritance, including autosomal dominant (adRP), autosomal recessive (arRP), X-linked (XLRP), and mitochondrial (mtRP) . Thus far, more than 57 genes with about 3,000 RP-causative variants have been identified .
In this study, RP1 (Gene ID 19888; OMIM 603937) and RLBP1 (Gene ID 6017; OMIM 180090) were identified through exome sequencing (ES) to cause RP in five consanguineous Jordanian families. ES represents a highly sensitive and efficient strategy for molecular diagnosis in phenotypically and genetically heterogeneous monogenic diseases, such as RP , with a detection rate ranging from 36% to 80% .
This study was performed according to the principles of the Declaration of Helsinki, following the ARVO statement on human subject studies in Ophthalmic and Vision Research and was approved by the Institutional Review Board of the Cell Therapy Center, The University of Jordan. All participants provided written informed consent. Five extended consanguineous Jordanian families affected by retinal diseases were recruited for the study. Pedigrees F1, F2, and F5 present a first-degree cousin marriage, while pedigrees F3 and F4 have first- and second-degree cousin marriages. Of these families, 12 affected individuals participated as follows: three patients in F1, two patients in F2, one patient in F3, four patients in F4, and two patients in F5. Thirteen unaffected relatives were included as controls: parents and siblings in families F1 to F4 and one sibling in family F5 (Figure 1).
The inclusion criteria for participants were RP-related symptoms, such as night blindness or decreased night vision, low visual acuity, and reduction in ERG amplitudes. Furthermore, the affected individuals underwent ophthalmic examinations, including visual acuity (VA), best-corrected visual acuity (BCVA), slit-lamp biomicroscopy (BM 900; Haag-Streit, Koeniz, Switzerland), dilated fundus examination, fundus photography (200Tx; Optos, Dunfermline, Scotland, UK), optical coherence tomography (OCT) to measure retinal thickness and integrity (Optovue RTVue, Fremont, CA), Pentacam for RP-related keratoconus detection (Pentacam Typ70700; Oculus, Wetzlar, Germany), and full-field flash electroretinograms (ffERGs) to measure photoreceptor electric activity (Color Ganzfeld Q450 C; Roland Consult, Brandenburg an der Havel, Germany). ERGs were recorded following the standards of the International Society for Clinical Electrophysiology of Vision.
Blood samples for molecular genetic testing were collected from affected and unaffected relatives of families F1, F2, F3, F4, and F5. Genomic DNA was extracted from peripheral blood leukocytes using a QIAprep Spin Miniprep Kit (Qiagen GmbH, Duesseldorf, Germany) according to the manufacturer’s instructions. The quantity and quality of the extracted DNA were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). The DNA samples were stored at −80 °C.
Exome sequencing and data analysis-- Exome sequencing was conducted for a single proband from each family. Exome capture was completed using an Agilent Sure Select Human All Exon 65 Mb kit V5 (Agilent, Santa Clara, CA), according to the manufacturer’s protocols. The DNA libraries of the probands were sequenced using the HiSeq 2500 platform (Illumina Inc., San Diego, CA) to generate 100-bp paired-end reads at the Partners HealthCare Personalized Medicine (PPM) Translational Genomics Core (Cambridge, MA). Sequence reads were aligned to the human reference genome (GRCh37) using the Burrows-Wheeler Aligner, and variants were called using the Genomic Analysis Tool Kit (GATK). To identify the candidate disease-causing variants, the findings were annotated and filtered against four databases: NCBI CCDS, RefSeq, Ensembl, and Encode. The Variant Call Format (vcf) files were analyzed using the Illumina basespace variant interpreter tool. The filtered variants were subsequently queried in ClinVar and the Human Gene Mutation Database (HGMD). To evaluate the possible deleterious effects of the variants, in silico prediction analyses were performed using Alamut Visual prediction software (SIFT, MutationTaster, and PolyPhen-2). Alamut annotations were also used to estimate the evolutionary conservation of the respective amino acid positions. The variants were classified according to the interpretation guidelines of the American College of Medical Genetics and Genomics (ACMG).
Validation and segregation analysis-- Sanger sequencing was used to confirm the pathogenic variants revealed by ES and to perform segregation analysis in affected and unaffected family members. Primers were designed using Primer 3.0 and synthesized by IDT (Appendix 1). Genomic DNA was PCR-amplified using Platinum PCR SuperMix (Invitrogen) with (i) 94 °C for 2 min, (iii) 94 °C for 30 s and 60.0 °C for 15 s and 68 °C for 30 s (35 cycles), (iii) 68 °C for 5 min, and the PCR products were then purified using a GeneJET PCR purification kit (Invitrogen). Sequencing was performed using a BigDye Terminator V3.1 Cycle Sequencing kit (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA) on an ABI 3500 genetic analyzer, and the sequence data were analyzed with SeqA software (Applied Biosystems) and Chromas Pro software (Technelysium LTD, South Brisbane, Australia).
Five consanguineous Jordanian families were recruited, including 12 individuals between 25 and 60 years affected with RP. Patients underwent thorough ophthalmic examinations; the clinical results are summarized in Table 1, Table 2, and Appendix 2. The five pedigrees were clinically categorized into two phenotypically correlated groups, typical RP and retinitis punctata albescens (RPA), based on the fundus examination, initial symptoms, and disease progression. RP1 was the candidate underlying disease-causing gene in families F1, F2, and F3 (Figure 1). Disease onset presented as initial symptoms of nyctalopia, between the age of 5 and 10 years. The pathology progressed to peripheral vision loss and central vision loss between 22 and 30 years. However, patient F2-V5, age 27, did not present with central vision loss. Fundus images revealed the typical RP triad of attenuated retinal vessels, intraretinal pigment deposits, and waxy optic disc pallor (Appendix 2). ERG results showed moderate (F1-V6 and F3-V4) and severely reduced electrical responses (F1-V5, F1-V9, F2-V5, and F2-V5; Figure 2). We noticed that the ERG degrees of severity did not correlate with the patients’ decreased VAs. The VAs for F1 and F3 members were light perception (LP) or hand motion (HM), with a central vision loss at the ages of 25 and 22, respectively. However, only the OCT images for family F1 members demonstrated moderate to severe foveal atrophy. The foveal thickness of the proband F3-V4 was pathologically unremarkable (Appendix 2). Analysis of the left eye of patient F1-V5 was not available due to unsuccessful retinal detachment surgery, and the OCT of her right eye showed thinned outer retinal layers (Appendix 2). Her right eye was diagnosed with posterior subcapsular cataract (PSCC) and keratoconus. Furthermore, patient F3-V4 was affected by PSCC and keratoconus in both eyes (Appendix 3). Disease progression was slower in family F2. The onset of central vision loss occurred at age 30 in this family; therefore, F2-V5 was not affected at age 25. His VA was ≥0.3 logMAR. The OCTs of this family showed normal or moderately thinned retinal layers, and only F2-V6 was diagnosed with PSCC (Table 1 and Appendix 2). Patient F2-V5 was prescribed daily vitamin A supplementation by an ophthalmologist between the age of 10 and 13 and daily every other month in the following time. The F1 members’ lipid profiles were assessed after sequencing and segregation analyses to evaluate the clinical effects of the variant in TTPA (Gene ID 7274; OMIM 600415). Low-density lipoprotein (LDL) concentrations were moderately high, and high-density lipoprotein (HDL) concentrations were low (Appendix 4).
RLBP1 was the candidate gene in the two other extended consanguineous pedigrees, families F4 and F5 (Figure 1). The initial symptom was nyctalopia at the age of 2 in both families. Although disease onset was early in life, disease progression was slow (Table 2). Members of family F4 presented partial peripheral and central vision and better paracentral vision, whereas family F5 members lost their peripheral and central vision at the age of 51 and 60 (Table 2). All members of family F4 had headaches upon sun exposure, which is associated with photophobia. Fundus examinations verified an atypical form of RP and RPA (Appendix 2). This condition is expressed through a large number of discrete, small, yellow-white dots at the level of the RPE and concentrated in the retinal midperiphery . The ERG results were severely reduced (Figure 2), whereas the VAs were at the level of counting fingers (Table 2). The OCT images revealed atrophic changes at the macula (Appendix 2). Patient F5-V6 exhibited the most advanced phenotype. He lost his peripheral vision at the age of 27 and central vision at the age of 41. His current VA was only hand motion, and the OCT showed severe atrophic maculopathy. Additionally, he was the only patient diagnosed with PSCC in both eyes. Further investigations of his fundus showed no RPA, but typical RP, unlike his sister and family F4 members. This may be due to advanced retinal degeneration. All patients in this study were evaluated with slit-lamp biomicroscopy, and only F4-V13 had a congenital blue dot cataract (Table 2). The mode of inheritance for congenital blue dot cataract is often autosomal dominant, but cases of autosomal recessive and X-linked mutations have been reported .
Exome sequencing was performed on the probands of the five families affected with arRP (F1-V9, F2-V5, F3-V4, F4-V13, and F5-V6; Figure 1). The mean depth across the targeted regions ranged from 59 to approximately 149, and all targets were covered at ≥10X. To identify candidate pathogenic variants, ES results were filtered based on the following criteria: (i) Out of the total variants, we filtered variants in the critical genomic regions (exons and splice sites), including loss of function and missense variants with a total read depth ≥10X. (ii) We filtered the disease-causing genes that are associated with retinopathies . (iii) The pedigrees demonstrated an autosomal recessive mode of inheritance (Figure 1). Therefore, variants with homozygous or compound heterozygous status were prioritized. (iv) We then filtered the variants with a minor allele frequency (MAF) <1% in the ExAC, GnomAD, 1000 Genomes Project, and NHLBI Exome Sequencing Project (ESP; Table 3). We performed in silico prediction analyses for the filtered candidate variants using the Alamut Visual prediction software tool. These variants were further prioritized based on the information available in public variation databases, including HGMD, ClinVar, and Ensembl. The variant filtration strategy identified a total of seven variants across the five probands in five different RP genes (RP1, RLBP1, MERTK [Gene ID 10461; OMIM 604705], TTPA, and IFT140 [Gene ID 9742; OMIM 614620]), all of which were homozygous (Table 4). To further assess the candidate variants, segregation analyses were performed in the affected and available unaffected family members (Figure 1 and Figure 3). This led us to determine the disease-causing variants in each family (Table 4). For family F1, the variant in RP1 (c.1126C>T; p.Arg376Ter) segregated in the homozygous state with RP in affected family members and the RP1 missense variant (c.607G>A; p.Gly203Arg) was found in families F2 and F3. Two disease-causing frameshift variants in RLPB1 were responsible for RPA in families F4 and F5 (c.79delA, p.Thr27ProfsTer26 and c.398delC, p.Pro133GlnfsTer126), respectively. Both variants were observed to segregate in the homozygous state with the disease in the affected members, while heterozygous carriers were unaffected. For family F5, segregation analysis was available for one sibling.
The pedigrees for the five participating families demonstrated an autosomal recessive RP inheritance pattern (Figure 1). The disease was diagnosed in every affected member of the five consanguineous Jordanian families. The initial symptoms were nyctalopia and decreased VA. The subjects showed phenotypic heterogeneity, depending on the causative genes. The fundus examinations revealed different subtypes of RP: typical RP in families F1, F2, and F3 for whom we identified variants in RP1 and RPA in families F4 and F5 in whom we identified variants in RLBP1. In cases of typical RP, the disease starts with nyctalopia in the first decade of life, followed by peripheral vision loss (second decade) and central vision loss (third decade). Disease progression was slower in patients from family F2. Patient F2-V5 has been treated with continuous vitamin A supplementation for 15 years. It was reported in the literature that vitamin A supplementation could have positive effects on the visual field or the retina electrophysiology, or on both, depending on the patient’s genotype . Further studies on the positive effects of vitamin A on the retina must be conducted, and its effects cannot be confirmed in this case. For all participants, the ffERG confirmed the early dysfunction of retinal photoreceptors, but it did not correlate with the disease severity in its advanced stages. RP severity was presented in the VA, BCVA, and OCT images (Table 1 and Appendix 2). Across the five participating families, PSCC or keratoconus was present in some of the affected individuals (Appendix 3, Table 1, and Table 2). It has been reported that keratoconus correlates with hereditary retinal dystrophies, including RP .
Through ES, we identified two variants in the RP1 gene for participants affected by typical RP and verified them with Sanger sequencing. Oxygen-regulated photoreceptor 1 (RP1; OMIM 603937), which is located on chromosome 8 (8q11.2-q12.1), consists of four exon regions, and encodes a 2,156-amino acid protein . Mouse models and human tissue analysis revealed that RP1 is differentially expressed in the photoreceptor outer segment architecture and is localized in connecting cilia rods and cones [23,24]. The protein encoded by RP1 plays a role in transporting proteins between the inner and outer segments of the photoreceptors where it participates in regulating c-Jun N-terminal kinase (JNK) signaling cascades . As reported in RetNet, variants in RP1 cause 5% to 10% of the arRP cases . Eighty-three pathogenic and likely pathogenic variants, including frameshift, missense, and nonsense variants, have been identified in RP1 (ClinVar). Pathogenic and likely pathogenic variants are mostly located in exon 4 . In family F1, we identified a homozygous nonsense variant in the last RP1 exon (c.1126C>T; p.Arg376Ter; Figure 1, Table 4). This variant was previously reported to cause arRP in a Pakistani pedigree . The highest MAF for this variant were 0.013% (gnomAD) and 0.01213% (ExAC) in the South Asian population (Appendix 5). Segregation analysis showed that RP1 (c.1126C>T; p.Arg376Ter) is homozygous in the affected individuals only and heterozygous in the unaffected family members (Figure 1, Table 4, Figure 3). We excluded the candidate variant in the TTPA gene after segregation analysis for being homozygous in an unaffected family member. However, elevated LDL and lower HDL concentrations might be related to impaired vitamin E function caused by the variant in TTPA .
In families F2 and F3, the ES results identified another previously reported missense variant in RP1 (c.607G>A; p.Gly203Arg; Figure 1, Table 4). This variant has been reported in the Iranian population to cause arRP . It is conserved across various related species (Alamut Visual Software; Orthologs Ensemble; Appendix 6). In silico analysis tools predict this variant to be disease-causing. It is currently classified in ClinVar as likely pathogenic (rs786205589). The validation and segregation analyses showed that five patients in families F2 and F3 were homozygous for RP1 (c.607G>A; p.Gly203Arg), and their unaffected relatives were not (Figure 1 and Figure 3). The present analyses further supported a disease-causing role for this variant.
Clinical investigations for families F4 and F5 revealed a form of atypical RP: RPA. Those clinical manifestations differed from the RP phenotypes in families F1, F2, and F3. The disease onset was at an earlier age, 2 years old, but with slower vision deterioration. All affected subjects in family F4 were 26 to 34 years old and still had peripheral and central vision. The older proband of family F5 presented peripheral vision loss at the age of 35 years and central vision loss at the age of 50. Compared with the affected individuals in family F5, F5-V6 presented fewer symptoms of RPA, and his retinal degeneration was faster (Table 2). This suggests variable expressivity among the affected individuals within the same family. This finding also suggests that RPA could be correlated with slower disease deterioration.
We identified variants in RLBP1 to be disease-causing in families F4 and F5. Retinaldehyde-binding protein 1 (RLBP1) on chromosome 15 (
In this study, we detected a novel frameshift variant in RLBP1 (c.398delC; p.Pro133GlnfsTer126) as disease-causing in family F5 (Table 4). This variant segregated in the affected members and was heterozygous in the unaffected sibling (Figure 1, Table 3, and Figure 3). Furthermore, we classified the ophthalmic outcome as likely RPA-causative.
In this study, we identified a novel pathogenic variant in RLBP1 (c.398delC; p.Pro133GlnfsTer126) and three recurrent variants in the RP1 and RLBP1 genes. We thoroughly investigated their consequent clinical manifestations using a battery of ophthalmic tests. We described heterogenic disease phenotypes of RP depending on the affected genes: typical RP for variants in RP1 and RPA for variants in RLBP1. This study also demonstrated that ES is a valuable ophthalmic molecular diagnostic tool and expands the causative variant spectrum in patients with RP. Genetic analyses enable us to perform targeted diagnostic testing and identify therapeutic approaches for gene therapy.
Appendix 1. Primer sequences for Sanger Sequencing.
Appendix 2. Fundus and OCT images for F1-F5 members.
Appendix 3. Pentacam images for the right (OD) and left (OS) eyes for families F1 to F5.
Appendix 4. Lipid profiling of family F1 members.
Appendix 5. Minor Allele Frequencies (MAF) and in silico predictions candidate variants.
Appendix 6. Amino acid conservation for candidate missense and frameshift variants across related species using Alamut Visual prediction software. The missense variants are TTPA (c.599C>T) and RP1 (c.607G>A). The frameshift variant is IFT140 (c.3955_3960delGCCAAG).
Supported by a grant from the Deanship of Academic Research at the University of Jordan (Grant No. 2013/39), Abdul Hameed
Shoman Foundation (Grant No. 2017/05), Amman, Jordan. Neither of the funders had any role in the data collection, analysis,
interpretation, or manuscript writing. Dr. Belal Azab (firstname.lastname@example.org