Novel 25 kb Deletion of MERTK Causes Retinitis Pigmentosa With Severe Progression

Abstract

Purpose: Retinitis pigmentosa (RP) describes a complex group of inherited retinal dystrophies with almost 300 reported genes and loci. We investigated the genetic etiology of autosomal recessive RP (arRP) in a large kindred with 5 affected family members, who reside on the island of Newfoundland, Canada.

Methods: Genetic linkage analysis was performed on 12 family members (Infinium HumanOmni2.5-8 BeadChip). Whole exome sequencing analysis (Illumina HiSeq) was performed on one affected individual. A custom pipeline was applied to call, annotate, and filter variants. FishingCNV was used to scan the exome for rare copy number variants (CNVs). Candidate CNVs subsequently were visualized from microarray data (CNVPartition v.3.1.6.). MERTK breakpoints were mapped and familial cosegregation was tested using Sanger Sequencing.

Results: We found strong evidence of linkage to a locus on chromosome 2 (logarithm of the odds [LOD] 4.89 [θ = 0]), at an interval encompassing the MERTK gene. Whole exome sequencing did not uncover candidate point mutations in MERTK, or other known RP genes. Subsequently, CNV analysis of the exome data and breakpoint mapping revealed a 25,218 bp deletion of MERTK, encompassing exons 6 to 8, with breakpoints in introns 5 (chr2:112,725,292) and 8 (chr2:112,750,421). A 48 bp insertion sequence was buried within the breakpoint; 18 bps shared homology to MIR4435-2HG and LINC00152, and 30 bp mapped to MERTK. The deletion cosegregated with arRP in the family.

Conclusions: This study describes the molecular and clinical characterization of an arRP family segregating a novel 25 kb deletion of MERTK. These findings may assist clinicians in providing a diagnosis for other unsolved RP cases.

Retinitis pigmentosa (RP) is a broad term meant to describe a complex group of hereditary retinal dystrophies. These dystrophies are characterized by the progressive degeneration of photoreceptor cells and/or retinal pigment epithelium, which typically leads to severe visual impairment or blindness by the fourth decade of life.¹ Symptoms of RP typically include early difficulty with night vision and dark adaptation, progressive loss of peripheral and/or central fields of vision, and declining visual acuity in the later stages. On examination of the fundus, RP often manifests with a triad of clinical features: a waxy pallor of the optic disc, attenuation of retinal blood vessels, and pathognomonic pigmentary aggregates called bone spicules.^2,3 There can be great variability in the clinical manifestation of RP, owing to the large number of causative genes and metabolic pathways affected.¹ Moreover, carriers of pathogenic RP mutations sometimes can display phenotypic variability within families, a phenomenon likely due to a combination of genetic modifiers and environmental influences.^4–6 Depending on the causal gene or mutation(s), patients with RP can experience onset of visual symptoms in a range from early childhood to mid-adulthood, with the progression and severity of disease influenced by the particular molecular pathways involved. The genetic etiology of inherited retinal dystrophies is complex, with nearly 300 genes and loci currently reported (RetNet, http:www.sph.uth.tmec.edu/RetNet/; in the public domain).⁷ The investigative approaches to uncovering these causal genes have been well-described.⁸ 

Mutations of the MERTK gene explain approximately 1% of cases of autosomal recessive RP (arRP).⁹ Pathologic mutations of this gene were first described in the Royal College of Surgeons (RCS) rat,¹⁰ and subsequently in human RP patients.¹¹ MERTK has an essential role in maintaining retinal photoreceptor viability by facilitating phagocytosis of shed photoreceptor outer segments.^12,13 In humans, MERTK mutations are a well-documented cause of arRP,^14–21 displaying a phenotype that often involves an early onset of symptoms (night blindness) within the first decade of life and a rapid progression and severity of disease. Mutations of MERTK occur along the entire length of the gene.²² Several point mutations and small indels in MERTK have been described, in addition to three large exonic deletions. These deletions include a 91 kb deletion encompassing exons 1 to 7,²³ a 9.86 kb deletion of exon 8,²⁴ and a 1.73 kb deletion of exon 15 with a complex re-arrangement.²⁵ 

We present findings of a fourth exonic deletion of MERTK; a 25 kb deletion encompassing exons 6 to 8, which was identified using whole exome data supplemented by copy number variant (CNV) and linkage analyses. The deletion segregates in a large consanguineous arRP family with five affected individuals, who live on the island of Newfoundland, Canada (Fig. 1). Affected individuals have an early-onset form of RP that affects peripheral and central fields of vision, and leads to a rapid decline in visual acuities and eventual blindness. 

Figure 1

 

A large arRP family from Newfoundland, Canada, segregating the 25 kb (c.845-1450del; p.Ala282_His483del) deletion of MERTK. Shaded symbols indicate clinically manifested RP. Screening for the deletion shows deletion carriers (wt/del), affected homozygotes (del/del), and wild-type noncarriers (wt/wt).

Materials and Methods

Patient Recruitment

Following ethical approval (HIC# 11.060), family members gave informed consent in adherence with tenets of the Declaration of Helsinki for participation in a research study. Genomic DNA was extracted from peripheral leukocytes using standard protocols for whole blood DNA extraction. The family was enrolled in the Finding of Rare Disease Genes in Canada (FORGE Canada) Consortium study. 

Genotyping and Linkage Analysis

Genomic DNA samples of 12 family members (Fig. 1; individuals IV-10, V-4, V-7, V-8, V-10, V-12, V-13, VI-2, VI-3, VI-8, VI-9, VI-10) were genotyped for 2,379,855 single nucleotide polymorphisms (SNPs), using the Infinium HumanOmni2.5-8 v1.0 BeadChip (Illumina, Inc., San Diego, CA, USA) according to the manufacturer's protocols at The Centre for Applied Genomics (Toronto, Canada). Briefly, 200 ng of DNA (4 μL at 50 ng/μL) was independently amplified, labeled, and hybridized to BeadChip microarrays and then scanned with default settings using Illumina iScan. Analysis and intrachip normalization of the resulting image files was performed using Illumina's GenomeStudio Genotyping Module software (v.2011) with default parameters. Genotype calls were generated using the Illumina-provided genotype cluster definitions file (HumanOmni2.5-8v1_C.egt) with a Gencall cutoff of 0.15. The CNVPartition v.3.1.6 module in GenomeStudio was used for CNV and (copy number neutral) loss of heterozygosity (LOH) analysis with default parameters. 

Following quality control and filtering protocols, 17,718 SNPs across the autosomes and X chromosome were suitable for linkage analysis. We assumed a recessive model of inheritance with a disease allele frequency of 0.1 and penetrance of 0.2, 0.2, and 99% for 0, 1, and 2 copies of the disease-causing allele, respectively. This corresponds to a disease prevalence of approximately 1%. A multipoint linkage analysis then was performed using a two-stage approach. First, the pedigree was broken down into two subpedigrees and then analyzed using an exact multipoint calculation. Subsequently, regions showing evidence of linkage were further analyzed using approximate methods and the full pedigree. 

Whole Exome Analysis and FishingCNV

The genomic DNA of one affected patient (Fig. 1; individual VI-9) was chosen for whole exome sequencing analysis. Whole exome capture and high-throughput sequencing was performed at the McGill University and Genome Québec Innovation Centre (Montréal, Canada). The Agilent SureSelect (Agilent Technologies, Santa Clara, CA, USA) 50 Mb All Exon Kit (V3) was used for target exome enrichment, and high-throughput sequencing was achieved using the Illumina HiSeq platform. An in-house annotation pipeline was applied to call and annotate sequence variants. Candidate homozygous variants were reviewed manually for evidence of pathogenicity. Copy number variants (CNVs) were identified in the whole exome dataset using FishingCNV.²⁶ Briefly, this program compares exon-level read counts in a test sample against a population of control samples to establish rare CNVs. Exon-level read counts were determined using GATK (v1.0) depth of coverage for individual VI-9, with comparison against 150 in-house control exomes. These exomes consisted of DNA samples that were sequenced during other rare disease gene projects carried out by FORGE, using the same target exome enrichment and sequencing protocols as per individual VI-9. 

Breakpoint Analysis and Carrier Screening

Following the identification of a candidate CNV in MERTK, the genomic position of the deletion breakpoints were mapped using primer walking. Primer sequences were designed using Primer3.²⁷ The primer sequences used to identify deletion breakpoints are as follows: Forward 1 primer (ACCTTCATCCTCACCACACC), Reverse 1 primer (TCAATTCCACAGCAGACACC), Forward 2 primer (CCTGCCCTCATCCATAAAGA), and Reverse 2 primer (GACACTGAAGCAGGGAGGTC). A PCR protocol was developed to capture deletion-specific amplicons in patients carrying the MERTK deletion (Supplementary Methods). Twenty family members were screened by this custom PCR protocol, to distinguish noncarriers from heterozygous and homozygous deletion carriers (Supplementary Fig. S1). To mitigate the possibility of PCR reactions yielding false-negative results, we repeated each reaction in triplicate for every individual tested. Sequence chromatograms for deletion carriers were inspected visually for alignment to regions of MERTK according to the GRCh38.p3 human reference genome. 

Results

Affected patients of this family were initially seen by GJJ and JSG in 1983, after referral to an Ocular Genetics clinic. Referrals initially were received for three separate families; however, genealogical analysis later revealed a shared and complex ancestry (Fig. 1). 

Individual V-11 (Fig. 1) was first seen in 1975 at age 28 with central vision loss. Visual acuities were 20/120 in both eyes at age 29, and by age 30, vision was reduced to hand movements only. At age 35, posterior capsular cataracts developed in the left eye, and funduscopy revealed pigmentary clumping, but no pallor of the optic discs or attenuation of retinal blood vessels. At 36, his fundus had a beaten-bronze appearance, with granular gray and whitish flecks in the midperiphery. Many bone spicules also were seen in the periphery. Individual VI-10 (Fig. 1) reported that he was always night blind. At age 11 (1984), he had temporal constriction of the left visual field and a full field of vision in the right eye. Visual acuities were 20/20 in both eyes, and he had a color vision defect with 10/20 errors using the Ishihara test. His fundus had an albinoid appearance, with a marked decrease in chorioretinal pigmentation, and no bone spicules were found on examination. The following year at age 12, visual acuities decreased to 20/25 in both eyes, and his fundus had a salt and pepper appearance. Findings suggestive of delayed dark adaptation and nyctalopia were noted at this time. At age 16, visual acuities in both eyes further decreased to 20/40, and he reported difficulties in school, having trouble seeing the blackboard. At this time, his color vision continued to deteriorate, such that he scored 19/20 errors on the Ishihara test, and had strong red-green and blue-yellow defects, seeing only one plate using the Hardy, Rand, and Rittler Pseudoisochromatic plate test. By age 19, he had a large central scotoma in both eyes, with peripheral vision only (nasally). By age 37, he had central scotomas to 30° to 40° in both eyes, and visual acuities were confined to hand movements only. The fundus displayed extensive chorioretinal degeneration. Alternating exotropia was noted. Individual VI: 3 (Fig. 1) first experienced a decrease in central vision at age 7. The defect rapidly progressed, such that by age 12 the patient was enrolled at a school for the blind. Visual acuities continued to decrease, and color vision deficits were noted, as well as decreased night vision. The patient reported being sensitive to light for as long as he could remember. In 1980, at the age of 30, the patient had typical changes of RP, with macular degeneration as well. Visual acuities were reduced to hand movements only in the right eye, and light perception in the left. At age 35, fundi showed dense bone spicules at the equator, with an atrophic area in the midperiphery with grayish spots. At the macula, there was a small atrophic area with the appearance of choroidal sclerosis. Optic discs had a gray pallor and retinal blood vessels were narrowed. Visual acuities further decreased to light perception in both eyes. The patient had a marked exotropia, mainly of the left eye, which did alternate at times. Individual VI-9 (Fig. 1) was first seen in 1986 (age 14) as his younger brother (VI-10) already was experiencing symptoms. At this time, VI-9 had 20/20 visual acuities in both eyes, and decreased chorioretinal pigmentation and stippling was noticed on funduscopic examination. He stopped playing hockey at age 15 due to deteriorating vision, and reported he was always night blind. By age 17, uncorrected visual acuities were 20/80 in the right eye and 20/40 in the left. He had difficulties with classes and left school by grade 12. At age 20, he had developed a significant loss of central vision, such that only peripheral islands of vision (nasally) remained. Visual acuity was reduced to counting fingers only by age 30, and the patient subsequently lost all vision. Funduscopic surveillance of this patient at age 44 highlights the classic features of RP, including bone spicules, attenuation of retinal vessels, and pallor of the optic disc (Fig. 2). A summary of the clinical findings of this family is provided in Supplementary Table S1. 

Figure 2

 

Fundus imaging of RP patient (VI-9), showing characteristic features of RP, including bone spicules, attenuation of retinal blood vessels, and pallor of the optic disc, with significant atrophy of the retina. (A, B) Patient's left eye. (C, D) Patient's right eye.

A genetic linkage analysis of 12 family members (Fig. 1: individuals IV-10, V-4, V-7, V-8, V-10, V-12, V-13, VI-2, VI-3, VI-8, VI-9, VI-10) was performed to search for evidence of a disease-susceptibility locus in the family. Owing to the large size of the pedigree, we first performed an exact multipoint linkage analysis of subpedigrees, which revealed six regions on five chromosomes displaying logarithm of the odds (LOD) scores above 1.5 (Supplementary Table S2). A maximum LOD score of 3.71 (θ = 0) was observed on chromosome 2 encompassing markers rs7584136 and rs10177102 (chr 2:105833093-121145352). An approximate linkage analysis of these chromosomes subsequently was performed using the full pedigree, which increased the maximal LOD score to 4.89 (θ = 0), indicating strong evidence of linkage at chr2: 107493123-117426163 (Supplementary Table S3; Supplementary Fig. S2). The region covered markers rs4676093 and rs4525709 (Chr 2: 107493123 - 117426163), a 7.5 cM (10 Mb) interval that includes the MERTK gene. 

Next, we sequenced the exome of one affected patient (Fig. 1; individual VI-9), to screen for point mutations, small indels, and CNVs. In filtering the whole exome data, we included variants with minor allele frequency <5% and excluded variation that was synonymous or in untranslated regions (UTRs). This strategy produced 541 candidate variants; 32 of which were homozygous. Each homozygous variant then was assessed for evidence of pathogenicity by reviewing protein function, predicted mutation consequences, and evolutionary conservation scores. None of the 32 variants was in genes known to cause RP, in the interval with maximal LOD score, or appeared functionally relevant to an ocular phenotype (data not shown). Therefore, our exome filtering strategy did not produce any obvious candidate point mutations. 

To screen the exome for candidate CNVs, we next applied FishingCNV. This program yielded 9 candidate CNVs after Holm-Bonferroni correction (P < 0.05). These candidates included amplifications or deletions in ZFHX3, MERTK, AAK1, SMARCAL1, CCDC7, GIT2, NAALAD2, ZNF658, and FGGY (Supplementary Table S4). Among these, the CNV in MERTK was the only candidate functionally relevant to an ocular phenotype and also situated in a linked interval of LOD ≥ 1. This was a significant finding, given that aberrations of MERTK are known to cause RP, and this locus had the strongest evidence of linkage in the family, demonstrating a significant LOD score of (4.89 [θ = 0]). The CNV in MERTK is a large 25 kb deletion spanning chr2:112725714-112740570 and encompassing exons 6, 7, and 8. It is a novel deletion, which to our knowledge, is not reported in the literature or online databases. Thus, we identified a strong candidate mutation for arRP in this family; a novel in-frame deletion of 25 kb in the MERTK gene (c.845-1450del; p.Ala282_His483del), which removes 201 amino acids from the encoded protein, corresponding to two fibronectin type-III domains. 

To further investigate this mutation, we screened other family members for the deletion, by re-examining microarray data generated during the linkage analysis. This corroborated a heterozygous MERTK deletion in obligate carriers (Fig. 1: individuals IV-10, V-4, V-13), and a homozygous deletion in the affected individuals with RP (Fig. 1; V-10, VI-3, VI-9, VI-10), and no unaffected individuals were homozygous for the deletion. We then proceeded to map the breakpoints of the segregating MERTK deletion using primer walking. The 5′ breakpoint is located in intron 5 (chr2:112,725,292), while the 3′ deletion breakpoint is in intron 8 (chr2:112,750,421). The sequence surrounding the 5′ breakpoint maps to an L2B repeat of the LINE-2 family, while the 3′ breakpoint maps to a MER21A repeat sequence. The deletion spans a total of 25,218 base pairs (bp) according to the GRCh38.p3 human reference genome. 

We next optimized a custom PCR protocol to validate the MERTK deletion and screen additional family members, by amplifying deletion-specific fragments (Supplementary Methods). In brief, the PCR protocol is able to differentiate DNA samples with wild-type breakpoint sequences from heterozygous or homozygous deletion carriers (Supplementary Fig. S1), and, therefore, may prove useful for others who wish to screen for this deletion in unsolved arRP families. A total of 20 family members were screened and validated in this manner. As expected, the deletion was homozygous in affected individuals and heterozygous in obligate carriers (Fig. 1), thereby replicating our previous findings. Screening identified a total of 8 carriers in the family, who did not have a phenotype similar to their homozygous affected cousins, and were not routinely followed by ophthalmology. 

Interestingly, Sanger sequencing of the deletion-specific PCR amplicons revealed a 48 bp insertion sequence within the deletion breakpoints (Fig. 3). A Basic Local Alignment Search Tool (BLAST) search of this sequence (ATTACTAGGTGAAGCACAGTGGAGCACATGGCTTGGTATAGGAGACCC), showed that the terminal 30 bp (GTGGAGCACATGGCTTGGTATAGGAGACCC) have a 97% identity to intron 5 of MERTK. The remaining proximal 18 bp of the insertion (ATTACTAGGTGAAGCACA) mapped with 100% identity to intronic regions of MIR4435-2HG and LINC00152 (Fig. 4). MIR4435-2HG is a microRNA approximately 700 kb upstream of MERTK, while LINC00152 is a long noncoding RNA on the opposite arm of chromosome 2. 

Figure 3

 

Novel 25 kb deletion of MERTK detected in WES data (A) and validated using Sanger sequencing (B). Validation of the deletion breakpoints (B) revealed a 48 bp insertion sequence in this affected family member.

Figure 4

 

Diagrammatic representation of 25 kb MERTK (c.845-1450del, p.Ala282_His483del); an in-frame deletion encompassing exons 6 to 8, with a 48 bp insertion sequence at the breakpoints. Red font, the 5′ and 3′ breakpoints of the deletion; green font, 18 bp of insertion with homology to MIR4435-2HG; blue font, 30 bp region of homology to intron 5 of MERTK.

Discussion

This study provided compelling evidence for the discovery of a novel 25 kb deletion of MERTK causing arRP in this large family from the island of Newfoundland. We discuss our rationale for these findings below. First, we observed strong evidence of linkage disequilibrium at an interval that encompasses MERTK (LOD 4.89 [θ = 0]). Meanwhile, our whole exome filtering strategy did not uncover strong candidate mutations in known or putative RP genes. Subsequently, CNV analysis then uncovered a large exonic deletion of MERTK within the interval with the strongest linkage signal. Analysis of genotype data and a custom PCR protocol corroborated the segregation of this deletion in a pattern typical of autosomal recessive inheritance. Moreover, deletion of exon 8 is known to cause RP,²³ and the deletion identified here encompasses exons 6 to 8. The large size of our study family provides important insights into the clinical manifestation of arRP caused by MERTK mutations. 

The family is characterized by an early onset form of RP. Affected individuals all reported decreased or absent night vision early in life, and subsequently experienced a rapidly progressive deterioration of central and peripheral vision. Most individuals were blind by their 20th to 30th year. Deterioration of color vision also was noted in these patients, and funduscopy revealed classic changes of RP, with some variability in the detection of bone spicules (Supplementary Table S1). Indeed, other patients with MERTK mutations often experience earlier onset of symptoms compared to other forms of arRP. 

This family segregates only the fourth exonic deletion of MERTK to be reported to our knowledge. The 25 kb deletion (c.845-1450del; p.Ala282_His483del), encompasses exons 6 to 8 and overlaps portions of two other MERTK deletions reported in the literature: one encompassing exons 1 to 7 (91 kb),²⁴ and the other a deletion of exon 8 (9.86 kb).²³ Interestingly, some clinical overlap is seen in each of these families carrying MERTK deletions. For example, patients carrying the 91 kb deletion encompassing exons 1 to 7 also experienced onset of symptoms in the first decade of life, which led to deterioration of central and peripheral vision; a finding that also was seen in our patients. As well, authors of that study noted interindividual variation within their family with respect to retinal vessel attenuation and optic disc pallor. Interindividual variation is seen within our family as well, as individual V-11 (Fig. 1) did not display any obvious attenuation of retinal vessels or pallor of the optic disc, while these features were prominent in individual VI-3. It remains unclear if these observations are due to differences in disease manifestation or thoroughness of surveillance. We also observed variability between two affected individuals of the same sibship (Fig. 1: individuals VI-10, VI-9). Individual VI-10, the younger brother, appeared to have a more severe onset and progression of disease, as he began noticing decreased peripheral vision by age 11. Meanwhile, his older brother did not clinically manifest symptoms until age 14. Moreover, individual VI-3, a more distant relative, manifested symptoms at age 7 and experienced significant visual deterioration such that he was registered at a school for the blind by age 12. Thus, the phenotype seen in our family is one that is severe and early in onset, with first reported decline in visual acuities often by the second decade, ranging from age 7 to age 30. 

In patients segregating the MERTK exon 8 deletion,²³ the authors noted macular involvement, as the proband of their study had central vision deficits by age 13. Additionally, affected family members of their study experienced progressive color vision defects. For example, the proband of their study failed a color vision test with 24/24 errors at age 26, while the younger sibling, who also was affected, began experiencing mild generalized dyschromatopsia by age 8. Early color vision deficits also were noted in our study, as individual VI-10 experienced significant deficits at age 11 (10/20 errors), which markedly progressed by age 16 (19/20 errors). Moreover, individual IV-3 of our study also had early color vision deficits, documented at age 12, suggestive of early rod and cone involvement. 

A peculiar finding of this study was the discovery of a 48 bp insertion sequence buried within the deletion breakpoint. In patients with the previously reported exon 8 deletion of MERTK, the authors uncovered a complete AluY element insertion, leading them to hypothesize nonhomologous recombination between two AluY elements as the mechanism of the deletion. Intrigued by those findings, we performed a BLAST search of our 48 bp insertion sequence, but did not find homology to Alu elements. We did, however, find homology to three intronic regions on chromosome 2. Shown in Figure 4, 30 bps of the insertion sequence map to a stretch of intron 5, which is deleted in these patients, and appears to be the reverse complement compared to wild-type intron 5 reference sequences. The remaining 18 bp portion of the insertion sequence maps to two other intronic loci on chromosome 2: MIR4435-2HG and LINC00152. One possible explanation for these findings could be the insertion of a transposable element buried within one of these genes. The etiology of these exonic MERTK deletions appears complex, given that the third reported MERTK deletion is a removal of exon 15 (1732 bp), with a duplication and inversion event,²⁵ with no mechanism postulated for those findings. 

The deletion uncovered in this study (c.845-1450del; p.Ala282_His483del) removes 201 amino acids from encoded MERTK protein, corresponding to the removal of two fibronectin type-III domains. These domains function in the extracellular region of the protein, and together with the immunoglobulin domains, define MERTK within the TAM receptor tyrosine kinase family.²⁸ This deletion does not disrupt the reading frame, given that the 484th amino acid is retained as a glycine, rendering nonsense-mediated decay unlikely. It is likely, however, that the removal of the fibronectin domains could structurally alter the extracellular region of the MERTK protein, impairing its ability to bind GAS6 ligands. Binding of GAS6 ligands to MERTK is functionally important, as it is known to stimulate phagocytosis,^12,29 and loss of this ability could lead to the observed accumulation of photoreceptor outer segment debris in the subretinal space. 

This study highlights the importance of using alternative methods to screen arRP patients who have early onset and severe forms of RP. MERTK currently is believed to account for 1% of arRP cases; however, these numbers were estimated using a sequencing study that screened a cohort of 96 arRP cases.⁹ Therefore, iIt remains possible that the true proportion of arRP caused by MERTK mutations may differ when taking potentially undetected CNVs into account. Therefore, investigators should consider screening this gene for larger deletions in unsolved arRP families. The identification of novel pathogenic mutations and structural alterations remains an important process, particularly for the MERTK gene, as ongoing clinical trials continue to highlight the potential for the future application of gene therapy,³⁰ which will first require patients to have a genetic diagnosis. 

We concluded that the discovery of this 25 kb MERTK deletion, abolishing exons 6 to 8, is a novel cause of arRP which ultimately leads to an early onset and rapid progression of visual symptoms seen in this family. We provided an optimized PCR protocol (Supplementary Methods) that others might use to detect this deletion in unsolved arRP families. The clinical histories provided here are of benefit to clinicians and investigators in understanding the clinical manifestation of MERTK deletions, and this is one of the largest families segregating a MERTK deletion currently reported in the literature. Moving forward, we encourage others to consider screening MERTK thoroughly for larger CNVs in the investigation of unsolved arRP cases. 

Acknowledgments

The authors thank our study participants, as their assistance in providing clinical histories and DNA samples was an integral part of this study. They also acknowledge the contribution of the high throughput sequencing platform of the McGill University and Genome Québec Innovation Centre, Montréal, Canada. The authors also thank The Centre for Applied Genomics, The Hospital for Sick Children (Toronto, Canada) for their assistance with technical expertise with microarray analysis. 

Supported by the Government of Canada through Genome Canada, the Canadian Institutes of Health Research and the Ontario Genomics Institute (OGI-049). Additional funding was provided by Genome Québec, Genome British Columbia, and the McLaughlin Centre, as well as the Research & Development Corporation of Newfoundland & Labrador (RDCNL). 

Disclosure: D.R. Evans, None; J.S. Green, None; G.J. Johnson, None; J. Schwartzentruber, None; J. Majewski, None; C.L. Beaulieu, None; W. Qin, None; C.R. Marshall, None; T.A. Paton, None; N.M. Roslin, None; A.D. Paterson, None; S. Fahiminiya, None; J. French, None; K.M. Boycott, None; M.O. Woods, None 

Appendix

FORGE Canada Consortium Steering Committee, consisting of Kym Boycott (leader; University of Ottawa), Jan Friedman (co-lead; University of British Columbia), Jacques Michaud (co-lead; Université de Montréal), Francois Bernier (University of Calgary), Michael Brudno (University of Toronto), Bridget Fernandez (Memorial University), Bartha Knoppers (McGill University), Mark Samuels (Université de Montréal), and Stephen Scherer (University of Toronto). Janet Marcadier (Clinical Coordinator) and Chandree Beaulieu (Project Manager).