Identification of Novel Mutations in the Ortholog of Drosophila Eyes Shut Gene (EYS) Causing Autosomal Recessive Retinitis Pigmentosa

Abstract

Purpose.: Recently, a novel gene was cloned for autosomal recessive retinitis pigmentosa (arRP), EYS, on 6q12. This study was conducted to determine the spectrum and frequency of EYS mutations in 195 unrelated patients with autosomal recessive and autosomal dominant RP (adRP).

Methods.: All cases had a complete ophthalmic examination, and the clinical diagnosis of RP was based on visual acuity, fundus photography, and electroretinography findings. The DNA extracted from all participants was subjected to molecular genetic analysis entailing amplification of the coding regions and exon–intron boundaries of EYS by polymerase chain reaction, followed by direct sequencing. Bioinformatics analysis was undertaken to study the effect of the identified mutations on protein structure and function.

Results.: Eleven novel missense, nonsense, and splice site mutations were identified within EYS in 10 unrelated arRP patients, with probable allele frequency of 11%. However, no mutations were observed in the adRP panel. In addition, 53 single-nucleotide polymorphisms (SNPs) were found, of which 12 were previously unreported. Bioinformatics analyses revealed that all mutations were highly conserved across other species and/or involved important domains on protein structure. Intrafamilial phenotypic variability was also observed in a family with double heterozygous mutations.

Conclusions.: This is the first report of molecular genetic analysis of EYS in a cohort of unrelated British and Chinese patients with RP. The results further the initial hypothesis that EYS is a major causative gene for recessive RP and emphasize the role of different types of mutations in disrupting the function of EYS.

Progressive degeneration of rod photoreceptor cells is the hallmark of retinitis pigmentosa (RP; OMIM 268000; Online Mendelian Inheritance in Man; http://www.ncbi.nlm.nih.gov/Omim/ provided in the public domain by the National Center for Biotechnology Information [NCBI], Bethesda, MD). A disease known by extreme genetic and allelic heterogeneity, RP has an incidence of ∼1:3500, with the recessive form the most prevalent worldwide. ¹  

To date, 29 loci have been implicated in the etiology of autosomal recessive (ar)RP (http://www.sph.uth.tmc.edu/retnet/ RetNet: http://www.sph.uth.tmc.edu/RetNet; provided in the public domain by the University of Texas Houston Health Science Center, Houston, TX); accounting for ∼10% to 15% of all arRP cases. ² Exceptionally, the RP25 locus was identified as the genetic cause of 10% to 20% of arRP cases in Spain. ³ Later, it was also mapped in multiple families from various ancestral origins. ^4,5  

Recently, we have identified the RP25 gene EYS (an orthologue of Drosophila eyes shut [EYS])/spacemaker [SPAM]), on chromosome 6, region q12, spanning ∼2 Mb of the genomic DNA and encoding the protein SPAM. ⁶ Initially, six independent mutations were identified within EYS in five unrelated Spanish families with arRP. ⁶ Independently, two additional mutations were reported in three unrelated arRP patients of Dutch origin. ⁷  

The mutations identified in EYS to date include deletions, frame-shift mutations, and nonsense substitutions, all leading to premature truncation codons (PTCs). ^6,7 The position of PTCs in most of these mutations occur upstream of the last exon. Hence, nonsense-mediated decay, causing quick degradation of the protein, ⁸ could be the underlying mechanism of RP in these patients. 

The absence of missense substitutions or splice site changes within EYS, however, indicates that these types of mutations could be either population specific, extremely rare, or identifiable in another form of the disease, such as the autosomal dominant (ad)RP. This finding is not uncommon, since different mutations in previously reported genes such as RHO and RP1 have been identified as the genetic cause of both adRP and arRP through different mechanisms. ^9–11 Overall, the number of mutations identified thus far supports the concept that EYS is the primary major gene responsible for recessive RP; however, prevalence studies are essential for evaluating this assumption as well as for identifying whether EYS is also responsible for adRP. 

In this study, our objectives were to investigate the frequency and mutational spectrum of EYS in a panel of patients with ad- and arRP. Bioinformatics analysis of the identified mutations has been also performed to study their effect on the protein function. Finally, we sought to describe any specific phenotypes associated with EYS mutations in our patients. 

Materials and Methods

Patients and Clinical Evaluation

One hundred ninety-five British and Chinese patients with RP (100 with arRP and 95 with adRP), with no systemic manifestations, participated in the study. In addition, at least 200 control individuals of Caucasian and Chinese origins were included. An informed consent was obtained from all participants for clinical and molecular genetic studies. The research adhered to the tenets of the Declaration of Helsinki. 

All cases were seen at Moorfields Eye Hospital with the exception of three Chinese patients who were examined at the Department of Ophthalmology and Visual Sciences in Hong Kong, as previously published. ⁵ A full ophthalmic examination was performed, and the clinical diagnosis was based on visual acuity, fundus photographs, and electroretinography (ERG) findings. The clinical phenotype of one of the Chinese families segregating a mutation within EYS was revisited. 

Molecular Diagnosis

DNA Extraction and PCR Amplification.

Genomic DNA was extracted from peripheral blood lymphocytes according to the standard procedures. The coding regions, together with the splice sites (GT/AG) and the 5′ UTR of EYS were amplified by polymerase chain reaction (PCR) with our previously published primers. ⁶ In addition, we designed one pair of primers (5′-3′), F: ccaagaagcatcagccttgt and R: ggtactaagagaccccgttcaa, with the Primer 3 Output program (http://frodo.wi.mit.edu/primer3/) ¹² to screen the alternatively spliced exon 42 published by Collin et al. ⁷ Each PCR was performed in 25-μL reaction mixture containing genomic DNA (100 ng), primers (0.4 μM each), MgCl₂ (1.5–2.5 mM), deoxynucleoside triphosphate (dNTPs; 0.2 mM), 1× PCR buffer (Bioline, Luckenwalde, Germany), and Taq polymerase (0.5 U; Bioline). Amplification reactions were performed under the following conditions: 3 minutes of denaturation at 94°C followed by 35 cycles of denaturation at 94°C for 30 seconds, annealing at 57°C to 65°C for 30 seconds, extension at 72°C for 30 seconds, and a further extension step at 72°C for 5 minutes. For GC-rich regions, buffer 3 (containing 22.5 mM MgCl_2; supplied by Roche Diagnostics GmbH, Mannheim, Germany) together with dimethyl sulfoxide DMSO 5% (Sigma-Aldrich, Poole, UK) or PCR master mix (Ready mix AB-0795; ABgene, Epsom, UK) was used. Also, DNA polymerase (KOD Hot Start; Novagen-Merck, Nottingham, UK) was used whenever necessary. 

Mutation Detection.

Initially, the PCR products were purified by adding shrimp 1 U alkaline phosphatase (SAP; Exosap Q; Amersham LifeScience, Buckinghamshire, UK; ExoSAP-IT, USB Corp. Cleveland, OH) and 1 U exonuclease I (USB Corp.) to 1 μL of the PCR product and incubated at 37°C for 15 minutes and then at 80°C for another 15 minutes to deactivate the enzyme. A sequencing reaction was then performed as previously described. ¹³ Direct sequence analysis of all participants was performed on the automated fluorescence DNA sequencer (ABI 3730; Applied Biosystems, Foster City, CA), according to the manufacturer's instructions. Subsequently, the data were analyzed (SeqMan 4.03 software; DNASTAR, Inc., Madison, WI). 

Mutation Nomenclature.

All changes were assigned a nucleotide number starting at the first translation base of EYS according to the GenBank entry NM_001142800.1 (http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). 

Confirmation of Mutations.

Since all the identified mutations did not create or abolish any of the restriction enzyme sites, direct sequence analysis was used to test the status of these mutations in at least 400 ethnically matching control chromosomes. 

Splice Site Scores and Splice Site and Exonic Splicing Enhancer (ESEs) Prediction.

Splice site scores were calculated with Splice Site Score Calculation software (http://rulai.cshl.edu/new_alt_exon_db2/HTML/score.html/ an opensource web-based program). We also searched for the presence of ESE motifs within EYS exons with missense mutations, using the ESE finder software (http://genes.mit.edu/burgelab/rescue-ese/ provided in the public domain by the Massachusetts Institute of Technology, Cambridge, MA). 

Multiple Sequence Alignment.

To assess the significance of the identified changes, we performed an alignment of the amino acid sequence of human SPAM with that from other species using CLUSTAL W, ver. 1.82, software (provided in the public domain by European Bioinformatics Institute, European Molecular Biology Laboratory, Heidelberg, Germany; available at http://www.ebi.ac.uk/clustalw/). 

Results

Clinical Examination

The clinical data of the three Chinese patients, belonging to families RP10, RP112, and RP116, were showing a typical picture of RP, as previously described. ⁵ However, in the RP10 family where a compound heterozygous mutation was identified, one member (III-3, Fig. 1) had no RP symptoms, and she also had a normal fundus appearance until the age of 29 (Figs. 2A, 2B) but showed extinguished ERGs. Her clinical appearance is in comparison to her mother (II-2; Fig. 1) whose fundus presented typical signs of RP (Figs. 2C, 2D). This variability in the clinical picture between the mother (II-2) and her daughter (III-3) could be due to the different stages of the disease. However, this example was not the only one of clinical variability in this family since the age of onset of RP in the mother (II-2) and her brother (II-4; Fig. 1) was at 43 and 36 years, respectively, in comparison to their sister (III-5) who had RP symptoms by the age of 20 years. Therefore, this family could represent an example of RP with intrafamilial phenotypic variability. 

Figure 1.

 

A pedigree of a Chinese family (RP10) showing the segregation of a compound heterozygous change, c.6416G>A and c.6557G>A, within EYS in all members. The pseudodominant appearance in this family may be attributable to undeclared consanguinity between family members II-1 and II-2.

Figure 2.

 

Fundus photographs of a Chinese family (RP10) with arRP showing phenotypic variability between affected members. (A) Right and (B) left eyes of the proband's affected daughter at the age of 29 years, revealing a normal fundus appearance. (C) Right and (D) left eyes of the proband aged 60 years showing bone-spicule–like pigmentation in the periphery of the retina typical of RP.

All remaining patients had a classic disease course and a clinical picture typical of RP, including variable grades of low visual acuity, waxy pallor of the optic disc, attenuation of the retinal blood vessels, bone spicule pigmentation, and flat or absent ERGs. The pattern of inheritance in all participating patients was diagnosed based on the pedigree structure. 

Mutation Screening

EYS Sequencing in 100 Patients with arRP.

Sequence analysis of EYS led to the identification of 11 novel mutations in 10 unrelated patients with arRP (Table 1). Four of the identified mutations were compound heterozygous, one was homozygous, and the remaining six were heterozygous for which the second mutant allele escaped detection. 

Table 1.

 

Summary of the Mutations Identified in EYS in 100 arRP Patients

Family ID/Patient No.	Exon	DNA Change	Protein Change	Type of Protein Change	Type of Mutation	Status of the Change
RP10/II-2	31	[c.6416G>A+c.6557G>A]	p.C2139Y	Sulfhydryl→hydroxyl	Missense	Heterozygous
	32		p.G2186E	Non polar→acidic	Missense	Heterozygous
RP28	11	[c.1765A>G c.7665C>G]	p.R589G	Basic→nonpolar	Missense	Heterozygous
	39		p.Y2555X	Truncated protein	Nonsense	Heterozygous
RP56	22	c.3443+1G>A	NA	NA	Splice site	Homozygous
RP32	36	c.7205G>A	p.R2402K	Basic→basic	Missense	Heterozygous
RP42	44	c.8492C>T	p.T2831I	Hydroxyl→non polar	Missense	Heterozygous
RP44	44	c.9194G>T	p.W3065L	Non polar→non polar	Missense	Heterozygous
RP23	40	c.7810C>T	p.R2604C	Basic→uncharged	Missense	Heterozygous
RP67	44	c.8492C>T	p.T2831I	Hydroxyl→non polar	Missense	Heterozygous
RP90	41	c.8054G>A	p.G2685E	Non polar→acidic	Missense	Heterozygous
RP7	8	c.1299+3A>C	NA	NA	Splice site	Heterozygous

All compound heterozygous changes were missense (p.C2139Y, p.G2186E, and p.R589G) apart from a single nonsense change (p.Y2555X). The first compound heterozygous mutation p.[C2139Y]+[G2186E] was identified in the proband of one of the previously linked Chinese families (RP10, Table 1). It occurred in the form of c.6416G>A (Figs. 3A, 2B) and c.6557G>A (Figs. 3C, 2D) transitions at the second nucleotide position of codons 2139 and 2186, respectively. Both changes segregated with the disease phenotype in the studied family (Fig. 1) and were not detected in 192 ethnically matching control chromosomes. Even though these changes led to nonconservative amino acid substitutions they were also found to be located within ESEs in exons 31 and 32 of EYS. 

Figure 3.

 

Electropherograms of EYS mutations in the arRP patients and the corresponding normal sequence in control subjects. (A, C) Control sequences at nucleotide positions c.6416 and c.6557, respectively; (B, D) the heterozygous c.6416G>A and c. 6557G>A substitutions in the proband of family RP10, respectively; (E, G) normal sequences at positions c.1765 and c.7665, respectively; (F) heterozygous c.1765A>G and (H) c.7665C>G nucleotide substitutions in the proband of family RP28; (I) normal sequence and (J) homozygous c.3443+1G>A in the proband of family RP56; (K) control sequence and (L) heterozygous c.7205G>A in the proband of family RP32; (M) normal sequence and (N) c.8429C>T heterozygous substitution in the proband of family RP42; (O) normal control and (P) c.9194G>T heterozygous replacement in the proband of family RP44; (Q) normal sequence and (R) heterozygous c.7810C>T in the proband of family RP23; (S) control sequence and (T) heterozygous c.8054G>A exchange in the proband of family RP90; and (U) normal sequence and (V) heterozygous c.1299+3A>C in the proband of family RP7.

In a sporadic patient (RP28) of Caucasian origin, another double heterozygous mutation, p.[R589G]+[Y2555X], was observed (Table 1). The p.R589G is not only predicted to result in a nonconservative amino acid substitution that would be likely to affect SPAM function, but also lies in the splice site at the end of exon 11 (Figs. 3E, 3F). Calculation of the splice site scores showed a reduction from 6.2 for the natural site to 4.0 for the variant site, indicating that this alteration may affect splicing. The second heterozygous change in this patient was a transversion at nucleotide position 7665, resulting in replacement of a tyrosine at codon 2555 with PTC (Figs. 3G, 3H). 

A fifth mutation, a homozygous splice donor site change within intron 22 at c.3443+1G>A (Figs. 3I, 3J), was identified in a sporadic patient (RP56) of Caucasian origin (Table 1). The c.3443+1 mutation was found only in this patient and not in any of the control subjects, suggesting that it might affect splicing. This notion was supported by the calculated splice site scores: 5.7 for the normal splice site and −5.0 for the variant site. 

The remaining six mutations were heterozygous; for each, a second mutant allele was not found. Of these, three mutations, p.W3065L (Figs. 3O, 3P) p.R2604C (Figs. 3Q, 3R), and p.G2685E (Figs. 3S, 3T) were identified in three sporadic subjects of Caucasian origin (Table 1). The fourth mutation, p.T2831I (Figs. 3M, 3N), was identified in two patients (RP42 and RP67) of Caucasian and Asian origins, respectively (Table 1). The last two changes, p.R2402K (Figs. 3K, 3L) and the predicted splice site change (c.1299+3A>C; Figs. 3U, 3V) were found in two sporadic arRP patients of Asian origin (Table 1). The nucleotide substitution of A>C at c.1299+3 reduced the calculated splice site score from 9.7 to 3.0, and hence it probably affected the splicing. 

In addition, we found 53 SNPs, of which 12 were previously unreported (Table 2). 

Table 2.

 

Summary of the SNPs Identified in Mutation Screening of EYS

Exons	Nucleotide Position	Amino Acid Change	Reported/Novel	Allele Frequency (%)
1	c.-546−81A>C	—	Novel	2.0
	c.-508A>G		rs1490127	—
4	c.334G>A	p.V112I	Novel	1.0
	c.359C>T	p.T120M	rs12193967	—
	c.748+52 T>C	—	Novel	5.0
5	c.862+87 T>C	—	rs4710522	—
6	c.863−70A>G	—	Novel	1.0
	c.863−22ins TT	—	rs34154043	—
	c.911delT	p.304W>G	rs34676630	—
7	c.1146T>C	p.N382N	rs974110	—
9	c.1300−3C>T	—	rs1936439	—
	c.1459+103 C>T	—	rs9453265	—
10	c.1596A>C	p.K532N	rs61753611	—
	c.1599+96A>C	—	rs1502963	—
11	c.1600−79A>G	—	rs1502964	—
	c.1600−38G>A	—	rs1502965	—
	c.1712A>G	p.Q571R	rs61753610	—
	c.1766+61A>G	—	Novel	15
12	c.1809C>T	p.V603V	rs9345601	—
	c.1891G>A	p.G631S	rs9342464	—
	c.1922A>T	p.E641V	rs17411795	—
13	c.2137+114C>T	—	rs10455568	—
14	c.2157C>T	p.C719C	rs9453148	—
	c.2259+10C>A	—	Novel	5
16	c.2382−26C>G	—	rs9445437	—
	c.2555T>C	p.L852P	rs9294631	—
17	c.2733>T>C	p.N911N	Novel	10
18	c.2846+53ins TAAT	—	rs59518422	—
19	c.2847−24C>T	—	rs7743515	—
23	c.3444−5C>T	—	rs94445051	—
25	c.3787A>G	p.I1263V	rs17404123	—
26	c.3906C>T	p.H1302H	rs12663916	—
	c.3936A>G	p.T1312T	rs12662610	—
	c.3973C>G	p.Q1325E	rs12663622	—
	c.4026C>T	p.S1342S	rs12663619	—
	c.4081A>G	p.I1361V	rs17403955	—
	c.4256T>C	p.L1419S	rs624851	—
	c.4352T>C	p.I1451T	rs62415828	—
	c.4543C>T	p.R1515W	rs62415827	—
	c.4549A>G	p.S1517G	rs62415826	—
	c.4593G>A	p.E1531E	rs62415826	—
27	c.5705A>T	p.N1902I	rs9353806	—
28	c.5927+116T>C	—	Novel	43
29	c.5928−35C>T	—	rs587278	—
	c.6048+68A>G	—	Novel	60
34	c.8634+61T>G	—	Ens5NP7485951	—
35	c.6835−64C>T	—	rs1482457	—
	c.6977G>A	p.R2326Q	rs4710457	—
39	c.7723+64T>A	—	Novel	10
	c.7655G>C	p.V2553A	Novel	2
	c.7666A>T	p.S2556C	rs66462731	—
40	c.7800 A>G	p.P2600P	Novel
41	c.8071+84T>G	—	rs4710257	—

EYS Sequencing in 95 Patients with adRP.

Mutation screening of EYS did not reveal any pathogenic mutations in any of the adRP patients. However, most of the observed SNPs in the arRP panel were also seen in this panel. 

Pathogenicity of Mutations

Absence of Mutations in the Control and adRP Panels.

None of the identified mutations were detected in at least 400 ethnically matching control chromosomes, as determined by direct sequence analysis. In addition, the absence of these mutations in any of the adRP patients indicates their specificity to the recessive phenotype and confirms that they are likely to be disease-causing changes. 

Location of Mutations on Protein Domains.

All mutations identified were found to be localized within important domains in the protein SPAM and are thought to be essential for maintaining the protein's function (Fig. 4A). 

Figure 4.

 

(A) Predicted domain structure of SPAM and position of the mutations previously identified by us, ⁶ Collin et al., ⁷ and in the present study are shown in normal, italic, and bold type, respectively; (B) Alignment of the SPAM peptide sequence in human and in other species using CLUSTAL W (ver. 1.82) multiple sequence alignment software. Accession numbers of the protein sequences used for sequence comparison were as follows: chimpanzee, XM_527426.2 (RefSeq); horse, XM_001918159.1 (RefSeq); chicken, XM_426198.2 (RefSeq); dog, XM_848323.1 (RefSeq); and Drosophila, NP_001027571.1 (RefSeq).

Conservation of Mutations.

Multiple protein sequence alignment revealed that five of the identified mutations were highly conserved across other species (Fig. 4B). Two of the identified changes occurred at the highly conserved splice sites for the gene. Nonetheless, the conservation of the remaining mutations could not be studied due to the complete or partial absence of similar sequences in other species. 

Discussion

Here, we report on the molecular screening of EYS in a cohort of patients with autosomal recessive and autosomal dominant RP. Eleven novel mutations, eight missense, one nonsense and two splice site changes, were identified in 10 of 100 unrelated patients with recessive RP (probable mutation frequency of 11%). On the other hand, no disease-causing changes were observed in the 95 tested patients with adRP. In addition, 53 SNPs were identified, of which ∼22% were novel. 

Four of the identified mutations in this study were double heterozygous (Table 1) and hence would be sufficient to explain the recessive phenotype in their corresponding patients. This finding is attributable to the presence of two mutant alleles in each patient, as well as the pathogenicity of each of the identified changes in terms of protein conservation or creation of PTC and the existence of two of these changes p.[C2139Y]+[G2186E] within ESE sites. 

In addition, the homozygous splice site change identified in one patient is sufficient, per se, to disrupt the protein function. It is possible that this mutation could disrupt the splicing process as predicted by computational analysis and hence could either lead to exon skipping or intron retention. However, it was not possible to affirm this effect at the mRNA level due to lack of EYS expression in lymphocytes. ⁶  

On the other hand, the remaining six mutations were heterozygous. For each, the second mutant allele could not be detected by direct sequencing, implying that it could be due to a large deletion. It is also possible that the second mutation in these families was located within the regulatory elements of the gene and hence was undetectable by the experimental methodology we used. Even though rare, a single EYS mutation in combination with another mutation in a second gene could also explain the phenotype in these families. In addition, it is possible that in these families any of the EYS mutations could act as a second-site modifier for a primary mutated gene that remains to be identified. Finally, they may represent extremely rare polymorphisms. 

Nevertheless, these mutations in addition to their absence in any of the tested control chromosomes occurred at positions in the protein where the residue type is highly conserved across other species and/or involved important domains that are essential for maintaining the structure and function of the protein (Figs. 4). For example, the nonconservative p.T2831I substitution is located within one of the LamG domains predicted to be essential in maintaining SPAM function. Similarly, p.R2604C and p.G2685E are both nonconservative substitutions of basic for uncharged residues and nonpolar for acidic residues, respectively, and are highly conserved across other species (Fig. 4B). In addition, the p.G2685E was predicted to exist within an ESE site at the 3′ end of exon 41. 

In contrast, the p.R2402K and p.W3065L amino acid substitutions were chemically similar; nevertheless, their respective position within the EGF and LamG domains of SPAM could be sufficient to alter the protein stability. It is interesting to note that chemically related amino acid changes have been reported as disease causing. ^14,15 Finally, the c.1299+3A>C (Figs. 3U, 3V) occurred within the donor splice site region of exon 8, which is predicted to create a cryptic splice site that subsequently may lead to a frame-shift mutation and premature truncation of the protein. 

Remarkably, there are no mutation hot spots within EYS, since the mutations previously identified ^6,7 as well as those shown here are dispensed throughout the gene (Fig. 4A). In addition, all types of mutations were observed, which is probably consistent with loss of function as a mechanism of the disease. This finding is comparable to the previously identified genes PAX3 and CFTR for commonly inherited recessive conditions. ^16,17 Moreover, most of the mutations reported herein are missense and for these the deleterious effects are usually attributed to their effects on protein function. However, recent reports have suggested that the drastic effect on gene function is associated with the location of these nucleotide substitutions within ESEs which in turn affects the splicing machinery. ^18,19 This mechanism is probably the one by which three of the mutations described herein, p.C2139Y, p.G2186E, and p.G2685E, could lead to the RP phenotype. 

The changes observed in this study occurred in patients of Caucasian, Asian, and Chinese ethnic backgrounds. Therefore, adding the multiple ethnicity and the number of mutations identified in our study to those from previous reports ^6,7 supports our initial assumption that EYS is the first major gene for recessive RP. However, it is important to note that the allele frequency of the mutations presented herein should be regarded cautiously since 55% of these changes occurred in only one allele. Therefore, the presence of a second mutant allele is essential both for explaining the RP phenotype and for supporting the pathogenicity of the first mutant allele. Meanwhile, the absence of any pathogenic changes within the adRP panel does not exclude the role of EYS as a causative gene for this form of RP, since a larger panel of patients is needed to affirm this conclusion. 

This is the first report of molecular genetic screening of EYS in a large cohort of patients with recessive and dominant forms of RP revealing a probable mutation frequency of 11% for arRP. Future studies to determine the functional consequences of the identified sequence changes within EYS is an essential future step in allowing better understanding of the molecular mechanisms underlying the rod photoreceptor cell degeneration caused by defects in SPAM. 

Footnotes

 Supported by British Retinitis Pigmentosa Society Grant GR556; Foresight, Dubai; Fight For Sight, United Kingdom; Foundation Fighting Blindness; National Institutes of Health Research (NIHR) Biomedical Research Centre for Ophthalmology; and The Special Trustees of Moorfields Eye Hospital London and EU-Neurotrain Grant MEST-CT-2005-020235.

Footnotes

 Disclosure: M.M. Abd El-Aziz, None; C.A. O'Driscoll, None; R.S. Kaye, None; I. Barragan, None; M.F. El-Ashry, None; S. Borrego, None; G. Antiñolo, None; C.P. Pang, None; A.R. Webster, None; S.S. Bhattacharya, None

The authors thank the patients who participated in the study.