Screening of a Large Cohort of Leber Congenital Amaurosis and Retinitis Pigmentosa Patients Identifies Novel LCA5 Mutations and New Genotype–Phenotype Correlations

Introduction

Leber congenital amaurosis (LCA) is an infantile onset, severe retinal dystrophy that presents with profound visual impairment, inability to fixate and nystagmus from birth or within the first few weeks of life [von Leber, 1869]. It is the most severe form of a spectrum of disorders arising within the first few years of life, which affect both rod and cone photoreceptors, termed childhood or early-onset retinal dystrophy (EORD) [Gu et al., 1997]. The term juvenile retinitis pigmentosa (RP) has also been used to describe EORD. Autosomal recessive RP (arRP) falls into this spectrum of disorders with onset typically later in childhood or beyond. LCA accounts for 3%–5% of childhood blindness in the developed world and has an incidence of two to three per 100,000 live births [Heckenlively et al., 1988]. Depending on the age at diagnosis, the retinal appearance may be normal or there may be a variety of abnormalities including vascular narrowing, macular atrophy, peripheral white dots at the level of the retinal pigment epithelium (RPE), and retinal pigmentation. The full-field electroretinogram (ERG) is usually nondetectable (<10 μV) with conventional ERG testing or severely decreased before the age of 1 year [Foxman et al., 1985; Franceschetti and Dieterle, 1954].

LCA is predominantly inherited as an autosomal recessive trait, although rare dominant forms have been reported [Swaroop et al., 1999]. Nineteen causative genes have thus far been reported: AIPL1 [MIM #604392] [Sohocki et al., 2000], CABP4 [MIM #608965] [Aldahmesh et al., 2010], CEP290 [MIM #610142] [den Hollander et al., 2006], CRB1 [MIM #604210] [den Hollander et al., 2001], CRX [MIM #602225] [Freund et al., 1998], GUCY2D [MIM #600179] [Perrault et al., 1996], IMPDH1 [MIM #146690] [Bowne et al., 2006], IQCB1 [MIM #609237] [Estrada-Cuzcano et al., 2011], KCNJ13 [MIM #603208] [Sergouniotis et al., 2011], LCA5 [MIM #611408] [den Hollander et al., 2007], LRAT [MIM #604863] [Thompson et al., 2001], MERTK [MIM #604705] [Gal et al., 2000], NMNAT1 [MIM #608700] [Chiang et al., 2012; Falk et al., 2012; Koenekoop et al., 2012; Perrault et al., 2012], RD3 [MIM #180040] [Friedman et al., 2006], RDH12 [MIM #608830] [Janecke et al., 2004], RPGRIP1 [MIM #605446] [Dryja et al., 2001], RPE65 [MIM #180069] [Gu et al., 1997; Marlhens et al., 1997], SPATA7 [MIM #609868] [Wang et al., 2009], and TULP1 [MIM #602280] [Hanein et al., 2004]. Interestingly, mutations in some of these genes, such as CEP290 [Littink et al., 2010], CRB1 [den Hollander et al., 1999], LRAT [Thompson et al., 2001], MERTK [Gal et al., 2000], RPE65 [Morimura et al., 1998], and SPATA7 [Wang et al., 2009], may also be associated with milder forms of retinal dystrophy.

The LCA5 locus (MIM #604537) was mapped to chromosome 6q11-q16 by traditional linkage studies in a multigenerational kindred of the Old Order River Brethren [Dharmaraj et al., 2000]. Further homozygosity mapping in consanguineous and nonconsanguineous families led to a refinement of the critical region [den Hollander et al., 2007; Mohamed et al., 2003] and the causative gene, LCA5, was identified [den Hollander et al., 2007]. LCA5 comprises nine exons, seven of which encode Lebercilin, a 697 amino acid protein that contains four coiled-coil domains [den Hollander et al., 2007]. Lebercilin is widely expressed in human tissues including cilia. In situ hybridization of mouse embryos at day 12.5 postcoitum detected almost ubiquitous, low-level staining. At embryonic day 14.5, staining of the eye, inner ear, kidney, and regions of the central and peripheral neural system, gut and the ciliated epithelium of the nasopharynx, trachea, and lungs was more pronounced. In the adult eye, expression was limited to the photoreceptor cell layer. In ciliated cell lines, lebercilin localizes to the ciliary axoneme. In mouse and rat retina, it is found between the outer and inner segments of the photoreceptor layer. In a human embryonic kidney cell line, HEK293, recombinant lebercilin was found to interact with 24 proteins, many of which are associated with centrosomal or ciliary functions, and the intraflagellar transport machinery [Boldt et al., 2011]. Lca5 inactivation in mice led to very severe and rapid retinal degeneration. At 1 month postnatally, there was a significant reduction in photoreceptor cells, and after 4 months, the photoreceptor layer was completely absent [Boldt et al., 2011].

To date, 12 different mutations in LCA5 have been reported. In patients of all families except one, the phenotype has been typical of LCA [Abu-Safieh et al., 2013; Ahmad et al., 2011; den Hollander et al., 2007; Gerber et al., 2007; Jacobson et al., 2009; Li et al., 2011; Ramprasad et al., 2008; Vallespin et al., 2010a; Vallespin et al., 2010b]. Despite its ubiquitous expression and its crucial functions in the cilia [Boldt et al., 2011], the phenotype associated with LCA5 mutations has thus far been confined to the retina. In this paper, we present the results of screening for mutations in LCA5 in a large number of subjects with LCA and EORD (n = 797), and arRP (n = 211), and describe the associated phenotypes.

Materials and Methods

Mutation Screening

Nucleotide numbering reflects cDNA with +1 corresponding to the A of the ATG translation initiation codon (RefSeq NM_181714.3), according to journal guidelines (www.hgvs.org/mutnomen). The initiation codon is codon 1. Primers used to amplify the coding exons and the intron–exon boundaries of LCA5 were designed using those published previously [den Hollander et al., 2007]. In addition, 194 persons with LCA and 94 persons with arRP were analyzed for the presence of an exon 1 deletion identified in the original LCA5 family [den Hollander et al., 2007]. To this end, primers 5′-TTCCGTGCAATTTAGGGATT-3′ and 5′-CGGGGTTTTTGTTTTGTTTG-3′ were designed to amplify a deletion breakpoint fragment of 117 bp. All standard polymerase chain reactions (PCR) were performed in a total volume of 30 μl containing 200 μM dNTPs (Bioline, London, UK), 20 μM of each primer, 1× reaction buffer with 1.5 mM MgCl₂ (Bioline) with 1 unit of Biotaq (Bioline) and 100 ng of DNA. PCR was carried out on a PTC200 DNA engine thermal cycler (Bio-Rad, Hemel Hempstead, UK). Cycling conditions were as follows: 5 min denaturation at 94°C followed by 35 cycles of 94°C for 30 sec, annealing temperature for 30 sec, and extension at 72°C for 30 sec. A final extension of 72°C for 5 min completed the cycling conditions.

PCR products were visualized on a 2% agarose gel containing 0.05% ethidium bromide. The products were cleaned using multiscreen PCR filter plates (Millipore, Watford, UK) prior to sequencing. PCR products were sequenced directly using the ABI Prism Big Dye terminator Kit (V3.1, Applied Biosystems, Foster City, CA) in a 10-μl reaction. Samples were purified using the Montage cleanup kit (Millipore) prior to being run on an ABI applied biosystems 3730 DNA sequencer (Applied Biosystems). Electropherograms were analyzed for sequence changes using DNAStar computational software (DNAStar, Inc., Madison, WI). Any sequence changes identified were checked visually.

Mutations were tested in normal controls when possible, and cosegregation was performed in the families, when possible. Mutation frequencies were analyzed in the exome variant server database, 1000 genomes database, and dbSNP database. In silico analyses were then performed to assess pathogenicity, including Blosum62, Polyphen2, and PhyloP.

Patient #1 was originally identified to harbor a heterozygous mutation in LCA5 by microarray analysis using the Asper LCA Apex chip. His second mutation was identified following LCA5 sequence analysis. Patient #18 was selected for LCA5 sequence analysis because homozygosity mapping of two affected persons (IV:2 and IV:3) of family W12-2240 (Fig. 1) showed a large, significant homozygous region surrounding LCA5. To identify this region, we had employed the CytoScan® High-Density array containing probes to detect 750,000 single-nucleotide polymorphisms (Affymetrix, Santa Clara, CA).

Results

Clinical Phenotype

Of the 18 probands with LCA5 variants identified in this study, all but two patients had a clinical diagnosis of LCA, with onset of severe visual loss at birth or within the first few months of life, and nystagmus (Table 2). However, patients #7 and #16 received a diagnosis of EORD. Patient #7 was born with congenital nystagmus, and developed night blindness at the age of 3, visual field loss at the age of 6, and visual loss at the age of 8. In patient #16, the visual loss started after the age of 40 years. In addition, patient #16 did not have nystagmus. All of our LCA and EORD patients had nyctalopia and severe visual field constriction. General health was good in the majority of patients, except in patient #2 who had global developmental delay and patient #3 who had behavioral problems. Fourteen of the 16 LCA patients were legally blind with visual acuity ranging from 1.0 logMAR to light perception. One patient, #11, with LCA and nystagmus, had visual acuities of 0.48 right (20/60) and 0.54 left (20/70) at the age of 6 years (Table 2). Patient #16, with EORD, had a visual acuity of 0.30 (20/40) at the age of 55 years. Most patients had a hypermetropic refraction (Table 2). The high myopic correction in patient #6 may be attributed to keratoconus. A full-field ERG was performed in eight patients and was nondetectable (<10 μV) with conventional ERG testing in five patients. A small cone ERG response could be quantified in patient #14 and trace cone ERG responses (unquantified) could be detected in patient #16 (Table 2). Anterior segments were normal in all but one of the oldest patients (patient #6), who had bilateral keratoconus at the age of 37 years. This patient also developed bilateral nuclear sclerotic and posterior subcapsular lenticular opacities in adulthood.

Table 2. Clinical Characteristics of Persons with LCA5 Variants

				Age at		Latest	Latest	RE spherical	LE Spherical
			Age of	exam	General	VAR	VAL	equivalent	Equivalent	Anterior
Patient	Ethnicity	Diagnosis	onset	(years)	health	(LogMAR)	LogMAR)	(dioptres)	(Dioptres)	segment	Fundus	ERG
1	British Caucasian	LCA	3 months	17	Good	1	1	+4.00	+4.38	Normal	Widespread RPE atrophy, white dots, vascular attenuation, minimal BS pigmentation inferiorly	Undetectable
2	Afghani	LCA	Birth	4	Global developmental delay	LP	LP	+2.50	+2.00	Normal	White dots in retinal periphery, normal macula, normal discs	Rudimentary full-field rod function; undetectable cone function
3	Slovakian Romani	LCA	3 months	8	Behavioral issues	LP	LP	+6.50	+6.50	Normal	Widespread RPE atrophy, white dots, macular atrophy	Undetectable
4	British Caucasian	LCA	Infancy	8.5	Good	1.1	1.1	N/A	N/A	Normal	Widespread RPE atrophy, white dots, normal discs, and macula	N/A
5	Indian Mauritian	LCA	6 weeks	1	N/A	LP	LP	+6.00	+6.00	Normal	N/A	N/A
6	British Caucasian	LCA	Birth	37	Good	CF	LP	−13.26	−1.94	Bilateral KC; bilateral NS; PSCLO	Asteroid hyalosis LE, widespread RPE atrophy, white dots, minimal BS pigmentation, vascular thinning, normal macula	N/A
7	Iraqi	EORD	Birth	21	Good	1	1.2	N/A	N/A	Normal	Salt and pepper fundus, peripheral preserved RPE islands pattern, minimal BS pigmentation, pumpkin-colored macula	N/A
8	Spanish	LCA	Birth	3	Good	CF	CF	N/A	N/A	Normal	White dots in retinal periphery	N/A
9	Spanish	LCA	Birth	3	Good	CF	CF	N/A	N/A	Normal	White dots in retinal periphery	N/A
10	Mexican	LCA	Birth	3	Good	HM	HM	N/A	N/A	Normal	N/A	N/A
11	Chinese	LCA	Birth	6	Good	0.48	0.54	+3.25	+3.88	Normal	White dots in posterior pole	Undetectable
12	US	LCA	Birth	20	Good	HM	HM	+5.00	+5.00	Normal	Granularity throughout fundus	Undetectable
13	Chinese	LCA	Birth	16	Good	HM	HM	N/A	N/A	Normal	White dots in retinal periphery	Not able to perform
14a	US	LCA	Birth	19	Good	1	1	+2.25	+2.25	Normal	White dots in retinal periphery and posterior pole; minimal BS pigmentation; right optic disc drusen	Cone b-wave: 0.26 μV OD, 0.56 μV OS
15	US	LCA	Birth	1	Good	N/A	N/A	N/A	N/A	Normal	N/A	N/A
16	Taiwan	EORD	43 years	55	Good	0.3	1.66	−0.75	−0.13	Normal	Relatively pink optic disks, attenuated vessels with vascular sheathing, macular RPE atrophy with preserved foveae, widespread RPE atrophy, with superficial and deep intraretinal pigmentation	Trace of cone function in OS
17	British Caucasian	LCA	Birth	18.5	Good	LP	LP	+9.00	+9.00	Normal	Bilateral well-circumscribed atrophic macular lesions. Relatively normal hypopigmented mid-peripheral and peripheral retinae. No white dots, normal optic nerves	N/A
18	Pakistani	LCA	Birth	20	Good	LP	LP	N/A	N/A	Normal	N/A	Undetectable

• ^a Represents patient ID Berman–Gund Laboratory ID: 048–038. This patient was evaluated by full-field ERG testing using narrow band-passed filtering and computer averaging that can extend the range of detectability of 30 Hz cone ERGs to 0.05 μV [Berson et al., 1993].
• ArRP, autosomal recessive retinitis pigmentosa; BS, bone spicule; CF, counting fingers; ERG, electroretinogram; HM, hand movements; KC, keratoconus; LCA, Leber congenital amaurosis; LP, light perception; N/A, not available; NS, nuclear sclerosis; OD, right eye; OS, left eye; PPRPE, preserved para-arteriolar retinal pigment epithelium; PSCLO, posterior capsular lens opacity; RPE, retinal pigment epithelium; VAL, visual acuity left eye; VAR, visual acuity right eye.

Fundus examination most commonly revealed widespread retinal pigment epithelial atrophy and granularity, with peculiar, small, white dots at the level of the RPE in the retinal periphery, which were seen in 10 of 14 patients (71%) in whom funduscopic data were available (Fig. 2). Figure 2B and C shows the white dots clearly, and document the round, evenly spaced, similar-sized lesions. Intraretinal pigment migration, when present, was minimal, situated in the far retinal periphery, and seen in the oldest patients, patients #1, #6, #7, #14, and #16 (Fig. 2A, D, and E). Macular atrophy was noted in only three patients, #3, #16, and #17 (Fig. 2B, E, and F); the macula was otherwise normal on funduscopy in the remaining patients. In one patient, the macular lesion was particularly severe, with features of a “macular coloboma” (Fig. 2F). Optic discs were normal in appearance (pink disc color) in all but one patient (patient #14), who had optic disc drusen (Fig. 2D). In one family with arRP (patient #7), we identified a pumpkin-colored maculopathy and intraretinal pigmentation in the periphery with faintly preserved islands of RPE (Fig. 3).

OCT data, available for patients #1, #11, and #16, showed preservation of the central foveal inner segment/outer segment (IS/OS) junctions and of the outer segments in the two younger patients (#1 and #11) (Fig. 4A and B). In the older patient (#16), OCT imaging demonstrated macular atrophy, disruption of retinal lamination, and the presence of a hyporeflective well-circumscribed area in the outer nuclear layer, with a hyperreflective border. This may represent an area of outer retinal tubulation (also known as rosettes) (Fig. 4C). OCT image acquisition was otherwise not possible due to nystagmus. Fundus autofluorescence (FAF) imaging data were available in two patients (Fig. 5). FAF in patient #1 revealed an overall hypofluorescence in the macula, with a central hyperfluorescent signal in the fovea (Fig. 5A). FAF images of patient #16 showed severe hypofluorescence in the macula corresponding to significant RPE atrophy (Fig. 5B). Goldmann perimetry in patient #16 revealed a small central island, with a relatively large preserved nasal field and a small temporal field.

Discussion

LCA5 mutations are a rare cause of LCA. Previously, only 12 different mutations had been identified and published in 16 families with LCA. In the current study, which is the largest study to date, we have screened the coding region of LCA5 in 1,008 patients, 797 with LCA and EORD, and 211 with arRP, ascertained from eight different countries. We identified mutations in 18 new LCA families. We discovered 19 different LCA5 mutations, 17 of which are novel. Mutations in LCA5 thus account for ~2% of patients with LCA, EORD, and arRP in our cohort. All 29 different LCA5 variants have now been included in the new Leiden Open Variation Database (LOVD; http://www.LOVD.nl/LCA5). In two patients (#14 and #15), we identified heterozygous variants. The second allele may have been missed because the variant may be located outside the protein-coding exons, or because these variants are causative in a more complex inheritance model (such as digenic), or that these are coincidental findings of heterozygous variants that play no role in the LCA phenotype.

The majority of LCA5 mutations found both in this study and in those published previously are likely null mutations due to premature termination of the protein. Only one missense mutation in this gene (p.Ser202Pro) has been previously reported in LCA, but without accompanying clinical information [Vallespin et al., 2010b]. We also identified a missense variant, p.Met1Ile (the start codon), in patient #4, who carried a diagnosis of LCA, in combination with the most common LCA5 stop mutation p.Gln279*. If the next methionine in the open reading frame were to be used instead (at position 336), as the alternative start codon, then two coiled–coiled domains in the lebercilin protein would be lacking, very likely rendering the truncated lebercilin inactive. Therefore, this missense mutation may also be a null allele. We also identified another missense mutation, p.His164Arg. This mutation was found in combination with a novel nonsense mutation, p.Tyr23*, in LCA patient #16, but segregation data are lacking. Thus, we cannot exclude that these variants are in cis configuration and that they are not causative for the phenotype in patient #16.

Previously, and in the majority of patients in this study, the phenotype reported in association with LCA5 mutations is consistent with LCA with severely reduced vision at, or near birth, nystagmus, and a nondetectable (<10 μV) ERG [Ahmad et al., 2011; den Hollander et al., 2007; Dharmaraj et al., 2000; Gerber et al., 2007; Jacobson et al., 2009; Mohamed et al., 2003; Ramprasad et al., 2008]. High hypermetropia is common. The visual acuity is reported to range between 0.20 to light perception and there is extensive peripheral field loss. In our study, the majority of patients had a clinical diagnosis of LCA, with fundus examination revealing widespread atrophy of the retina and RPE but with little intraretinal pigment migration. On the clinical examination, we commonly identified scattered white dots at the level of the RPE (Fig. 2A–D—white arrows). White dots can be found in other genetic types of LCA, such as RPE65, LRAT, and CEP290. In RPE65-type LCA, Weleber et al. (2011) suggested that the white dots might represent accumulations of trans-retinyl esters. The biochemical make-up of the white dots associated with LCA5 mutations is as yet unknown. Our LCA5-associated retinal phenotypes are consistent with the retinal changes observed in other patients with LCA5 mutations (Figs. 2 and 3) [Ahmad et al., 2011; Dharmaraj et al., 2000; Gerber et al., 2007; Mohamed et al., 2003; Ramprasad et al., 2008].

However, in this study, for the first time, we identified an Iraqi family with two affected sibs who carried a diagnosis of EORD, because they maintained some useful vision until the age of 10 years. In addition, these siblings had preservation of the peripheral islands of RPE (Fig. 3), a phenotype similar to preserved para-arteriolar retinal pigment epithelium documented by Heckenlively (1982) and later found to be associated with LCA and arRP patients with CRB1 mutations [den Hollander et al., 1999; Heckenlively, 1982]. We have thus expanded the phenotypic spectrum associated with LCA5 mutations, which now includes both EORD and a new retinal phenotype consisting of preserved islands of peripheral RPE (Fig. 3).

There is little published data on retinal imaging such as OCT and FAF imaging in patients with LCA5 mutations. In this regard, we here contribute three new patients with LCA5 mutations and new OCT and FAF data. However, it is a limitation of the present study that we were unable to obtain good quality autofluorescence imaging and OCT on the entire cohort. This was, in part, due to the inherent difficulties of such imaging in patients with nystagmus, their ages and their remote living conditions. Jacobson et al. (2009) reported a low autofluorescent signal corresponding to macular atrophy in two siblings with a homozygous LCA5 mutation, in whom mutations in CRB1 had previously been excluded. We demonstrated similar macular hypoautofluorescence in one subject who in addition had a small area of hyperautofluorescence at the fovea, suggesting continued metabolic activity of the foveal cones and RPE. In our study, OCT imaging was possible in three subjects, and the findings revealed surprising preservation of the central foveal IS/OS junctions and of the outer segment structures (in the younger patients, of the age of 17 years and 6 years, respectively). In the older patient, of the age of 55 years, OCT imaging demonstrated macular atrophy, loss of the retinal lamination, and probable outer retinal tubulation (also known as rosettes), a late-stage process in the degeneration of photoreceptors that has been described recently in a number of retinal conditions (Fig. 4C) [Sergouniotis et al., 2012; Zweifel et al., 2009]. Previously published OCT findings have demonstrated loss of the photoreceptor and outer nuclear layers, and abnormal retinal lamination eccentric to the fovea, but with evidence of a retained photoreceptor layer at the fovea [Jacobson et al., 2009]. This suggests that in the early stages of the disease, there may be viable photoreceptors (cones) in the foveal region, which may be amenable to therapeutic intervention. It is of interest that patient #14 at the of age of 19, with an advanced stage of disease, still had a detectable 30 Hz cone ERG in each eye (Table 2) with narrow band-passed filtering and computer averaging. In future clinical trials, this technique of ERG recording, which is not performed in the present ISCEV protocol, could be considered as a method of objectively monitoring efficacy or lack of efficacy as undoubtedly most patients with remaining vision will have a detectable cone ERG with this technique.

Lebercilin is expressed in a number of tissues; however, there have been no reports of extraocular abnormalities in LCA5 mutations. One patient with LCA5 retinopathy died from asphyxia and the authors propose a possible link between the cause of death and the presence of lebercilin in the bronchial ciliated epithelial cells [Ramprasad et al., 2008]. In the present study, one patient had global developmental delay, whereas another experienced behavioral problems. Whether this is related to abnormal extraocular expression of lebercilin remains to be determined.

LCA5-associated LCA is a rare form of severe childhood onset retinal dystrophy. Most LCA5 mutations are null and are associated with a severe phenotype. Missense mutations are rare and may be associated with a milder EORD, suggesting that such mutations may result in some residual protein function. LCA5-type LCA may be amenable to treatment, but this is more likely to be successful in young patients.

URLs Used in This Study

Exome Variant Server database: http://evs.gs.washington.edu/EVS/

1000 genomes database: http://www.1000genomes.org/

PhyloP: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2808870/

Polyphen2: http://genetics.bwh.harvard.edu/pph2/

SpliceSiteFinder: www.genet.sickkids.on.ca/~ali/splicesitefinder.html

MaxEntSplice: http://genes.mit.edu/burgelab/maxent/Xmaxentscan_scoreseq.html

NNSplice: http://www.fruitfly.org/seq_tools/splice.html