The Phenotype of Leber Congenital Amaurosis in Patients With AIPL1 Mutations FREE

Leber congenital amaurosis (LCA) was first described by Theodore Leberin 1869¹ as a congenital form of retinitispigmentosa. It represents a clinically and genetically heterogeneous disorderwith severe visual impairment from birth.^{2 - 3} Fundusexamination results are not frequently initially normal, but chorioretinalatrophy, narrowing of the retinal vasculature, intraretinal pigment migration,white fundus flecks, and macular aplasia have been described.^{4 - 8} Theretinal basis of the visual loss is shown by absent or severely diminishedrod and cone responses on electroretinography (ERG).⁹ Nystagmus,enophthalmos, sluggish pupillary responses, keratoconus, cataracts, and hyperopiahave also been described.^{10 - 12}

Leber congenital amaurosis is usually inherited as an autosomal recessivetrait, although dominant inheritance has been reported.^{13 - 16} Currently,mutations in 6 different retinal genes have been shown to cause LCA. The genesinclude (1) retinal guanylate cyclase (GUCY2D),¹⁷ (2) retinal pigment epithelium–specific 65kDprotein (RPE65),¹⁸ (3)cone-rod homeobox (CRX),^{19 - 22} (4)crumbs gene homolog of CRB1,^{23 - 24} (5)retinitis pigmentosa GTPase regulator–interacting protein (RPGRIP-1),^{25 - 26} and (6) AIPL1, encoding the aryl hydrocarbon receptor interactingprotein-like 1 protein.^{27 - 28}

The AIPL1 gene consists of 6 exons and encodesa protein of 384 amino acids. This sequence includes 3 tetratricopeptide repeatmotifs thought to be associated with protein-protein interaction, and itssimilarity with aryl hydrocarbon interacting protein is suggestive of a proteinfolding function.^{27 - 28} The exactfunctions of the AIPL1 gene are not fully understood.However, recent data suggest that the protein may be involved in photoreceptordifferentiation during development and subsequent survival of photoreceptors.²⁹ Indeed, through interaction with the NUB1 protein,it might be involved in regulation of the cell-cycle progression during photoreceptormaturation.²⁹ Mutations in AIPL1 account for 7% of LCA.²⁸

Clinical outcomes differed for patients with LCA and GUCY2D mutations when compared with those with RPE65 defects^{30 - 33} interms of the natural history of this disorder. In addition, some heterozygouscarriers of GUCY2D mutations, who have offspringwith LCA, have been shown to have significant cone abnormalities on ERG results,with essentially normal rod ERG findings.³⁴ Mostheterozygotes with RPE65 mutations have normal ERGfindings.³²

The purpose of this large study is to describe the phenotype of LCAin patients with AIPL1 mutations and compare it withthe known phenotypes of patients with mutations in other LCA genes. The phenotypeof 26 patients with LCA of different ethnic origins with mutations in AIPL1 is described. The genotype of most patients has previouslybeen published.^{15 ,28} The ERG andclinical findings in a female heterozygous carrier are also reported.

Informed consent was obtained from all patients involved in this studyor from their legal guardians in accordance with the Declaration of Helsinki.The review and ethics boards of the institutions approved this study.

The clinical diagnosis of LCA was made on the basis of the followingdiagnostic criteria: severe visual impairment from birth or during early infancyaccompanied by nystagmus, absent or very sluggish pupillary responses, andabsent or markedly reduced rod and cone ERGs. All ERGs were performed accordingto the International Society for Clinical Electrophysiology of Vision standards.³⁵ The examinations were undertaken in 5 centers andincluded slitlamp biomicroscopy, retinoscopy, and indirect ophthalmoscopyfollowing pupillary dilation (Table 1).Clinical pictures were taken, and keratometry was performed.

DNA was extracted from peripheral blood leukocytes or cheek swabs. Acohort of 303 patients with LCA was screened for mutations in AIPL1. Patients were from a wide range of racial and ethnic backgrounds.The 6 exons of AIPL1 were screened using single-strandconformation polymorphism analysis (SSCP) followed by direct sequencing whenan aberrant migration pattern was noted on the SSCP gels. In 39 probands,direct sequencing was used to screen for mutations in AIPL1, while in the others, SSCP was initially undertaken using primersand conditions previously described.²⁷ Thegenotype of most of the patients with AIPL1-relatedLCA in this study has been published previously (Figure 1).^{27 - 28}

Mutations in AIPL1 were detected in 26 probandswith LCA (Figure 1). Seventeen probandswere homozygotes, while 9 were compound heterozygotes. Twenty-four of the52 mutated AIPL1 alleles carried the W278X mutation.All sequence changes identified in our patients were absent in 205 controlsamples.

Night blindness was reported in 13 probands and photoaversion in 4.Photoattraction (staring at lights) was noted in 2 probands (Table 1).

Visual acuities were found to vary between probands and ranged from20/400 to light perception. Nine patients had light perception. Seven patientshad hand motion vision (Table 1).Cycloplegic refractions performed in 10 patients showed hyperopia in 8 (+3.00diopters [D] to +7.00 D) and myopia in 2 (–0.50 D to –2.75 D).

Twenty-four probands with an AIPL1-relatedLCA genotype had some form of pigmentary retinopathy that ranged from mildmidperipheral salt and pepper-like retinopathy to diffuse and severe chorioretinopathy(Figure 2, Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7).The youngest patient with pigmentary changes was 4 months old. Two patients,a 2-year-old and a 3-year-old, had essentially normal retinas with indistinctfoveal reflexes. A maculopathy of variable appearance was noted in a significantnumber of patients (Figure 2, Figure 3, Figure 4, Figure 5, and Figure 7). Information about the macularappearance was available in 20 of the 26 probands. Maculopathy was noted in16 (80%) of 20 probands. In 4 probands, all young children (ranging from ages2-6 years), an abnormal indistinct foveal reflex was noted, which likely representsan early stage of maculopathy. This strongly suggests that a significant numberof patients with LCA and AIPL1 mutations developa maculopathy. The maculopathy ranged in appearance from mild foveal atrophywith variable degrees of macular stippling to aplasia. The youngest patientwith macular atrophy was 8 years old (Table1).

Information about the presence of keratoconus was available in 19 probands(Table 1). Keratoconus was diagnosedin 6 probands (32%), and cataracts were noted in association with the keratoconusin 5 of these 6 patients. Distinct hydrops with scarring and breaks in theDescemet membrane were noted in proband 17. The cataracts ranged from corticalchanges to posterior subcapsular cataracts. Of interest, keratoconus and cataractswere only seen in patients who were homozygous for AIPL1 mutations. Keratoconus was not observed in patients with compoundheterozygous mutations. The youngest patient with keratoconus and cataractwas aged 10 years.

Varying degrees of optic nerve pallor were noted in all patients afterthe age of 6 years. The optic nerve head appeared normal in children youngerthan 6 years, except in an infant (Table1).

The ERG findings obtained in the 3 sets of clinically normal parentsof probands 7, 10, and 26 who carry the AIPL1 mutationin a heterozygous state did not show any abnormalities. However, the ERG of1 carrier parent of proband 2 with the W88X mutation showed significant rodabnormalities (Figure 8). She didnot have any ocular complaints, and her clinical examination findings werenormal. This 47-year-old mother had vision of 20/20 OU. Although her retinalexamination results were unremarkable, full-field flash ERG showed rod b-waveamplitudes to be reduced to approximately one third of normal, with no changein implicit time. This is well below the lower limit of normal. The 30-Hzflicker and single flash cone responses were within normal limits (Figure 8). The ERG responses were reproducibleon repetition.

The LCA phenotypes with mutations in the other LCA genes (GUCY2D, RPE65, CRX, CRB1, and RPGRIP1) were comparedwith the LCA phenotypes of the current study and tabulated in Table 2, Table 3, Table 4, Table 5, and Table 6.The AIPL1-related LCA phenotype is severe in nature,with pronounced macular involvement in individuals older than 6 years withvarying degrees of optic nerve pallor. Additional findings of keratoconusand cataract could be present.

Both GUCY2D-related and AIPL1-related LCA phenotypes have markedly decreased visual acuities,visual fields, and ERGs.^{30 - 31 ,33} However,maculopathy, remarkable peripheral pigmentary changes, cataract, and significantoptic disc pallor were not detected in patients with mutations in GUCY2D.^{30 - 31 ,33} Keratoconuswas reported by El-Shanti et al³⁶ in a Jordanianpedigree. Compared with the reported GUCY2D phenotype,^{30 - 31 ,33} theAIPL1 phenotype appears to be similar in severity of visual loss. Phenotypicaldifferences exist in the pattern of pigmentary changes, cataract, and keratoconus,which are more frequent in AIPL1-related LCA (Table 2).

The RPE65 phenotype reported in earlier studies^{30 - 32 ,37 - 38} showsthat the visual acuities, visual fields, and ERG measurements were betterthan in the AIPL1 phenotype. Patients with RPE65-relatedLCA may develop a mild maculopathy, and the documented peripheral retinalchanges are characterized as grainy and/or salt and pepper-like. The maculopathyof patients with AIPL1-related LCA appears to bemore pronounced in all probands older than 6 years, while the peripheral retinalchanges range from mottling to bone spicule-like formation. Cataract and keratocunuswere present in one third of the patients with AIPL1-relatedLCA. Lorenz et al³² conclude that patientswith LCA and RPE65 mutations are distinguishableon clinical grounds, based on their measurable visual acuities, their transientvisual improvement in childhood followed by deterioration in later life, measurablecone ERGs (which also diminish in later life), measurable visual fields, andsignificant night blindness. The data from our study suggest that patientswith LCA and mutations in AIPL1 do not have a similarcourse (Table 3).

From the several reported cases of patients with LCA and CRX mutations, visual acuities of 20/300 to light perception, with1 case of 20/80, were described.^{15 - 16 ,19 - 22 ,31 ,33 ,39 - 43} Markedatrophy in the macula was recorded in 71% of CRX-relatedLCA, while in AIPL1-related LCA, maculopathy waspronounced in 80% of the patients after 6 years of age. Marked pigmentaryretinopathy was noticed in 84% of patients with AIPL1-relatedLCA unlike in CRX-related LCA, where it was observedin 33% (Table 4).

Compared with that of patients with CRB1 mutations,the phenotype of our patients with AIPL1 mutationsappears to be less variable and more severe. Small white dots and zonal retinal/choroidalhypoplasia were seen in the patients with CRB1-relatedLCA²³ but not in patients with AIPL1-related LCA (Table 5).The presence of cataract, keratoconus, and optic disc pallor were not reportedin the CRB1-related LCA phenotype. The constant featuresreported in the CRB1-related phenotype were moderateto high hyperopia, the relatively early appearance of white spots, and nummularpigment clumps in the retina.²³

The RPGRIP1-related LCA phenotype has beenreported in 3 patients.²⁵ Visual acuity waslight perception. Hyperopia and absence of intraretinal pigment migrationwere noted in 2 patients. However, bone spicule-like pigmentary deposits inthe midperipheral zone were noted in a third patient. No evidence of maculopathyas seen in the patients with AIPL1-related LCA wasobserved (Table 6).

The retinal phenotype of AIPL1-related LCAis that of a severe, congenital retinal dystrophy with a notable maculopathy.The retinal appearances in our patients ranged from near normal (in a 3-year-oldand a 6-year-old) to severely atrophic (and in all patients older than 6 years)with marked maculopathy and pigmentary retinopathy. Varying degrees of intraretinalpigment migration culminating in bone spicule-like pigment and gross pigmentclumps in the retinas were observed. Overall, a high prevalence of macularlesions was observed in our patients compared with patients with LCA causedby mutations in the other 5 genes implicated in this disease. Atrophic macularlesions were particularly frequent and were observed in 16 (80%) of 20 patients;11 harbored a premature stop-codon mutation, either in a homozygous or a heterozygousstate. Macular involvement as seen on ophthalmoscopy likely begins with anindistinct dull or irregular foveal reflex and progresses to a diffuse ill-definedarea of retinal pigment stippling and atrophy, leading to a marked atrophicmaculopathy. Owing to the differences in age at the time of first examination,it was not possible to determine the accurate age of onset of the maculopathy.

The heterozygous carrier parent of the W88X mutation was found to havea significant and reproducible rod ERG abnormality with essentially normalcone ERG results. These ERG findings are significantly different from theheterozygous carriers of GUCY2D mutations, who havesignificant cone ERG abnormalities but relatively normal rod ERG findings.³⁴ The rod ERG abnormalities in the AIPL1 carrier correlate with recent reports showing AIPL1 expression exclusively in rod photoreceptors in the differentiatedretina.⁴⁴ However, more ERGs in carriers of AIPL1 mutations need to be studied to better understandthe role of AIPL1 in relation to rod function.

The presence of keratoconus in patients with LCA has been well documented.^{45 - 48} Thehigh incidence of keratoconus in patients with a homozygous sequence changeof AIPL1 in our cohort may well be significant. Keratoconuswas observed in 6 probands, all with homozygous mutations. There is no definitiveconsensus about the origin of keratoconus in patients with LCA. The incidenceof keratoconus has been reported to be as high as 54.5 cases per 100.0 inthe general population, and it has been noted in 29% of children with LCAand 2% of all children with blindness.^{10 ,49} Keratoconusin patients with LCA occurred in 2% of 0- to 14-year-olds and in 30% of 15-to 45-year-olds, further illustrating the later onset of the pathologic cornealfeatures in comparison with the retinal dysplasia.⁵⁰ Theabsence of keratoconus prior to 9 years of age also has been well documented⁴⁶ and is the case in our cohort too.

Cataract has been associated with many different types of retinal dystrophy.Its association with retinitis pigmentosa has been well documented.^{51 - 52} Cataract has been noted at or beyondthe second decade of life in patients with LCA.⁴⁶ Inthis study, cataracts were observed in 5 probands (27%). Progressive retinaldegenerative changes in association with keratoconus and cataract have beenreported during the course of the disorder.^{46 - 47} Theincidence of both keratoconus and cataract increased with increasing age inour cohort.

The LCA phenotypes are highly variable^{15 ,23 ,31 - 32} andchange with age,⁴⁶ and the phenotypes associatedwith the currently known LCA genes overlap.^{31 - 33} Comparisonsbetween the reported LCA phenotypes of different studies^{23 ,25 ,30 - 33} arehampered by a lack of uniform assessment strategies, age matching, and uniformfollow-up. Despite these obvious difficulties, it is important to study theseLCA phenotypes in an effort to understand the evolution of disease based ongenotype.

In summary, patients with AIPL1-related LCAappear to have a particularly severe phenotype, characterized by marked visualimpairment, nondetectable fields and ERGs, optic disc pallor, maculopathy,peripheral retinal bone spicule-like pigmentation, and a significant prevalenceof keratoconus and cataract.

Mutations in AIPL1 disrupt the normal functionof photoreceptors. AIPL1 is not only expressed inmature rod photoreceptors⁴⁴ but also duringdevelopment in both rods and cones.²⁹ The dysfunctionalrole of AIPL1 in photoreceptor cell cycle progressionleads to photoreceptor cell death during development by disrupting the normalregulation of the cell cycle.²⁹ More detailedunderstanding of the pathogenesis of each molecular subtype of LCA will providefurther insight into treatment.

Correspondence: Sharola Dharmaraj, MD, FRCS, The Johns Hopkins Centerfor Hereditary Eye Diseases, Maumenee 517, Wilmer Eye Institute, Johns HopkinsMedical Institutions, 600 N Wolfe St, Baltimore, MD 21287-9237 (sdharmaraj@jhmi.edu).

Submitted for publication October 10, 2002; final revision receivedApril 3, 2003; accepted June 6, 2003.

This research was supported by grants from the Foundation for RetinalResearch, Highland Park, Ill; the Edel & Krieble Funds of the Johns HopkinsCenter for Hereditary Eye Diseases, Baltimore, Md; the Grousbeck Family Foundation,Stanford, Calif; the Fonds voor Research in Oftalmologie/Fonds de la Rechercheen Ophtalmologie, Edegem, Antwerp, Belgium (Dr Leroy); the Bijzonder Onderzoeksfondsof Ghent University, Ghent, Belgium (Dr Leroy); the Foundation Fighting Blindness-Canada,Toronto, Ontario (Dr Koenenkoop); the Canadian Institutes of Health Research,Ottawa, Ontario (Dr Koenenkoop); Fonds de la recherche en Santé duQuébéc, Montréal, Québéc (Dr Koenenkoop);the Kirchgessner Foundation, Los Angeles, Calif (Dr Sohocki); the KnightsTemplar Eye Foundation, Chicago, Ill (Dr Sohocki); the Foundation FightingBlindness, Owings Mills, Md (Dr Sohocki); Fight for Sight, New York (Dr Sohocki);the Research Division of Prevent Blindness America, Schaumburg, Ill (Dr Sohocki);and William R. Acquavella (Dr Sohocki).

Dr Sohocki is the William R. Acquavella Scholar of Ophthalmic Research,Columbia University, New York.

These authors contributed equally to the study: Sharola Dharmaraj, MD,FRCS, and Bart P. Leroy, MD.

We thank the families for support and cooperation. We also thank thephotography departments of all the institutes for their professional assistanceand to Olof Sundin, PhD, for reviewing the manuscript.