Mutations in the CRB1 Gene Cause Leber Congenital Amaurosis

Objectives  To test the hypothesis that mutations in the CRB1 gene cause Leber congenital amaurosis (LCA) and, if so, to describe the ocular phenotype of patients with LCA who harbor CRB1 sequence variations.

Patients  One hundred ninety probands with a clinical diagnosis of LCA were selected from a cohort of 233 probands ascertained in 5 different countries. The remaining 43 probands (18%) were excluded because they harbored sequence variations in previously identified LCA genes.

Methods  One hundred ninety unrelated individuals with LCA were screened for coding sequence mutations in the CRB1 gene with single-strand conformation polymorphism analysis followed by automated DNA sequencing.

Results  Twenty-one of the 190 probands (9% of the total cohort of 233) and 2(1.4%) of 140 controls harbored amino acid–altering sequence variations in the CRB1 gene (P = .003).

Conclusions  In our cohort of patients with LCA, coding sequence variations were observed in the CRB1 gene more frequently than in any of the other 5 known LCA-associated genes. Likely disease-causing sequence variations have now been identified in 64 (28%) of 233 subjects in this cohort.

Clinical Relevance  Molecular diagnosis can confirm and clarify the diagnosis in an increasing fraction of patients with LCA. As genotype data accumulate, clinical phenotypes associated with specific mutations may be established. This will facilitate the counseling of patients regarding their visual prognosis and the likelihood of associated systemic anomalies.

LEBER CONGENITAL amaurosis (LCA) is a term used to refer to a group of inherited retinal disorders characterized by severe, bilateral visual impairment in infancy. The pupillary responses are sluggish, nystagmus is frequently present, and the electroretinographic responses are markedly attenuated. The eyes of affected children appear grossly normal, with clear media, pink optic discs, and completely attached retinas. The fundus can initially appear normal, although many patients exhibit a degree of vascular attenuation when first examined. Pigmentary abnormalities, ranging from white dots to nummular dark pigment clumps and even bone spicule–like changes, can be seen in some patients. Coloboma-like lesions in the macula are a less common finding. High refractive errors are sometimes present, and older patients may develop keratoconus, presumably from the chronic trauma of the oculodigital reflex. Systemic disorders, most often neurological, are observed in a small number of individuals. The visual outcome can vary widely. Some children with LCA maintain measurable acuity for decades, while others are completely and permanently blind in infancy.^1-3

These disparate clinical findings are in part owing to the genetic heterogeneity of this disorder. To date, mutations in 5 genes have been reported to cause a subset of LCA. These include GUCY2D, encoding retinal guanylate cyclase⁴; RPE65, encoding a retinal pigment epithelium–specific 65 KD^5,6; CRX, encoding the cone-rod homeobox-containing gene⁷; TULP1, encoding the Tubby-like protein 1⁸;and AIPL1, encoding aryl-hydrocarbon interacting protein–like 1.⁹ The discovery of these genes has stimulated the search for novel therapies for this currently untreatable disease.¹⁰

The fraction of LCA resulting from these 5 genes varies widely. TULP1 has been associated with the LCA phenotype in only a single family from the Dominican Republic.⁸ The contribution of the other genes to the worldwide prevalence of LCA has been reported to be CRX, 2.8%; GUCY2D, 6.3%; RPE65, 6.8%²;and AILP1, 7%.¹¹ Thus, the molecular cause for most cases of LCA is still unknown.

Retinitis pigmentosa (RP) is a term used to refer to another clinically and genetically heterogeneous group of retinal degenerations that are closely related to LCA. In fact, the distinction between LCA and RP is largely based on age of onset of the visual dysfunction. Patients whose conditions are diagnosed when they are younger than 1 year are likely to be classified as having LCA, while those older than 1 year who develop photoreceptor degeneration are more likely to be diagnosed as having RP. The similarity of these conditions suggests that genes known to cause RP might also cause some cases of LCA. Indeed, 3 of the known LCA-associated genes—RPE65, CRX, and TULP1—are each known to cause some cases of RP.^11-14

CRB1 is a recently discovered gene that is responsible for a distinctive form of autosomal recessive RP referred to as RP12.¹⁵ Clinically, RP12 exhibits the unusual feature of preservation of the periarteriolar retinal pigment epithelium. Some individuals affected with RP12 first experience visual loss in childhood and some have hyperopic refractive errors.¹⁶ The purpose of this study was to determine whether mutations in CRB1 might cause LCA in some individuals.

Informed consent was obtained from all study patients or their legal guardians. Two hundred thirty-three probands with LCA were ascertained from the United States (146), Canada (43), India (41), Israel (3), and Switzerland(1). Forty-three of these patients were known from a previous study² to harbor a likely disease-causing sequence variation in a previously described LCA gene. These patients were excluded from the present study because the likelihood of finding additional disease-associated mutations was judged to be too low to warrant further consumption of their often irreplaceable DNA samples. The DNA was extracted from peripheral blood using a previously described protocol.¹⁷ One hundred ninety probands were screened for mutations in the coding sequence of the CRB1 gene with single-strand conformation polymorphism analysis. Ninety-four control subjects from Iowa and 46 control subjects from India were screened in an identical fashion. The primer sequences used for the single-strand conformation polymorphism screening have been previously described¹⁵ with the exception that, for exon 2, we used the following primer sequences: forward, GCAGCACAAAGGTCACAAG and reverse, TCCTGATGGCAAATACCTCC. The polymerase chain reaction (PCR) amplification products were denatured for 3 minutes at 94°C and then electrophoresed on 6% polyacrylamide, 5% glycerol gels at 25 W for approximately 3 hours. The gels were then stained with silver nitrate.^18,19 The PCR products from samples with aberrant electrophoretic patterns were then sequenced bidirectionally with fluorescent dideoxynucleotides on an automated DNA sequencer (ABI model 377; PE Applied Biosystems, Foster City, Calif). Clinical records of 19 probands and 1 affected sibling with amino acid–changing sequence variations in the CRB1 gene were available for review, and the resulting phenotypic information was tabulated.

Thirty instances of 20 different amino acid–altering sequence variations were observed in this study (Table 1). Twenty-eight of these changes were found among 21 LCA probands while, only 2 were observed in the heterozygous state in 2 control individuals (P = .003). Six LCA probands were each found to harbor 2 amino acid–altering changes (presumably on different alleles). One proband was homozygous (for Cys948Tyr) and another was homozygous for Cys480Gly. The other 4 probands were compound heterozygotes. Fifteen of the probands exhibited 1 heterozygous amino acid–altering change (Table 1).

A total of 5 synonymous codon changes were observed—each unique, 4 present in single LCA probands and 1 in an Indian control. These were significantly more likely to be found in an LCA proband with a single amino acid–changing CRB1 variation (2/15) than in LCA probands without an amino acid–altering CRB1 change (2/169; P = .03) or in controls (1/140; P = .03).

Four different intronic single nucleotide polymorphisms (SNPs) were found among the study participants (Table 1), but these were not significantly skewed toward LCA probands or controls (either singly or as a group). The most common intronic SNP was located 12 bases upstream from the start of exon 2 and was identically distributed(thymine, 40%; adenine, 60%) among the alleles of LCA probands and control individuals. A single instance of a 4–base pair deletion in intron 2 was observed in a single LCA proband.

The most common amino acid–changing variant that we observed in the LCA cohort (Cys948Tyr) has been previously observed in patients with RP12.¹⁵ The other 17 amino acid–changing variants have not been previously reported.

Clinical records were available for 18 of the LCA probands with amino acid–altering sequence variations and 1 affected sibling (Table 2). Nystagmus was noted in 18 of 19 patients, with "roving eye movements" recorded for the remaining patient. The visual acuity ranged from 20/40 in 1 eye of 1 patient, to light perception, with a median of 20/250. A refraction was recorded in 37 eyes of 19 patients, and the average spherical equivalent was +4.9 diopters (D) (range, −10.00 to +9.00 D). Electroretinogram(ERG) data were available for 12 patients. In all cases the full-field ERG was markedly reduced, and in 7 of 12 patients it was nondetectable. Detailed retinal notes were available for 14 patients. Nummular pigment clumps were seen in 9 of 14 patients (Figure 1), and white spots (Figure 1, Figure 2, and Figure 3) were specifically mentioned in the notes of 5 of 14 patients. Large zones of retinal pigment epithelium atrophy were present in 1 patient(Figure 4), and coloboma-like lesions of the macula were present in 3 others (Figure 5 and Figure 6). Keratoconus was present in 2 siblings, and periarteriolar preservation of the retinal pigment epithelium was seen in a different pair of siblings.

Leber congenital amaurosis is a severe, genetically heterogeneous, autosomal recessive retinal dystrophy. Including CRB1, mutations in 6 genes have now been shown to cause the LCA phenotype. In our complete cohort of 233 LCA probands, amino acid–changing CRB1 variants were found in 21 (9%), making CRB1 the most commonly mutated gene in this group of patients. However, most patients with LCA in our cohort do not harbor sequence variations in any of these 6 genes. This is in part because other LCA genes undoubtedly exist, and in part because some mutations will be missed by PCR-based assays of the coding sequences such as the ones employed in this study. Such assays will not detect deletions involving the primer sites or mutations in the promoters. Indeed, even among the 21 patients with amino acid–changing CRB1 variants in this study, we were unable to detect any CRB1 variations in 15 (36%) of the alleles. Further extensive investigation of these 15 alleles is warranted to try to identify additional regions of the gene that will be fruitful to reexamine in the entire cohort.

The distribution of the synonymous codon variants observed in this study is interesting. Although one would not usually expect such changes to contribute to an altered phenotype (because the predicted protein sequence is not changed), the significant distribution among patients with LCA vs controls (as well as among patients with LCA with CRB1 mutations vs those without mutations) suggests that these alleles may also be disease causing, perhaps by altering the stability or processing of the CRB1 messenger RNA. Functional assays will be necessary to support this hypothesis. Evaluation of a common SNP found within intron 1 does not show a disease association, which indicates that our mutation screening assay is unlikely to be missing a common CRB1 mutation that would account for a large portion of the as yet undetected mutant alleles.

The biochemical function of CRB1 is currently unknown. Comparison of CRB1 with the homologous Drosophila melanogaster crumbs protein (CRB) suggests that it may be involved in neuronal development of the retina.¹⁵ This hypothesis would be consistent with the early onset of retinal dysfunction seen in the patients in this study. Although the number of patients with LCA in whom we found CRB1 coding sequence variations is small, we observed 2 fairly constant phenotypic features: the presence of moderate to high hyperopia and the relatively early appearance of white spots and nummular pigment clumps. The preservation of the periarteriolar retinal pigment epithelium phenotype (characteristic of RP12¹⁵)was seen in 2 of our patients. This is a good example of an increasingly common situation in which an established, clinically based nomenclature system fails to correspond neatly with molecular reality at the DNA level. There are now numerous examples in the field of retinal degeneration research in which a gene that is initially shown to cause disease in patients with one clinical phenotype is later found to cause disease in patients with another—sometimes quite different—phenotype. These include (1) the RDS gene, which was initially found to cause RP but was later discovered to be associated with pattern dystrophy and other maculopathies^20-22;(2) rhodopsin and PDEB, which were initially associated with RP but later found to be associated with stationary night blindness^23,24; (3) CRX and GUCY2D, which each cause LCA and cone-rod dystrophy^4,7,25,26; and(4) RPE65, which causes LCA and RP.^5,12

This lack of perfect correspondence between clinical and molecular nomenclature systems is not a serious obstacle to accurate communication as long as one makes it clear which system is being used in a specific situation. Both systems serve a useful purpose. Infants with severely impaired vision, nystagmus, no striking ophthalmoscopic abnormalities, hyperopia, and a nonrecordable ERG will continue to be given the clinical diagnosis of LCA, and it remains an important goal to ultimately identify all the genes that are capable of causing this phenotype. In contrast, scientists studying the biological roles of individual genes like CRB1 will continue to be interested in identifying all the different phenotypes that can be caused by variations in that particular gene. In the present study, the frame of reference was the clinical diagnosis of LCA, and the clinically relevant finding is that 9% of a carefully defined LCA cohort harbor mutations in the CRB1 gene.

Less than 5 years ago, no patients with LCA could be molecularly diagnosed. With the addition of CRB1 to the panel of LCA genes, we now have a greater than 25% chance of detecting a disease-causing mutation in a new patient with LCA. Molecular diagnosis is important for providing accurate counseling about recurrence risk to affected families, as well as for the identification of specific molecular subsets of patients for future studies of novel interventions.

Accepted for publication January 5, 2001.

This study was supported in part by National Institutes of Health, Bethesda, Md, grants EY10539 and EY05627; the Foundation Fighting Blindness, Hunt Valley, Md; the Grousbeck Family Foundation, Stanford, Calif; the Carver Endowment for Molecular Ophthalmology, Muscatine, Iowa; the Daniel Matzkin Research Fund; the Grant Healthcare Foundation; and an unrestricted grant from Research to Prevent Blindness Inc, New York, NY. Dr Lotery is a recipient of a Research to Prevent Blindness Career Development Award.

The authors thank Creig Hoyt, MD; James Jan, MD; Byron Lam, MD; Ronald Carr, MD; John Heckenlively, MD; William Scott, MD; Douglas Fredrick, MD; Ehud Zamir, MD; Saul Merin, MD; Francis Munier, MD; Arlene Drack, MD; Terry Schwartz, MD; Jean Bennett, MD; Alessandro Iannaccone, MD; Maria Musarella, MD; and Benedetto Falsini, MD, for sharing their patients with us for this study. The authors also thank Luan Streb, BA; Christine Taylor, BS; Heidi Haines, MS; Louisa Affatigato, BS; and Gretel Beck, BA, for their excellent technical assistance and Jessica Emmons, BA; Elaine De Castro, BS; and Leigh Gardner, BA, for clinical coordination.

Corresponding author and reprints: Edwin M. Stone, MD, PhD, Department of Ophthalmology and Visual Sciences, The University of Iowa College of Medicine, 200 Hawkins Dr, Iowa City, IA 52242 (e-mail: edwin-stone@uiowa.edu).