New ABCR Mutations and Clinical Phenotype in Italian Patients with Stargardt Disease

Abstract

purpose. To assess the mutation spectrum in the ABCR gene and clinical phenotypes in Italian families with autosomal recessive Stargardt disease (STGD1) and fundus flavimaculatus (FFM).

methods. Eleven families from southern Italy, including 18 patients with diagnoses of STGD1, were clinically examined. Ophthalmologic examination included kinetic perimetry, electrophysiological studies, and fluorescein angiography. DNA samples of the affected individuals and their family members were analyzed for variants in all 50 exons of the ABCR gene by a combination of single-strand conformation polymorphism analysis and direct sequencing techniques.

results. Ten ABCR variants were identified in 16 (73%) of 22 mutant alleles of patients with STGD1. Five mutations of 10 that were found had not been previously described. The majority of variants represent missense amino acid substitutions, and all mutant alleles cosegregate with the disease in the respective families. These ABCR variants were not detected in 170 unaffected control individuals (340 chromosomes) of Italian origin. Clinical evaluation of these families affected by STGD1 showed an unusually high frequency of early age-related macular degeneration (AMD) in parents of patients with STGD1 (8/22; 36%), consistent with the hypothesis that some heterozygous ABCR mutations enhance susceptibility to AMD.

conclusions. Patients from southern Italy with Stargardt disease show extensive allelic heterogeneity of the ABCR gene, concordant with previous observations in patients with STGD1 from different ethnic groups. Half the mutations identified in this study had not been previously described in patients with STGD1. Screening of increasingly large numbers of patients would help to determine whether this can be explained by ethnic differences, or is an indicator of extensive allelic heterogeneity of ABCR in STGD1 and other eye diseases. In 6 (55%) of 11 families, the first-degree relatives of patients with STGD1 were diagnosed with early AMD, supporting the previous observation that some STGD1 alleles are also associated with AMD.

Stargardt disease (including fundus flavimaculatus, FFM) is one of the most common causes of macular disease in childhood and accounts for approximately 7% of all retinal dystrophies. ^{1 2} STGD1 is a form of autosomal recessive macular degeneration characterized by diminished central visual acuity in the first several decades of life; the appearance of small, yellowish lesions or flecks at the level of the retinal pigment epithelium (RPE) at the posterior pole; and atrophic changes in the macula. ³ Recently, an ATP-binding cassette (ABC) transporter gene, ABCR, was localized to chromosome 1p22 and fully characterized. ⁴ This gene is expressed exclusively and at high levels in the retina in rod but not cone photoreceptors. Mutations in ABCR have been described in a number of inherited eye disorders, including STGD-FFM, ⁴ retinitis pigmentosa 19, ⁵ cone–rod dystrophy, ⁶ and age-related macular degeneration (AMD). ⁷ Although several independent studies have confirmed that ABCR is the causal gene in STGD-FFM, ^{4 8} its involvement in AMD is currently under investigation. ^{9 10 11 12} All studies of ABCR in eye diseases report a broad mutation spectrum and high allelic heterogeneity. ^{4 13 14 15} It has been proposed that some of this phenomenon may be due to the ethnic variability of populations studied. ^{9 15} To evaluate the variation of ABCR alleles in an Italian population affected with STGD1, we studied 11 families segregating STGD1 or STGD-early AMD. 

Materials and Methods

Patients

Eleven families (Fig. 1) , some members of which were affected with autosomal recessive Stargardt disease or FFM, ¹⁶ were ascertained through the Retina Research Center of the Eye Clinic, Second University of Naples, Italy, during 1997 and 1998, from a pool of 500 families with hereditary chorioretinal dystrophies. Families were not selected specifically because of observable parental abnormalities. Research procedures were in accordance with institutional guidelines and the Declaration of Helsinki. Informed consent was obtained from all patients after the nature of procedures to be performed was explained fully. All 36 individuals listed in Table 1 underwent complete eye examinations. Examinations included history of the patient and his or her family, visual acuity, central and peripheral visual fields, fluorescein angiography, electro-oculography, electroretinography, and color vision. Visual acuity was examined with Snellen visual chart, visual fields were examined with Goldmann kinetic perimetry, and electroretinography was recorded by means of a corneal contact lens electrode with Ganzfield stimulator according to international clinical standards. ¹⁷ The color test was performed using Ishihara tables. 

STGD1 and FFM were defined according to the description by Gass. ¹⁸ Briefly, the essential features of STGD1-FFM were considered to be the following: 1) retinal disorder in a family with more than one affected individual and compatible with autosomal recessive inheritance; 2) onset of symptoms in childhood or early adulthood; 3) bilateral central vision loss with “beaten-metal” foveal changes and/or yellow-white “fish-tail” flecks scattered through the posterior pole and peripheral retina (FFM); 4) normal caliber of the retinal vessels and no pigmented bone spicules in the retinal periphery; 5) normal electroretinogram; and 6) typical dark choroid in fluorescein angiography. Some patients had all essential features of STGD1 except for the absence of the typical dark choroid.“ Early” age-related maculopathy was defined as the presence of soft indistinct or reticular drusen or as the presence of any drusen type except hard indistinct, with RPE degeneration or increased retinal pigment in the macular area and in the absence of signs of “late” age-related maculopathy (neovascularization or geographic atrophy). ^{2 19 20}  

Mutation Detection

All 50 exons of the ABCR gene ²¹ were screened for mutations by a combination of single-strand conformation polymorphism (SSCP) and heteroduplex analysis in all affected probands from 11 families segregating STGD1. In all cases in which a pattern different from the wild-type SSCP was identified, the corresponding exon was sequenced. Sequencing was performed with a kit (Taq Dyedeoxy Terminator Cycle Sequencing; Applied Biosystems, Foster City, CA), according to the manufacturer’s instructions. Sequencing reactions were resolved on an automated sequencer (model 373A; Applied Biosystems). Segregation analysis and screening of controls were performed by SSCP ²² analysis under optimized conditions. Genomic DNA samples (50 ng) were amplified (AmpliTaq Gold polymerase; Perkin–Elmer, Foster City, CA) in 1× polymerase chain reaction buffer supplied by the manufacturer in the presence of[α -³²P] dCTP. Samples were heated to 95°C for 10 minutes and amplified for 35 to 40 cycles of 96°C, 20 seconds; 58°C, 30 seconds; and 72°C, 30 seconds. Products were diluted in 1:3 stop solution, denatured at 95°C for 10 minutes, chilled in ice for 5 minutes, and loaded on gels. Gel formulations include 6% acrylamide-Bis (2.6% cross-linking), 10% glycerol at room temperature, 12 W; and 10% acrylamide-Bis (1.5% cross-linking), at 4°C, 70 W. Gels were run for 2 to 16 hours (3000 volt-hours/100 bp), dried, and exposed to x-ray film for 2 to 12 hours. Sequences of all primers used for SSCP have been deposited with Human Genome DataBase (GDB). ²²  

Results

Molecular Analysis

Ten variants were identified in 16 (73%) of 22 alleles (Table 1) . None of these variants was detected in 170 control individuals (340 chromosomes) from Italy and had not been identified in more than 400 control individuals in previous studies. ^{4 7 14} Mutations were not identified by the means of standard mutation detection techniques (SSCP and direct sequencing) on the remaining six alleles. Eight (80%) of 10 variants were missense alterations. The remaining two were a splice site mutation and an insertion, resulting in a frameshift (Table 1) . Five mutations were unique for this study: missense changes V767D (exon 15), T897I (exon 18), E1399K (exon 28), insertion of four nucleotides CAAA at position 250 in exon 3, and a splice site variant 5018 + 2T/C in exon 36. These five variants accounted for 8 of 16 identified mutant alleles (Table 1) . Both disease-causing alleles were identified in 6 (55%) of 11 families, including two families in which mutations were identified in homozygous state, indicating either known (pedigree 431) or possible (pedigree 631) cases of consanguinity (Fig. 1 , Table 1 ). 

Genotype–Phenotype Correlation

It has been proposed that patients with FFM do not possess inactivating mutations in the ABCR gene ¹³ and that patients with earlier onset of STGD1 tend to have mutations in the 5′ portion of the gene. ¹⁴ In our study, we identified only one ABCR allele in two pedigrees that could be unequivocally classified as a truncating mutation (250insCAAA, exon 3; Table 1 ). In one case (pedigree 632), the patient was found to have STGD/FFM in the first decade of life (Table 1 , Fig. 2 ), which contradicts the observation by Rozet et al. ¹³ In the other case (pedigree 260; Table 1 , Fig. 1 ), two siblings were diagnosed with STGD1 relatively late in life (ages 35 and 38, Figs. 3A 3B ), which is clearly an exception to the observation made by Lewis and others. ¹⁴ These patients present all essential features of STGD1 except the absence of the typical dark choroid in fluorescein angiography. However, detection of ABCR mutations in both alleles of these patients unequivocally supports the diagnosis of Stargardt disease. Several cases of patients with STGD1 have been reported with the absence of dark choroid, including those who possessed mutations in the ABCR gene. ²³ Taken together, these recent findings modify the diagnosis of Stargardt macular dystrophy. Of interest, the second allele in this pedigree was identified as harboring the G1961E mutation. Moreover, in the two other families heterozygous for the G1961E variant (pedigrees 624 and 636, Table 1 ) the age of diagnosis was 15 years or more. 

Clinical data are reported in Table 1 . Ten of 18 patients with STGD also manifested FFM. The age of onset for patients corresponded mainly to the first decade, except for the families 624, 628, 636, and 260 in which the clinical appearance ranged from the second to the fourth decades of life. A wide spectrum of macular changes was present, including either a minimal ophthalmoscopic loss of the foveal reflex or major changes in the central macular area with widespread atrophy with a beaten-metal appearance or marked geographic atrophy with choriocapillaris involvement. The degree and pattern of RPE atrophy in the central macular area also varied and did not correlate with the degree of visual loss. Characteristic yellowish RPE flecks around the fovea were sometimes scattered throughout the posterior pole and extended to the midperiphery (Fig. 3A) . Flecks presented either an irregular pattern of fluorescence or appeared nonfluorescent. Fluorescein angiography showed areas of dark or silent choroid near the posterior pole and midperiphery. The electroretinogram and the electro-oculogram were of normal or slightly reduced amplitude in all patients and therefore of limited diagnostic value in patients with STGD1. 

Clinical evaluation of families affected by STGD1 showed an unusually high frequency of early AMD, in 8 (36%) of 22 parents of patients with STGD1 (Figs. 3 4 and 5) . Among those individuals, there were 6 women and 2 men with the mean age of 56 years. These subjects showed a disorder of the macular area of the retina characterized by any of the following primary features: areas of increased pigment or hyperpigmentation or areas of depigmentation or hypopigmentation of the RPE, with no visible choroidal vessels, associated with drusen. These changes appeared not to be secondary to another disorder (e.g., ocular trauma, retinal detachment, high myopia, chorioretinal infective or inflammatory process, or choroidal dystrophy) Fluorescein angiography revealed hyperfluorescent and/or hypofluorescent areas in the macular region (Figs. 3C 4B) . In addition, in the consanguineous 431 pedigree the father of the patients with STGD1, as well as his mother, had high myopia with chorioretinal degeneration and macular involvement. 

Discussion

Eleven families from southern Italy segregating Stargardt disease were analyzed for mutations in the ABCR gene. The detection rate of allelic variants was approximately 73%, favorably correlating with mutation detection rates from analogous studies, ^{14 15} in which similar techniques (SSCP and direct sequencing) were used. Alternative methods for detection of large genomic rearrangements (e.g., Southern blot analysis) were not used in this study. Analysis of mutations in ABCR in Italian patients with STGD1 showed broad allelic heterogeneity and a prevalence of missense mutations, a pattern similar to that described for patients from other ethnic groups. ^{4 13 14 15} Of note, half the mutations (5/10) identified in this study had not been described in patients with STGD1. ^{4 13 14 15} Screening of increasingly large numbers of patients would help to determine whether this can be explained by ethnic differences or is an indicator of extensive allelic heterogeneity of ABCR in STGD1 and other eye diseases. Occurrence of two newly identified alterations (250insCAAA and 5018 + 2T/C) in 2 independent families of only 11 studied (Table 1) suggests that these variants could be specific for, or more prevalent in, the Italian population. 

The previously reported G1961E mutation segregated with the disease in 3 (27%) of 11 families (Fig. 1) . This alteration has been reported to be one of the most frequent variants (>10%) in patients with STGD1 of white origin in North America. ^{4 11 14} In addition, this variant has been associated with AMD. ^{7 12} In two of three pedigrees (260 and 636), parents of patients with STGD1 harboring this mutation in heterozygous state had diagnoses of early AMD. In the third family (pedigree 624), the only living parent did not carry the alteration and was declared disease-free at the age of 74, whereas the other parent had died at 60 years of age with no known ophthalmic information available. Although limited numbers and the relatively early age (below or near 60 years) of individuals studied prevents us from making definitive conclusions, this observation supports the association of the G1961E variant with AMD. It is of interest that in all three families segregating the G1961E allele patients were found to have Stargardt disease at the age 15 or more (Table 1) . Furthermore, in family 260, the disease developed in two siblings at a relatively late age (35–38 years), although the second ABCR allele in this case is most likely inactivated because of an insertion, resulting in a frameshift. To date, no patient with STGD1 homozygous for the G1961E mutation has been identified, although this mutation is one of the most frequent in patients with STGD (>10%), and the number of screened patients with STGD1 exceeds several hundred. ^{4 11 14} Altogether, these data allow classifying the G1961E mutation as a“ mild” alteration ¹⁵ and individuals homozygous for this variant may not manifest the Stargardt disease phenotype. 

In summary, 10 mutations in the ABCR gene were identified in 16 alleles of 22 possible in 11 Italian pedigrees. Five mutations were unique for this patient collection and had not been described before. The G1961E variant was found to be the most frequent in Italian patients with STGD, in correlation with the previous data from patient collections with different ethnic backgrounds. New information about the clinical manifestations of ABCR mutations and knowledge of the complete mutation spectrum is essential to understanding the pathophysiology of Stargardt disease and retinal dystrophy in general. 

 Supported in part by a grant from “Regione Campania legge 41” (ER). Research of Rando Allikmets is supported by a Research to Prevent Blindness Career Development Award and The Ruth and Milton Steinbach Fund.

 Commercial relationships policy: N.

 

Figure 1.

 

Pedigrees of families with STGD1. Squares indicate males, circles indicate females, and diamonds indicate sex unknown. A diagonal line indicates a deceased individual. A double horizontal line between a mating pair indicates consanguinity. The family number is shown above each pedigree. Specific individuals in a given family are identified by a number below the symbol. Filled symbols represent individuals affected with STGD1. A dot within a symbol indicates a diagnosis of early AMD. A square within a symbol indicates a diagnosis of high myopia. Alphabetical small letters indicate mutation analysis.

Table 1.

 

Clinical Characteristics and Segregation of Mutations in 11 Italian STGD/FFM Pedigrees

Pedigree	Patients	Age	Age of Onset	Visual Acuity	Diagnosis	Allele 1	Allele 2
431	431 S	18	10	20/200	STGD	R212C	R212C
	433 D	29	8	20/200	STGD	R212C	R212C
	432 F	63	10	LE 20/40RE LP	High myopia with macular involvement	R212C	R212H(p)
	858 Gm	87	10	LP	High myopia with macular involvement	wt	R212H(p)
	774 M	60	58	20/25	Pigmentary abnormalities and drusen	R212C	wt
	260 D	41	35	20/200	STGD	250▿CAAA	G1961E
	759 S	39	38	20/100	STGD	250▿CAAA	G1961E
	760 M	60	57	20/40	Pigmentary abnormalities and drusen	wt	G1961E
	761 Gs			20/20	Normal	wt	G1961E
	762 Gs			20/20	Normal	250▿CAAA	wt
631	631 S	18	3	20/200	STGD / FFM	5018+ 2T → C	5018+ 2T → C
	777 F	59	58	20/20	Soft distinct drusen	5018+ 2T → C	wt
	779 M	52	50	20/20	Hard distinct drusen	5018+ 2T → C	wt
624	624 D	40	18	20/200	STGD	R1640Q	G1961E
	625 S	36	20	20/200	STGD	R1640Q	G1961E
	834 M	74		20/20	Normal	R1640Q	wt
636	636 S	22	15	20/400	STGD / FFM	E1087K	G1961E
	778 M	43	43	20/20	Hard distinct drusen	wt	G1961E
632	632 D	24	8	20/200	STGD / FFM	250▿CAAA	V767D
628	628 S	27	18	20/200	STGD	T897I	N/D
	4 F	63		20/20	Normal	wt	N/D
	5 M	62	62	20/20	Pigmentary abnormalities and drusen	T897I	N/D
633	633 D	12	8	20/400	STGD / FFM	A1038V	N/D
	776 S	15		20/20	Normal	A1038V	N/D
	634 D	20	10	20/400	STGD / FFM	A1038V	N/D
	3 F	60	60	20/20	Pigmentary abnormalities and drusen	wt	N/D
	2 M	49		20/20	Normal	A1038V	N/D
615	615 S	22	8	20/200	STGD / FFM	5018+ 2T → C	N/D
	616 D	23	10	20/200	STGD / FFM	5018+ 2T → C	N/D
	764 D	25		20/20	Normal	5018+ 2T → C	N/D
	763 F	60		20/20	Normal	5018+ 2T → C	N/D
	765 M	55	55	20/20	Pigmentary abnormalities and drusen	wt	N/D
629	629 D	23	10		STGD / FFM	E1399K	N/D
622	627 D	47	12	20/400	STGD	N/D	N/D
	622 D	35	8	20/400	STGD / FFM	N/D	N/D
	623 C	19	10	20/200	STGD / FFM

 S, son; D, daughter; F, father; M, mother; C, cousin; Gm, grandmother; Gs, grandson; (p), polymorphism; wt, wild type; N/D, not determined.

Figure 2.

 

Fluorescein angiograms from patients with STGD-FFM in pedigree 632. Note a ring of faint hyperfluorescence in the macular region sparing the fovea, coalescent patches on midperipheral fundus, and dark choroid.

Figure 3.

 

Fluorescein angiograms from patients in pedigree 260. (A) Patient 260. RPE atrophy of the macular region with rare paracentral flecks. No detectable dark choroid. (B) Patient 759. A ring of faint hyperfluorescence in the macular region. No dark choroid (corneal opacities interfere with a good visualization of retinal alteration). (C) Patient 760. Note hyperfluorescent and hypofluorescent areas in the macular region for RPE atrophy and hypertrophy. Mild stretching of retinal vessels for initial preretinal fibrosis.

Figure 4.

 

Fluorescein angiograms from two patients in pedigree 636. (A) Patient 636 with STGD. Dark choroid and transmitted hyperfluorescence centered at the fovea, indicating RPE atrophy. (B) Patient 778 with early AMD. Hyperfluorescent small area for RPE atrophy near the fovea and hyperfluorescent spots in the posterior pole corresponding at drusen.

Figure 5.

 

Fluorescein angiograms from two patients in pedigree 615. Patient 765 with early AMD. Scanning laser ophthalmoscope angiography. Hyperfluorescent and hypofluorescent areas for RPE atrophy and hypertrophy. Patient 616 with STGD-FFM. Dark choroid and transmitted hyperfluorescence centrally and extending in coalescent patches to the midperipheral fundus.