![]() Received 8 March 2005 | Accepted 29 October 2005 | Published 2 November 2005 |
Download Reprint |
![]() |
Sequence variations in the retinal fascin FSCN2 gene in a Spanish population with autosomal dominant retinitis pigmentosa or macular degeneration
María José Gamundi,1 Imma Hernan,1 Miquel
Maseras,2 Montserrat Baiget,3 Carmen Ayuso,4 Salud
Borrego,5 Guillermo Antiñolo,5 José María
Millán,6 Diana Valverde,7 Miguel
Carballo1
1Servei de Laboratori, Laboratori de Biologia i Genètica Molecular, Hospital de Terrassa, Ctra. Torrebonica, Terrassa, Spain; 2Servei d'Oftalmologia de l'Hospital de Terrassa, Spain; 3Servei de Genètica Molecular, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 4Servicio de Genética, Fundación Jiménez Díaz, Madrid, Spain; 5Servicio de Genética y Diagnóstico Prenatal, Hospital Virgen del Rocío, Sevilla, Spain; 6Servicio de Genética, Hospital de la Fe, Valencia; 7Área de Genética, Facultad de Ciencias, Universidad de Vigo, La Coruña, Spain
Correspondence to: Miguel Carballo, Hospital de Terrassa, Carretera de Torrebonica s/n, 08227 Terrassa, Spain; Phone: +34937312420; FAX: +34937319045; e-mail: mcarballo@csdt.es
Abstract
Purpose: Only one mutation in the retinal fascin gene (FSCN2) has so far been associated with autosomal dominant retinitis pigmentosa (adRP) and macular dystrophy (adMD), in a Japanese population. Our study was designed to identify mutations in the FSCN2 gene among Spanish persons with adRP or adMD.
Methods: Denaturing gradient gel electrophoresis and direct genomic sequencing were used to evaluate the complete coding region and flanking intronic sequences of the FSCN2 gene for mutations in 150 unrelated adRP and 15 adMD index patients, and in 50 sporadic cases of retinitis pigmentosa, together with 50 controls. Ophthalmic and electrophysiological examination of retinitis pigmentosa patients and their relatives was carried out according to pre-existing protocols.
Results: Sixteen nucleotide substitutions were detected in the coding sequence of the index patients. Nine of these, His7Tyr, Ala122Thr, Ser126Phe, His138Tyr, Arg149Gln, Ala240Thr, Ala323Thr, Asn331His, and Phe367Leu are missense mutations, one is a nonsense mutation (Lys302Stop), and six are silent mutations. Co-segregation of the mutations in the families showed no direct relation between mutation and disease.
Conclusions: The photoreceptor-specific FSCN2 gene showed a relatively high number of sequence variations. The mutation 208delG in FSCN2, the only mutation so far associated with adRP or adMD, and which presumably causes a null allele, was not detected in these Spanish families. The nonsense mutation, Lys302Stop, detected in one adRP Spanish family is not the cause of the disease. These findings support the fact that the kind and frequency of the mutations depend on the ethnic population.
Introduction
Retinitis pigmentosa (RP) is the name given to a group of hereditary retinal degeneration diseases with a worldwide incidence of about 1 in 4000 individuals [1,2]. Clinical characteristics include night blindness, loss of the peripheral visual field, characteristic changes in the ocular fundus and depression of the normal ocular electrophysiological responses [3,4]. Genetically, RP is heterogeneous and the disorder may be inherited through an autosomal dominant (adRP), autosomal recessive (arRP), X-linked (XLRP) [5-7] or digenic trait [8]. Complex inheritance patterns such as tri-allelism [9] or uniparental disomy [10,11] have been reported. Mutations within seven genes (RHO, peripherin/RDS, ROM1, RP1, NRL, CRX, and FSCN2) that encode proteins specifically expressed in photoreceptor cells have been reported to cause adRP [12-18]. These proteins are involved with specific functions in the retina, such as the visual transduction cycle, structural components of the rod, and cone photoreceptor cells or transcription factors. Studies of genetic linkage and mutation detection have resulted in the characterization of mutations in genes with ubiquitous expression implicated in adRP, such as the pre-mRNA splicing factors PRPF8, PRPF31, PRPF3, and PAP-1 [19-22] or the IMPDH1 and CAIV genes [23,24]. The clinical and genetic heterogeneity of RP is demonstrated by the fact that different mutations within genes associated with adRP can cause substantially different retinal degeneration phenotypes. For example, mutations within the RDS gene have been associated with RP, different forms of the macular dystrophies or the cone-rod dystrophy that is also caused by some of the mutations found in the CRX gene [25]. However, an extreme example of the clinical heterogeneity has recently been reported in the FSCN2 gene; the only mutation so far found in FSCN2 associated with adRP (208delG) has recently been associated with autosomal dominant macular dystrophy (adMD) in a Japanese population [18,26]. Similar heterogeneity was seen for the mutation 1147delA in the arrestin gene in both Oguchi disease and arRP, also detected in Japanese patients and thus supporting the importance of ethnic variation [27].
The photoreceptor-specific gene FSCN2, located on chromosome 17q25, encodes 516 amino acids [28]. Retinal fascin is associated with the assembling of actin structures of the connecting cilium plasma membrane, and it plays an important role in photoreceptor disc formation [28,29]. The human FSCN2 and actin ACTB genes lie within 200 kb of each other on chromosome 17q25, although they seem to show independent gene regulation. Though possible linkage of FSCN2 to the RP17 allele at distal 17q has been excluded [30], a mutation in the FSCN2 gene associated with adRP and adMD has been found [18,26]. Analysis of this gene in western populations, however, is lacking. We screened for mutations in the retinal fascin FSCN2 gene in 150 adRP, 15 adMD families, and 50 sporadic cases of RP (SRP), together with 50 controls. Interestingly, although we detected a relatively high sequence variation in FSCN2, none of the mutations detected seem to be directly causative of retinal disease.
Methods
Patients
The participants were all Caucasian from different regions of Spain. We determined the pattern of disease inheritance of the patients from their family history and ophthalmic examination, which consisted of a funduscopic exam, visual field, visual acuity, dark-adapted sensitivity and, in most cases, electroretinograph analysis, all according to previously established protocols. Patients with autosomal dominant disease usually had an affected parent or child. Prior to inclusion, all the patients and their relatives who were participating were informed about the aims of the study, which was approved by all the participating institutions. Informed consent, which adhered to the tenets of the Declaration of Helsinki, was obtained from all the adults and from the children's parents or tutors.
Screening for mutations and sequencing conditions
Genomic DNA was prepared from peripheral blood lymphocytes using QIAmp DNA Blood Mini Kit (Qiagen, Valencia, CA). The coding region of the FSCN2 gene (NM_012418) was amplified using primers located in the flanking intronic region (Table 1). Screening for mutations in FSCN2 was carried out by denaturing gradient gel electrophoresis (DGGE). A CG clamp consisting of a sequence of 40 CG nucleotides was included in the 5' sequence of the forward or reverse primer to have a better resolution in the DGGE analysis. The running conditions of the DGGE are given in Table 1. Polymerase chain reaction (PCR) was performed in a 50 μl volume of buffer (20 mM Tris-HCl pH 8.55, 16 mM (NH)2SO4, 1.5 mM MgCl2 150 mg/ml BSA) containing 200-500 ng of human genomic DNA, 25 pmol each primer, 10 nmol each deoxyribonucleoside triphosphate, and 1.5 units of Taq polymerase (Ecotaq, Ecogen, Barcelona, Spain). Incubation was performed for 40 cycles consisting of 30 s (or 1 min) at 94 °C, 30 s (or 1 min) for annealing at different temperatures (Table 1) and 30 s (or 1 min) at 72 °C, followed by 1 min at 94 °C and 5 min at 72 °C. Electrophoresis of 8 μl of final PCR reaction volume was performed on a 1.5% agarose gel to test the amplification reaction. For DNA sequencing, PCR products were purified using Qiaquick columns (Qiagen). DNA sequencing was carried out using the OpenGene Automated DNA sequencing system from Visible Genetics and Thermo Sequenase Cy5.5 Dye Terminator Cycle Sequencing Kit (Amersham Pharmacia Biotech, Barcelona, Spain). The sequencing primers were the same as those used in PCR amplifications.
Results
Screening for mutations in the flanking and coding sequences of five exons of the FSCN2 gene was carried out by DGGE in 150 adRP and 15 adMD index patients, 50 SRP patients, and 50 controls. We detected 16 sequence variations in the adRP index patients (Table 2). Nine of these mutations are missense mutations, one produces a premature stop codon and the other six are silent mutations in the coding region. Only one missense mutation (His138Tyr), detected in FSCN2, occurred in a residue conserved among other members of the fascin gene family (Figure 1). Most of these sequence variations were either unique or had a low frequency in the population screened, and should thus be considered as rare sequences or variations, except the variation in the promoter region (34%) and the Ala323Thr polymorphism, which were also found in the control population (Table 2).
We checked co-segregation of all these sequence variations with adRP and adDM in the families in this study. We considered a major criterion of non co-segregation to be the absence of the mutation in one or more patients in the family. All the families proved informative for this criterion. Even the family carrying the nonsense Lys302Stop mutation in the FSCN2 gene lacked co-segregation (Figure 2). However, in this family we also detected the mutation Gly182Ser within the rhodopsin (RHO) gene. This mutation in RHO co-segregated with the disease in this family (Figure 2), and it has been reported to cause adRP in other families.
Discussion
The fascins are a highly conserved family of actin-bundling proteins identified across a wide range of species [31]. Fascin expression has been shown to be specific for distinct cell types, including neurons, macrophages, dendritic, and other cells. Fascins are thought to play a specialized role in forming or stabilizing dynamic cell extensions. The retina-specific fascin has been identified in bovine [32] and human [28] retinas. The human retinal fascin FSCN2 gene is located on chromosome 17q (17q25-qter), with close physical linkage to the cytoplasmic actin ACTG1 gene. Interestingly, the human FSCN1 gene shows a similar close linkage with the actin ACTB gene at 7p22, suggesting that these two fascin/actin genes derive from a single chromosome duplication event.
The fascin FSCN2 gene is a retina-specific gene. The promoter region of the human FSCN2 gene contains potential binding sites for retina-specific transcription factors. Thus, a potential retinoic acid response element (RARE) and consensus sequences for the retina-specific CRX and NRL transcription factors are observed in the flanking 5' sequences of the FSCN2 gene. Although the localization of FSCN2 to distal 17q is where retinitis pigmentosa 17 (RP17) was mapped [30], linkage analysis studies in two large RP17 families have shown that FSCN2 and RP17 are not linked [28]. However, FSCN2 is a candidate gene for RP and screening for mutations has been carried out in a Japanese population. One mutation, 208delG in FSCN2, was detected in four unrelated families (14 patients) with adRP [18]. This mutation was found in 3.3% of adRP patients in Japan, and it was suggested that it might be relatively common in Japanese patients with adRP. The identical 208delG mutation has recently been found in four patients from two adMD Japanese families [26], showing the remarkable clinical heterogeneity of this FSCN2 mutation. No other sequence variation in FSCN2 has been reported in association with retinal disease. However, our analysis of the FSCN2 gene in a Spanish population detected 14 sequence variations in the coding sequence and two polymorphisms, one in the promoter region (34%) and a second (13%), Ala323Thr, in the coding sequence. One of the mutations, His138Tyr, corresponds to a conserved residue in the retinal fascin of different species. This His138Tyr mutation in the conserved fascin residue was detected in a patient with macular degeneration. However, this mutation was also observed in the patient's unaffected father but was absent in one affected brother. We detected the His138Tyr mutation in a simplex case of RP. The rest of the missense mutations were also analyzed in the families, but none of the mutations could be directly associated with the disease, because one or more patients in the family were not carriers of the mutation (data not shown).
We detected a nonsense mutation, 904A->T, in the FSCN2 gene, which introduces a premature stop codon (Lys302Stop), presumably producing a truncated protein. If the translation product of this mutant allele is stable, nearly half the C-terminal lacks the encoded protein. While the Lys302Stop mutation does not seem to be causative of adRP, as seen from its co-segregation in the family (Figure 2), we nevertheless detected in this family an additional mutation (Gly182Ser) in the RHO gene, which has been previously reported in association with adRP and that co-segregates with RP (Figure 2). Thus, the cause of the disease in the family in question is probably this rhodopsin mutation rather than the mutation in the FSCN2 gene. Furthermore, clinical examination of the patients carrying both mutations in the rhodopsin and FSCN2 genes showed no significant differences in phenotype compared with the carriers of the mutation in the rhodopsin gene only.
Because neither the mutation 208delG nor Lys302Stop are located in the final exon of the FSCN2 gene, it is unlikely that their products could be translated stably [33]. Both mutations probably lead to the loss of a functional allele in the carriers. However, while the mutation 208delG causes retinopathy in a Japanese population, Lys302Stop is unlikely to be the cause of the disease in the Spanish family. Recently, targeted disruption of the FSCN2 gene in a mouse model has been reported to produce retinopathy [34]. Furthermore, the generation of one mouse line of homozygous 208delG showed no retinal expression of FSCN2 while heterozygous murine carriers of the mutation showed progressive photoreceptor degeneration [34]. These findings indicate a pathogenic mechanism of haploinsufficiency for the mutation 208delG. Whether the mutation Lys302Stop also produces haploinsufficency in a mouse model leading to retinal degeneration remains to be established. It is well known that truncated mutations located in exon 4 near the 5' region of RP1 cause adRP [35], and our results (data not shown). However, in a Chinese family a truncated mutation detected in the C-terminal region of RP1 did not cause the disease [36]. A similar situation is seen with the nonsense mutations of FSCN2, with the difference that the mutations in RP1 are located in the last exon of the gene, with a possible stable translation product, and a deleterious effect of such mutants is not ruled out, although haploinsufficiency has been postulated as the most probable pathogenic mechanism for these mutations [37].
In summary, our report of the mutational analysis of the FSCN2 gene in a Spanish population suggests similar analyses in other populations to clearly associate mutations in the FSCN2 gene with retinal degeneration. Compared with the other known genes associated with adRP, in our population the FSCN2 gene shows a relatively high proportion of sequence variation that is not correlated with retinal degeneration disease. This variation, together with the clinical heterogeneity shown for the 208delG mutation in a Japanese population, suggests that the frequency and kind of mutations depend on ethnic populations and raises the possibility that unknown genetic factors may be linked to FSCN2 and modulate its mutant expression in retinal degeneration. Further studies of phenotype and genotype correlation in persons with adRP and adMD carrying mutations in the FSCN2 gene will be necessary to clarify the situation.
Acknowledgements
We thank the families for their participation in this research and Ian Johnstone for editing this manuscript. This work was partially supported by grants from the Fondo de Investigaciones Sanitarias (01/0081-01, FIS G03/018), ONCE and Fundación ONCE.
References
1. Humphries P, Kenna P, Farrar GJ. On the molecular genetics of
retinitis pigmentosa. Science 1992; 256:804-8.
2. Dryja TP, Li T. Molecular genetics of retinitis pigmentosa. Hum
Mol Genet 1995; 4 Spec No:1739-43.
3. Berson EL. Retinitis pigmentosa. The Friedenwald Lecture. Invest
Ophthalmol Vis Sci 1993; 34:1659-76.
4. Wang Q, Chen Q, Zhao K, Wang L, Wang L, Traboulsi EI. Update on
the molecular genetics of retinitis pigmentosa. Ophthalmic Genet 2001;
22:133-54.
5. Inglehearn CF. Molecular genetics of human retinal dystrophies.
Eye 1998; 12(Pt 3b):571-9.
6. Bessant DA, Ali RR, Bhattacharya SS. Molecular genetics and
prospects for therapy of the inherited retinal dystrophies. Curr Opin
Genet Dev 2001; 11:307-16.
7. Farrar GJ, Kenna PF, Humphries P. On the genetics of retinitis
pigmentosa and on mutation-independent approaches to therapeutic
intervention. EMBO J 2002; 21:857-64.
8. Kajiwara K, Berson EL, Dryja TP. Digenic retinitis pigmentosa due
to mutations at the unlinked peripherin/RDS and ROM1 loci. Science 1994;
264:1604-8.
9. Katsanis N, Ansley SJ, Badano JL, Eichers ER, Lewis RA, Hoskins
BE, Scambler PJ, Davidson WS, Beales PL, Lupski JR. Triallelic
inheritance in Bardet-Biedl syndrome, a Mendelian recessive disorder.
Science 2001; 293:2256-9.
10. Thompson DA, McHenry CL, Li Y, Richards JE, Othman MI, Schwinger
E, Vollrath D, Jacobson SG, Gal A. Retinal dystrophy due to paternal
isodisomy for chromosome 1 or chromosome 2, with homoallelism for
mutations in RPE65 or MERTK, respectively. Am J Hum Genet 2002;
70:224-9.
11. Rivolta C, Sharon D, DeAngelis MM, Dryja TP. Retinitis pigmentosa
and allied diseases: numerous diseases, genes, and inheritance patterns.
Hum Mol Genet 2002; 11:1219-27. Erratum in: Hum Mol Genet. 2003;
12:583-4.
12. Dryja TP, McGee TL, Hahn LB, Cowley GS, Olsson JE, Reichel E,
Sandberg MA, Berson EL. Mutations within the rhodopsin gene in patients
with autosomal dominant retinitis pigmentosa. N Engl J Med 1990;
323:1302-7.
13. Farrar GJ, Kenna P, Jordan SA, Kumar-Singh R, Humphries MM, Sharp
EM, Sheils DM, Humphries P. A three-base-pair deletion in the
peripherin-RDS gene in one form of retinitis pigmentosa. Nature 1991;
354:478-80.
14. Kajiwara K, Hahn LB, Mukai S, Travis GH, Berson EL, Dryja TP.
Mutations in the human retinal degeneration slow gene in autosomal
dominant retinitis pigmentosa. Nature 1991; 354:480-3.
15. Sullivan LS, Heckenlively JR, Bowne SJ, Zuo J, Hide WA, Gal A,
Denton M, Inglehearn CF, Blanton SH, Daiger SP. Mutations in a novel
retina-specific gene cause autosomal dominant retinitis pigmentosa. Nat
Genet 1999; 22:255-9.
16. Pierce EA, Quinn T, Meehan T, McGee TL, Berson EL, Dryja TP.
Mutations in a gene encoding a new oxygen-regulated photoreceptor
protein cause dominant retinitis pigmentosa. Nat Genet 1999; 22:248-54.
17. Bessant DA, Payne AM, Mitton KP, Wang QL, Swain PK, Plant C, Bird
AC, Zack DJ, Swaroop A, Bhattacharya SS. A mutation in NRL is associated
with autosomal dominant retinitis pigmentosa. Nat Genet 1999; 21:355-6.
18. Wada Y, Abe T, Takeshita T, Sato H, Yanashima K, Tamai M.
Mutation of human retinal fascin gene (FSCN2) causes autosomal dominant
retinitis pigmentosa. Invest Ophthalmol Vis Sci 2001; 42:2395-400.
19. Vithana EN, Abu-Safieh L, Allen MJ, Carey A, Papaioannou M,
Chakarova C, Al-Maghtheh M, Ebenezer ND, Willis C, Moore AT, Bird AC,
Hunt DM, Bhattacharya SS. A human homolog of yeast pre-mRNA splicing
gene, PRP31, underlies autosomal dominant retinitis pigmentosa on
chromosome 19q13.4 (RP11). Mol Cell 2001; 8:375-81.
20. McKie AB, McHale JC, Keen TJ, Tarttelin EE, Goliath R, van
Lith-Verhoeven JJ, Greenberg J, Ramesar RS, Hoyng CB, Cremers FP, Mackey
DA, Bhattacharya SS, Bird AC, Markham AF, Inglehearn CF. Mutations in
the pre-mRNA splicing factor gene PRPC8 in autosomal dominant retinitis
pigmentosa (RP13). Hum Mol Genet 2001; 10:1555-62.
21. Chakarova CF, Hims MM, Bolz H, Abu-Safieh L, Patel RJ,
Papaioannou MG, Inglehearn CF, Keen TJ, Willis C, Moore AT, Rosenberg T,
Webster AR, Bird AC, Gal A, Hunt D, Vithana EN, Bhattacharya SS.
Mutations in HPRP3, a third member of pre-mRNA splicing factor genes,
implicated in autosomal dominant retinitis pigmentosa. Hum Mol Genet
2002; 11:87-92.
22. Maita H, Kitaura H, Keen TJ, Inglehearn CF, Ariga H, Iguchi-Ariga
SM. PAP-1, the mutated gene underlying the RP9 form of dominant
retinitis pigmentosa, is a splicing factor. Exp Cell Res 2004;
300:283-96.
23. Kennan A, Aherne A, Palfi A, Humphries M, McKee A, Stitt A,
Simpson DA, Demtroder K, Orntoft T, Ayuso C, Kenna PF, Farrar GJ,
Humphries P. Identification of an IMPDH1 mutation in autosomal dominant
retinitis pigmentosa (RP10) revealed following comparative microarray
analysis of transcripts derived from retinas of wild-type and Rho(-/-)
mice. Hum Mol Genet 2002; 11:547-57.
24. Rebello G, Ramesar R, Vorster A, Roberts L, Ehrenreich L, Oppon
E, Gama D, Bardien S, Greenberg J, Bonapace G, Waheed A, Shah GN, Sly
WS. Apoptosis-inducing signal sequence mutation in carbonic anhydrase IV
identified in patients with the RP17 form of retinitis pigmentosa. Proc
Natl Acad Sci U S A 2004; 101:6617-22.
25. Sohocki MM, Daiger SP, Bowne SJ, Rodriquez JA, Northrup H,
Heckenlively JR, Birch DG, Mintz-Hittner H, Ruiz RS, Lewis RA,
Saperstein DA, Sullivan LS. Prevalence of mutations causing retinitis
pigmentosa and other inherited retinopathies. Hum Mutat 2001; 17:42-51.
26. Wada Y, Abe T, Itabashi T, Sato H, Kawamura M, Tamai M. Autosomal
dominant macular degeneration associated with 208delG mutation in the
FSCN2 gene. Arch Ophthalmol 2003; 121:1613-20.
27. Nakamachi Y, Nakamura M, Fujii S, Yamamoto M, Okubo K. Oguchi
disease with sectoral retinitis pigmentosa harboring adenine deletion at
position 1147 in the arrestin gene. Am J Ophthalmol 1998; 125:249-51.
28. Tubb BE, Bardien-Kruger S, Kashork CD, Shaffer LG, Ramagli LS, Xu
J, Siciliano MJ, Bryan J. Characterization of human retinal fascin gene
(FSCN2) at 17q25: close physical linkage of fascin and cytoplasmic actin
genes. Genomics 2000; 65:146-56.
29. Saishin Y, Ishikawa R, Ugawa S, Guo W, Ueda T, Morimura H, Kohama
K, Shimizu H, Tano Y, Shimada S. Retinal fascin: functional nature,
subcellular distribution, and chromosomal localization. Invest
Ophthalmol Vis Sci 2000; 41:2087-95.
30. Bardien S, Ramesar R, Bhattacharya S, Greenberg J. Retinitis
pigmentosa locus on 17q (RP17): fine localization to 17q22 and exclusion
of the PDEG and TIMP2 genes. Hum Genet 1997; 101:13-7.
31. Tubb B, Bryan J. Fascin. In: Kreis T, Vale R, editors. Guidebook to the cytoskeletal and motor proteins. 2nd ed. Oxford University Press, Oxford; 1999. p. 91-4.
32. Saishin Y, Shimada S, Morimura H, Sato K, Ishimoto I, Tano Y,
Tohyama M. Isolation of a cDNA encoding a photoreceptor cell-specific
actin-bundling protein: retinal fascin. FEBS Lett 1997; 414:381-6.
33. Hentze MW, Kulozik AE. A perfect message: RNA surveillance and
nonsense-mediated decay. Cell 1999; 96:307-10.
34. Yokokura S, Wada Y, Nakai S, Sato H, Sagara y, Takahashi M, Nakamura y, Tamai M, Noda T. Target disruption of FSCN2 gene induces retinopathy in mice. ARVO Annual Meeting; 2005 May 1-5; Fort Lauderdale (FL).
35. Berson EL, Grimsby JL, Adams SM, McGee TL, Sweklo E, Pierce EA,
Sandberg MA, Dryja TP. Clinical features and mutations in patients with
dominant retinitis pigmentosa-1 (RP1). Invest Ophthalmol Vis Sci 2001;
42:2217-24.
36. Xiaoli Z, Weiling F, Pang CP, Yeung KY. Screening for mutations
in a novel retinal-specific gene among Chinese patients with retinitis
pigmentosa. Chin Med Sci J 2002; 17:225-30.
37. Liu Q, Lyubarsky A, Skalet JH, Pugh EN Jr, Pierce EA. RP1 is
required for the correct stacking of outer segment discs. Invest
Ophthalmol Vis Sci 2003; 44:4171-83.