Clinical and genetic findings in Hungarian patients with X-linked juvenile retinoschisis

Navigate by section:

• Introduction
• Methods
• Results
• Discussion

Abstract

Purpose: To determine clinical phenotypes, examine the age dependency of X-linked juvenile retinoschisis (XLRS), and identify mutations in the retinoschisis1 gene (RS1) in 13 Hungarian (Caucasian) families with this disease.

Methods: This study included 72 members in 13 families. Complete ophthalmological examinations, including optical coherence tomography (OCT) and full-field and multifocal electroretinography (ERG), were performed on 20 affected males, 13 female carriers, and 27 healthy controls. The patients were divided into two age groups (Group I <25 years and Group II >25 years), retrospectively, to assess the possible effects of age. Correlations among genotype, age, best corrected visual acuity (BCVA), OCT, and ERG results were analyzed. A modified classification scheme was done to identify the different phenotypes of the disease. In each of the 72 family members and 100 age-matched male controls, all exons and introns of RS1 were amplified by polymerase chain reaction (PCR) and directly sequenced.

Results: Foveal retinoschisis was detected in 25 eyes (62.5%) of patients by funduscopy, and in 29 eyes (72.5%) by OCT, while macular lamellar schisis was recognizable only by OCT in 30 eyes (75%) of patients. Foveal thickness (FT) and total macular volume were significantly increased in younger (Group I) patients only. For patients younger than 26 years, large inner nuclear central cysts were observable by OCT, while after 26 years, foveas were atrophic. White flecks and dots, which were like that seen in fundus albipunctatus, were detected in both eyes of one patient. In both patient groups, characteristically decreased b-waves of standard combined ERG were recorded without any significant difference between the patient groups. The BCVA and ERG parameters of all patients and the OCT of younger patients were significantly worse (p<0.05) than those of age-matched controls. A significant difference between the two age groups was found in case FT, total macular volume, and amplitudes of rod b-wave only. Moderate negative correlation (r=-0.54, p<0.001) was detected between age and FT, while only low negative correlation (r=-0.33, p<0.05) was detected between age and standard combined b-wave amplitudes of full-field ERG. BCVA LogMAR did not show any obvious correlation with age (r=-0.14, p=0.39) or with the type of mutation. Nine different mutations were identified in 25 male patients and 31 female carriers of 13 families: six known and one novel missense mutation (c.575C>T, p.Pro192Leu), one insertion mutation (c.579dupC, p.Ile194Hisfs29ext43), and one frameshift, causing splice site mutation (c.78+1G>C) were detected. These mutations were absent in the 100 age-matched male control samples.

Conclusions: Foveal cystic schisis was found more often by OCT than by funduscopy (+10%), while flat macular lamellar schisis was recognizable only by OCT. Advancing age inversely influenced the size of cavities (FT), and standard combined b-wave amplitudes of full-field ERG, while BCVA, response density, and implicit times of multifocal electroretinography did not show any obvious correlation with age. The atrophic stage of the disease was observable after 26 years of age. The lesions that appeared to be indicative of fundus albipunctatus were proven to be palisades between the splitted retinal layers. Our modified classification scheme was helpful in assessing the prevalence of disease types. In these Hungarian patients, one novel and eight known mutations were detected. The distribution of mutations in RS1 was different to that reported in the literature, because the greatest number of different mutations was in exon 6 instead of exon 4. Two mutation hot spots were found: between c.418–422 in exon 5 and between c.574–579 in exon 6. Genotype-phenotype correlation was not demonstrable.

Introduction

X-linked juvenile retinoschisis (XLRS; OMIM: 312700) is one of the most common X-linked recessively inherited, bilateral, progressive vitreoretinal dystrophies. Limited almost exclusively to males, the incidence is between 1:5,000 and 1:25,000 [1–3]. Macular changes are present in almost all cases [4]. In the fundi, radially oriented intraretinal foveomacular cysts are seen in a spoke-wheel configuration, with the absence of foveal reflex in most cases (Figure 1A). In addition, approximately half of cases have bilateral peripheral retinoschisis in the inferotemporal part of the retina (Figure 1B) [4]. Aside from the typical fundus appearance, strabismus, nystagmus, axial hyperopia, defective color vision (red-green dyschromatopsia) and foveal ectopy can be present [4,5]. The most important complications were vitreous hemorrhage, retinal detachment, and neovascular glaucoma [1,4,6]. The diagnosis of the disease is most often made in boys aged 5–10 years, who may also exhibit symptoms of uncorrectable visual disturbance and reading difficulties [4].

Optical coherence tomography (OCT) images can be variable depending on the disease stage. The cysts are located in different layers of the retina (Figure 1C), namely the nerve fiber (NFL), ganglion cell (GCL), inner nuclear (INL), outer plexiform (OPL), outer nuclear (ONL) and photoreceptor layers (PRL) [7–9]. In the course of time the coalescing cysts form a large central cavity (Figure 1C), which can pass into a nonspecific macular atrophy (Figure 1D) [3,5,10,11]. Disease severity and progression are highly variable.

Full-field electroretinograms (ERGs) of the affected individuals are characterized by predominant reduction of standard combined b-wave amplitude (negative-type ERG: b-wave to a-wave ratio less than or equal to 1) in the dark-adapted eye (Figure 1E) [2,12]. The a-wave may also be reduced with disease progression due to the increasing retinal pigment epithelium (RPE) atrophy [4,13]. The response densities of multifocal ERG (mfERG) diminish mainly in the central rings (Figure 1F) and the implicit times are extended. Thus, ERG is a useful diagnostic method in almost all cases [2].

XLRS is caused by mutation in the RS1 gene (localization: Xp22.2-p22.1, GenBank AF014459), which was identified by positional cloning in 1997 (Figure 2) [14]. This is the only known gene to be associated with XLRS so far [15]. RS1 has six exons and encodes a 224 amino acid (AA) secretable extracellular adhesion protein, called retinoschisin (RS1), which is primarily present in photoreceptors and in bipolar cells [6]. It may be involved in cellular adhesion and cell-cell interactions on membrane surfaces [16]. Retinoschisin includes a secretory leader sequence (LS; 23 AA), an Rs1 domain (Rs1D; 39 AA), a highly conserved discoidin domain (a main structural feature of RS; 157 AA), and a C-terminal segment (5 AA) [2,16–19]. The discoidin domain is essential for the normal development of the retina, and if functional RS1 is lacking, schisis develops in different retinal layers [14,15].

Methods

Informed consent was obtained from all participants after a full explanation of the nature and possible consequences of the study. All studies were conducted in accordance with the tenets of the Declaration of Helsinki. Molecular genetic examinations were approved by the Ethics Committee for Human Genome Research of Semmelweis University.

Results

Discussion

The BCVA of patients was significantly impaired from childhood and, contrary to the results of Pimenides et al., [21] further significant impairment was not detectable with age. Although foveal abnormalities were found in all patients using both indirect funduscopy and OCT, foveal cystic schisis was more often found (+10%) by the OCT, and appeared to mainly occur in the INL and PRL retinal layers. Aside from the presence of cysts in GCL and INL, significant impairment of BCVA was detected if photoreceptors (PRL) were affected also. The flat macular lamellar schisis was recognizable only by OCT. Presence of cysts was evident in more than 3/4 of the eyes, while the nonspecific atrophic form in the macula was found in almost 1/4 of the eyes. The high overall prevalence of the cystic form in these results is undoubtedly due to the relatively young age of our patients (mean 24.5±16.6 years). The most common phenotype of the disease was the association of foveal cysts and macular lamellar schisis (type 3). The prevalence of peripheral retinoschisis was found to be lower than generally reported in the literature [4]. White flecks and dots, characteristic of fundus albipunctatus, were associated with the biggest foveal and peripheral retinoschisis. Lesions that are like that seen in fundus albipunctatus are a known, uncommon fundus appearance in XLRS [22]. Our calculations confirmed that white lesions seemed to be identical to palisades visible by OCT between the splitted retinal layers.

Patients’ OCT images differed greatly, and a large interocular and intrafamilial variability was found. The time-course of the disease can be characterized by the OCT parameters (Figure 4). This study is the first to present the exact time of changing from the cystic to atrophic stage of the disease. Before 26 years of age, large inner nuclear central cysts (with significantly elevated FT and TMV) were characteristic (Figure 1C). After 26 years of age a nonspecific foveal and perifoveal atrophy with unchanged parafoveal retinal thickness was more common (Figure 1D). In older age the whole macula was markedly atrophic.

Regarding the change from the cystic to atrophic stage (after 26 years), a moderate negative correlation was detectable between age and FT; but within each stage there was only low and little negative correlation with age (Figure 4). Excluding the remarkable FT change about 26 years of age the aforedescribed results showed a rather limited age dependency of FT alterations. Examining the possible functional effects of the morphological changes found by OCT between the younger and older patient groups, there were no significant differences in BCVA, in amplitudes and implicit times of full-field (except the rod b-wave amplitudes) and mfERGs. The lack of correlation between OCT and BCVA was probably due to inability to fixate and the common eccentric fixation. As with other reports, we found the changing of central macular region visible by OCT was not accompanied by further visual impairment [12,23].

In two boys (family 1 IV:2, 12 years of age; family 2 III:1, 7 years of age) improvement of BCVA (three lines on the Snellen chart) was detected during a one-year follow-up, but without any notable structural changes (OCT), or functional improvement (ERG) of the retina. This was probably due to more efficient use of the ectopic parafoveal retina, and would occur only in young patients as a result of the plasticity of the developing visual system.

While previous research has demonstrated that RS1 is synthesized mainly in and secreted from the inner segment of photoreceptors [6,24–26], Takada et al. identified in healthy mouse retina that RS1 mRNA was expressed in the ganglion cells first, and later in all remarkable cells locally in the outer layers of the retina [26]. This molecular mechanism explains why retinoschisis may affect several different layers of the retina [26]. In agreement with the latest publications we found retinoschisis mainly in the INL (Müller and bipolar cells) and in the PRL, using OCT [7,9,12,27]. The dysfunctional, defective RS1 may accumulate in these layers, both intracellular and extracellular layers, leading to cysts and schisis [3,21]. In cases of nonspecific foveal atrophy, electrophysiological and molecular genetic examinations are needed to confirm the diagnosis.

Nonspecifically all the amplitudes of scotopic and photopic responses of full-field ERG were significantly decreased in both patient groups compared with the controls. The only appropriate clinical sign to confirm the diagnosis of XLRS was the pronouncely decreased standard combined b-wave amplitudes, and in advanced cases negative type full-field ERGs (in 50% of patients). Significant difference between full-field ERG responses of the two patient groups was found in case rod b-wave amplitudes only. Standard combined b-wave amplitudes of full-field ERG showed a low negative correlation with age, while BCVA LogMAR did not show any obvious correlation with age or mutation type (Figure 8).

The mainly centrally decreased P1 amplitudes of mfERG probably indicate the lack of functioning cells in the central part of the foveal schisis, while the mainly peripherally delayed implicit times may be caused by the functionally damaged but still living cells. Previous results, based on ERG examinations, found Müller cells to be primary locations of the disease [28]. Recent results demonstrate that the mfERG (FOK) P1-waves originate mainly from ON- and OFF-bipolar cells, with a smaller contribution from cone photoreceptors [28]. Thus bipolar cells seem to be affected rather than Müller cells in which, furthermore, no RS immunoreactivity was detected [24,26]. Our abnormal ERG responses also suggest that the retinal damage affects mainly the INL and the PRL layers, where the biggest schisis was found by OCT and where, normally, the RS1 can be detected in the greatest amount [12,24,25]. Flat schisis and functional impairment was detected by OCT and mfERG in a greater part of the retina than would have been expected by funduscopy.

To identify patients with uncertain diagnosis and carriers without any symptoms, molecular genetic examinations are necessary (90% certainty) [29]. On the basis of previous genetic examinations, most of the mutations are missense mutations and localized to exons 4–6 of RS1 within the discoidin domain (dmd) [2]. Exons 1–3 tend to have mainly translation-truncating nonsense mutations [15].

This article is the first study to estimate the genetic background of patients with XLRS in Hungary. All the tested patients and carriers were found to have RS1 coding or splice site mutations. Most of the mutations were found to be localized in the conservative discoidin domain, which caused protein misfolding and finally retention of RS1 in the endoplasmic reticulum [18]. The remaining mutations were localized in the RS1 domain, which interfered with the normal oligomerization of RS1 [18]. Mutations in the leader sequence (not found in our study) prevented the translocation of the polypeptide chain across the endoplasmic reticulum membrane as part of the secretion process [18]. In each case the mutant RS1 was unable to function as an extracellular cell-adhesion protein.

Of the nine different mutations found, the c.421C>T missense mutation (exon 5) was detected most frequently. That a de novo mutation (c.421C>T) was found in patient II:2 of family 4 together with the already known disease-causing variant (c.574C>T, inherited from his mother), illustrates the remarkable new mutation rate of RS1 and the high mutation heterogeneity of the disease. Although the aforedescribed case with simultaneous presence of two mutations and the consequent two AAs substitution led to pronounced change in the tertiary structure of the RS1 (based on the CBS CPH model), neither the clinical condition nor the examination results were affected more than in those cases with separate presence of these two mutations. It is presumed the two mutations led to intracellular degradation of the RS1 even separately [21,30]. Considering that female carriers are unaffected, it seems that the absence of a functional RS1, rather than the presence of a mutant protein, was responsible for retinoschisis in affected males [31].

We highlight the fact that a novel missense mutation (c.575C>T) was found in family 7. Although missense, insertion, and splice site mutations have also been identified, no genotype-phenotype correlation could be detected. The relatively low number of our samples, the intracellular degradation and lack of secretion of mutant RS1 (loss of protein function) could explain why phenotypes are independent of the position or type of mutation [3,21,30]. The great phenotype variability in the case of the same mutation within one family, or between two eyes of the same person, presumably reflects the influence of other factors or genes [32].

Contrary to the data of the Retinoschisis database (dmd; exon 4: 42.7%; exon 5: 16.3%; exon 6: 26.9%), the frequency of various types of mutations per exons and introns were different in these Hungarian patients (exon 4: 11.1%; exon 5: 33.3%; exon 6: 44.4%; and intron 2: 11.1%). Two mutation hot spots were found: between c.418–422 in exon 5 and between c.574–579 in exon 6. Almost 70% of the mutations were within this part of RS1.

Although XLRS is one of the most frequent causes of juvenile macular degeneration, it is still underdiagnosed and often misdiagnosed. Serious visual deterioration can be prevented by regular follow-up and surgical intervention in case of sudden retinal complications. Molecular genetic examinations are of particular importance to identify carriers without any symptoms and timely genetic counseling may help in family planning decisions. Although there is no current therapy for the disease, the successful gene therapy performed in mice brings hope for the future [27,31,33].

Acknowledgments

We thank Andreas Gal, István Peták, and Hanno Bolz for providing the opportunity to sequence the RS1 gene, and we thank the patients and their families for their kind cooperation in the study.