Truncation of retinoschisin protein associated with a novel splice site mutation in the RS1 gene

Clinical studies

All studies were conducted in accordance with the tenets of the Declaration of Helsinki. Voluntary informed consent was obtained from all participating individuals (or from their parents) after providing a full explanation of the procedures. The molecular genetic examinations were approved by the Ethics Committee for Human Genome Research of Semmelweis University. We analyzed this family only, because no other family was found with this splice site mutation. There were no more relatives in the family. In spite of the low pedigree size, founding the mutation in the RS1 gene of the X-chromosome can prove the X-linked inheritance in this family. Our pedigree excludes the dominant inharitance pattern and on the basis of our knowledge there was no other inheritance pattern published if an RS1 mutation is found. The rare cases, like consangious marriage, X-chromosome inactivation, Turner syndrome can be excluded. So the inheritance can be X-linked recessive only.

A Hungarian family (Figure 2) with XLRS was referred to the study by the Department of Ophthalmology, Semmelweis University, Budapest, Hungary. Complete ophthalmic examinations including OCT, full-field and multifocal ERGs were performed on three males and two female obligate carriers. Two of the three males (III/2 and III/3) are fraternal twins and only one of them had visual difficulties.

Best corrected visual acuity (BCVA) was determined using a Snellen chart. Dilated indirect ophthalmoscopy of the posterior pole was performed with fundus photography.

OCT was performed in both eyes of each family member, using six 6-mm long OCT scans centered on the fovea by a Stratus OCT device (Carl Zeiss Medical Inc., Dublin, CA). The foveal thickness was calculated on the basis of OCT scans by the built-in software of the device and shows the retinal thickness between the internal limiting membrane and the pigmentepithel layer at the fovea. The FT map is a color-coded map of foveal thickness and uses to evaluate the size of macular edema, but also suitable to visualize macular cyst and check the central fixation. The total macular volume was also calculated by the built-in software and shows the volume between the above mentioned two layers at the macula and also capable to evaluate the size of edema or cyst.

Full-field ERGs (Roland Consult RetiPort, Wiesbaden, Germany) with white single flash stimuli were recorded according to International Society for Clinical Electrophysiology of Vision standards [23,24]. Pupils were dilated by instilling one drop of cyclopentolate 0.5% solution in each eye, and corneal jet electrodes were placed after instilling 0.4% oxybuprocaine chlorate as a topical corneal anesthesia. The amplitudes and implicit times of scotopic bright-flash (maximal response) a- and b-waves were measured, and the ratios of the b/a amplitudes were calculated.

The multifocal ERGs (Roland Consult RetiScan, Wiesbaden, Germany) were recorded with the same electrodes, using a 21” video stimulating display. Before insertion of the contact lens electrode, the subjects were optically corrected for the viewing distance if it was possible. The central 60° diameter part of the retina was stimulated in each eye separately by 61 hexagons under photopic conditions. Amplitudes and implicit times of the b-waves (P1) of first order kernels were measured.

All OCT and ERG results were compared with the values of 35 male and 35 female control subjects, with a mean age±SD: 22±10 years, respectively. To confirm the pathogenetic effects of the identified mutation 50 control male subjects (including the previous mentioned 35 control males), with a mean age±SD: 22.8±10.3 years were genetically examined. All controls were examined using the same clinical protocol. All of them had normal vision, had not any acquired or genetically inherited eye diseases and not the members of XLRS families. Inclusion criteria in the normative database were: preserved vision and absence of any acquired or genetically inherited eye diseases. A simple comparison was performed with the mean±2SD values of the normative database.

Molecular genetic studies

Peripheral blood (6 ml) was collected in EDTA-K3 tubes from all family members and 50 healthy male controls. Genomic DNA was isolated from leukocytes according to the manufacturer’s instructions, using the QIAamp DNA Mini Kit (Qiagen GmbH; Hilden, Germany). Isolated DNA samples were stored at a temperature -20°C. The 50 male controls include the 35 ones from the OCT and ERG controls. For mutation screening all exons and the flanking intronic regions of RS1 were amplified by the polymerase chain reaction (PCR) process using standard primers. Details of the primers, size of the amplified fragments, and annealing temperatures are given in Table 1.

PCR was performed using genomic DNA in a Termo Hybaid PxE thermal cycler (Thermo Hybaid, Franklin, MA). The reaction volume of 50 µl contained the following: 5X GoTaq Reaction Buffer (pH 8.5, containing 1.5 mM MgCl₂; Promega Corporation, Madison WI), 25 pmoles of each primer (Invitrogen Life Technologies, Glasgow, UK), 0.2 mM each dNTP (Promega), and 2.5 U GoTaq Polymerase (Promega). PCR cycles for the retinoschisin gene were as follows: pre-denaturation at 95 °C for 2 min, followed by 35 cycles of denaturation at 95 °C for 1 min, annealing at the appropriate temperature for 1 min, and 1 min elongation at 72 °C. The final extension was performed at 72 °C for 10 min.

The amplicons were analyzed on 1% agarose gel and stained with ethidium bromide. The resulting DNAs were purified using Roche High Pure PCR Purification Kit (Roche Diagnostics GmbH, Mannheim Germany) according to the manufacturer’s protocol.

PCR-amplified DNA of RS1 were sequenced by direct nucleotide sequencing using the Big Dye Terminator Cycle-Sequencing v3.1 Kit (Applied Biosystems, Foster City, CA) and run on an automated sequencer (ABI Prism® 310 Genetic Analyzer; Perkin Elmer™; Applied Biosystems).

The results were compared with the reference sequence from the X-linked retinoschisis sequence variation database and the sequence of the 50 controls.

The putative donor splice site was predicted with NNSPLICE, 0.9 version, at the website of the Berkeley Drosophila genome project.

Clinical studies

The clinical data of patients are reported in Table 2. A pedigree of the family members involved in the study is given in Figure 2.

Fundoscopy showed typical bilateral spokewheel-like foveal retinoschisis (Figure 3) characterized by radially arranged microcysts in the inner layers of the retina with the absence of foveal reflex in patients IV/1 and III/2. Neither peripheral retinoschisis nor any retinal tears or retinal detachments were evident. Golden-like fundus reflex, also called the Mizuo-Nakamura phenomenon, was found in both fundi of patient IV/1.

BCVA parameters of male patients, which varied from 0.25 to 0.42, were below the normal BCVA range of 0.98±0.06. At his first examination, patient IV/1 could hardly find the numbers on the Snellen chart. His BCVA was 0.25 in the right eye and 0.3 with eccentric fixation in the left eye. One year later, the fundoscopic findings had not changed, but his visual acuity had improved: BCVA in this patient’s right eye and left eye were both 0.42 without any correction and with eccentric fixation in the left eye.

OCT images of patient IV/1 showed huge low reflective cysts in the fovea surrounded perifoveally by many smaller ones in the inner nuclear layer (INL), and another deeper schisis in the photoreceptor layer (Figure 3). Next to the huge central cysts, some cavities were evident in the outer plexiform layer (OPL). The cysts were separated by multiple, highly reflective, vertical, thin-walled strands. In the right eye there was a small cyst in the ganglion cell layer also. Both OCT scans and the foveal thickness map proved the eccentric fixation in the left eye.

OCT images of patient III/2 (Figure 3) showed a flattened schisis with smaller cysts bordered by vertical bridging retinal elements in the INL in the right eye. In the left eye there were huge centrally localized cysts besides the smaller ones in the INL bordered by vertical bridging retinal strands with eccentric fixation in the left eye. Small cysts could also be found in the ganglion cell layer in both eyes. Foveal thickness and total macular volume parameters of the two patients were increased (FT: 328% of the normative values; TMV: 118%). OCT parameters of other family members were normal.

Full-field ERG recordings of patient IV/1 and III/2 (Table 2) showed markedly reduced standard combined b-wave amplitudes (57% of the normative values) with the relative preservation of standard combined a-wave amplitudes (91%) in both eyes. The standard combined b/a amplitude ratios were also decreased (63%), but the typical “negative-type” ERG was not observed.

Response densities (RDs) of mfERGs of patient IV/1 and III/2 (Figure 4A,B) were characterized by decreased P1 (b) amplitudes (IV/1>III/2 in the first rings) across the whole examined retinal area, but mainly in the three central rings as compared with controls. RDs of carriers were within the normal range. Implicit times of patients and carriers were delayed in all rings compared with controls.

Molecular genetic studies

A novel putative RS1 splice donor site mutation (c.78+1G>C) of intron 2 was identified in the two male patients (III/2, IV/1) and in two female carriers (III/1, II/1; Figure 2 and Figure 5; Table 1). There was no other disease-causing mutation in the exons or in other flanking introns of RS1. The asymptomatic fraternal twin (III/3) had no mutation in RS1. There were no any SNPs found in the disease family in or near the RS1 gene locus.

Cryptic splice site activation (intron retention)

Conserved GT signals are important for normal gene expression as they are involved in mRNA processing [33]. By mutating a conserved splice donor site from GT to CT, a novel cryptic site may become activated during RNA splicing, and thus the splice is probably shifted by ten nucleotides (donor site prediction score: 0.91) downstream to the next GT in the intronic sequence. Consequently, exon 2 becomes longer by ten nucleotides (CUAUGUAACA) inserted between positions 78 and 79, and the insertion encodes Leu+Cys+Asn amino acids and a residual adenine (A) nucleotide. The frameshift hypothetically results in a consecutive premature termination codon (PTC) in the beginning of exon 3 (after codon 30). The mRNA with its PTC is probably destroyed by a quality-control surveillance mechanism called nonsense-mediated mRNA decay (NMD). It protects the cell from mutations that could yield to truncated proteins by eliminating mRNA that contains a PTC [34].

Exon skipping

Nonsense, but not missense mutations, can alter splicing pathways, resulting in failure to include the affected exon in the final mRNA (a phenomenon known as exon skipping). The presumed 26 bp deletion at position 52 (exon 2 skipping) may cause frameshift and a PTC at the beginning of exon 3, as with the case of the 10 bp insertion described in the previous section. The mRNA with PTC is probably eliminated by NMD.

PTCs that occur because of the frameshift caused either by the 10 bp insertion or the 26 bp deletion may result in the total absence of RS1 protein by NMD and the resultant development of XLRS.

In a comparison of the clinical results of the two patients with our own retinoschisis database, we found the absence of the entire RS1 protein (completely lacking DD) did not seem to be accompanied by more severe symptoms than the less serious structural lesions of RS1 protein caused by other mutations. This is in agreement with previous findings for many X-linked recessive diseases: Considering that female carriers are unaffected, it seems that the absence of a functional RS1 protein in males, rather than the presence of a gain-of-function mutant protein, is responsible for retinoschisis in affected males [35]. Consequently, the main point in avoiding retinoschisis is the presence of the functional RS1 protein.

The Tennessee Mouse Genome Consortium generated the 44TNJ mutant mouse, the first murine model of X-linked retinoschisis in which the gene is expressed by using an ENU-based mutagenesis screen to produce recessive mutations [22]. The retinoschisis-1 homolog (Rs1h) cDNA, which was reverse transcribed from mRNA isolated from eyes, and the genomic DNA, obtained from tails, were sequenced directly. The gene Rs1h revealed a T->C splice donor site mutation at the second base of intron 2. This generated splice site mutation is very similar to our human mutation. The amplification of the mutant Rs1h cDNA by PCR revealed two new splice products in the retina in addition to the wild type. The first alternative transcript had a 10 bp insertion between positions 88 and 89, leading to elongation of exon 2 by the first 10 nucleotides of intron 2. The second alternative transcript had a 26 bp deletion at position 62 (entire exon 2). Both splice variants lead to frameshift and PTC at the beginning of exon 3. In both cases another open reading frame was found after only a few bases.

On the basis of the simultaneous presence of two different types of mRNA in 44TNJ mutant mouse and the very similar mutations, we hypothesized that there may also be two different types of mRNA simultaneously in our XLRS patients.

Since identification of RS1 splice variants in human is difficult because of the lack of tissue-specific cDNA sequence data, we feel further animal studies are needed to clarify whether the truncated Rs1h peptides or the peptides encoded by the new open reading frames are present in the retina or whether the mutant mRNAs are degraded by NMD. However, our findings are very similar to those found in the above mentioned animal model of XLRS and may help to understand the genetic background of the disease in our family with XLRS.