Genetic analysis and clinical features of X-linked retinoschisis in Chinese patients

Many mutations in the retinoschisis (RS1) gene have been identified, but there are limited clinical data relating to the different genotypes. This study investigated the genotype, clinical phenotype and therapies for X-linked juvenile retinoschisis (XLRS) patients in China to evaluate the effects of gene mutations and therapies on the prognosis of the disease. Thirty patients were recruited in the study. Genetic examination identified 8 novel RS1 gene mutations. Twenty-four patients were identified as missense mutation, which was the most common gene mutation in XLRS patients. Amino acids 102 and 209 were the most common mutation areas, accounting for a total 35.7% of all patients. Mutations affecting amino acid 102 were associated with poor results on the flash electroretinogram (ERG). Sixteen patients had various complications. Anti-vascular endothelial growth factor (VEGF) drugs were given to four patients with hemorrhage or other complications, and serious adverse events did not occur. Our outcome demonstrates that missense mutation was the leading cause of XLRS and more than half of the patients with this missense had various complications. Anti-VEGF drugs may be an effective and safe way to prevent deterioration of XLRS with certain complications. There is wide genotypic and phenotypic variability in Chinese patients with XLRS.

shows typical XLRS images of 1 patient.
Electrophysiology. Flash-ERG was performed on 28 eyes of 14 patients (Table 3). ERG demonstrated reduced b-wave amplitudes with relative preservation of a-wave amplitudes in 14 eyes (50.0%). Both a-and b-wave amplitudes were reduced in 12 eyes (42.9%); 2 eyes (7.1%) had relatively normal a-and b-wave amplitudes. The b/a ratio was reduced (< 1.2) in 24 eyes (85.7%).  detachment in 2 eyes, scleral buckling and cryotherapy for retinal detachment in 4 eyes, intravitreal injection of ganciclovir (IVG) for exudative retinal detachment in 2 eyes and intravitreal injection of bevacizumab (IVB) for vitreous hemorrhage in 1 eye. Vitrectomy, cataract surgery and laser therapy were performed in 1 eye each ( Table 2). The average VA after surgery was 0.87 ± 0.38. There was no significant difference in the VA change between preoperative and postoperative tests (P < 0.18) with a mean follow-up visit of 19.7 ± 13.3 months.

Discussion
Our results confirm that RS1 gene mutation is the primary factor in the development of XLRS. XLRS patients possess the characteristic of high incidence of various complications. Anti-vascular endothelial growth factor (anti-VEGF) drugs may be valid to control hemorrhage and other complications in XLRS. There is wide genotypic and phenotypic variability in Chinese patients with XLRS. More than one hundred RS1 mutations have been identified (http://www.dmd.nl/rs). In our study, missense mutation was the leading cause of XLRS, with a proportion of 85.7% in all the mutations for the patients, which is similar to previous reports 7 . Interestingly, a large proportion of mutations (28.6%) were novel, which suggests that the mutations may occur at different frequencies in different ethnic groups. The novel mutations included Arg102Cys, Gly91Cys, Arg102Pro, Tyr93Cys, Arg209Cys, splicing mutations (675 + 14C > T), p.95 deletion (c.282_284del), and a nonsense mutation (Ser73Ter). According to our previous study, severe XLRS phenotypes are associated with Arg102Gln, Asp145His, Arg209His, and Arg213Gln mutations 9 . In this investigation, we found that Arg102Gln and Arg209His together accounted for a considerable proportion (35.7%) of mutations. High incidence was associated with these specific mutations for XLRS patients in our study. However, patients with same genetic mutations did not always have the same degree of pathological changes in our study. We presumed that other genetic influences (i.e., environmental factors), might also contribute to disease severity 7 . Sporadic XLRS cases and novel mutations have been reported in China. Recently, two novel mutations were found in a cohort of six unrelated Chinese families 10 . In a similar study, six different RS1 mutations were identified in Chinese families with a clinical diagnosis of XLRS 11 . In that article, a new splice site mutation was described. Yangyan Xiao et al. investigated a Chinese family with XLRS and identified a novel mutation in the RS1 gene 12 . The mutation co-segregates with disease in the family and likely results in loss of RS1 protein production. The mutation expands the mutational spectrum of XLRS, and the phenotypes of the affected members of this family also provide evidence of potential genetic or environmental factors on XLRS prognosis. A study screened RS1 mutations from 16 patients belonging to 16 Chinese families 13 . In that study, sixteen different mutations were identified, nine of which had not been previously reported. The investigators concluded that no obvious genotype-phenotype association was observed in their Chinese patients. Our previous study identified severe XLRS phenotypes associated with some frameshift mutations, splice donor site mutations, and missense mutations 9 . We described the wide variability in the phenotype in Chinese patients with XLRS and expanded the information on clinical manifestations. Undoubtedly, the work described above will facilitate the diagnosis, appropriate early therapy, and genetic counseling and therapy based on the prognosis of XLRS.
Eyes with typical ERG manifestation (relatively normal a-wave and reduced b-wave amplitudes) accounted for 50.0% of all those that underwent ERG examinations in our study. The result was similar to those of previous studies 7, [14][15][16] . Eyes with reductions in both a-and b-wave amplitudes composed a large proportion (42.9%) of the sample. A decreased response may be related to serious retinoschisis and complications including vitreous hemorrhage and cataracts. It is noted that specific mutation may have severity ERG results. In our study, patients with amino acid 102 mutations (patients 1, 2, 4, 21) had lower a-and b-wave amplitudes compared with other mutations. The measurements were consistent with their clinical manifestations. ERG is an important approach contributing to differential diagnosis. One patient in our study was suspected to have indistinct retinal detachment initially. However, it proved to be merely retinoschisis, as ERG test showed that the suspected eye had relatively normal a-and b-waves compared with the other eye without detachment. In the era of gene therapy, ERG is becoming an important indicator of the effects of therapy for both humans and animal models. Our study provides a large amount of clinical data for the utility of ERG. Many animal experiments use ERG as an important reference standard of eye function; however, many patients in the clinic have normal ERG manifestations, as we have mentioned. For these patients, we need to seek a more comprehensive measurement instead to evaluate the effects of therapy.
In our study, 64.0% of the patients had various complications, and 1/3 of them required surgical intervention. Out of 16 cases with complications, 14 cases belonged to missense mutation. Patients with missense mutation tend to have severe complications and require surgery. The retinoschisis that occurs within the inner retina in XLRS leads to the decrease in vision in the children, but there used to be no effective therapy 7,17 . Scleral buckling for retinal detachment was performed in 4 eyes with rhegmatogenous retinal detachment as a traditional method 18 . In our patients, abnormal vasculature with hemorrhage was not rare; this condition can cause surface   wrinkling or traction macular detachment due to the abnormal vascular proliferation and leakage 19,20 . In this study, three patients were treated a total of 7 times with anti-VEGF drugs. Patient 2 (Fig. 3), who had typical retinopathy of prematurity (ROP) and cytomegalovirus (CMV) infection, was treated 5 times with binocular IVR, requiring more repeat dosing than simple ROP 21 . Furthermore, in a study by Spaide RF et al., intravitreal injection of bevacizumab in patients with proliferative diabetic retinopathy complicated by vitreous hemorrhage resulted in rapid resolution of vitreous hemorrhage 22 . In our study, three patients (patients 1, 9, 16) received IVR and/or IVB treatment for hemorrhages by an average of 1.8 times each eye. Remarkably, the hemorrhages reduced rapidly, and few complications were found after the treatment (Fig. 4). This group achieved relatively preserved VA at follow-up visits more than one year later. The results support the above conclusion and suggest that anti-VEGF drugs may be an effective way to control hemorrhage and other complications in XLRS. It is also interesting to notice that patient 25, who had severe exudative retinal detachment (CMV infection), underwent scleral buckling and cryotherapy in the left eye for proliferative retinopathy after binocular transscleral drainage of subretinal fluid initially. After that, the subretinal fluid relapsed in both eyes. The aqueous humor tested negative for CMV-IgG by polymerase chain reaction (PCR), but the subretinal fluid tested positive. The disease was eventually improved and stabilized by 4 treatments with binocular IVG. Systemic adverse reactions did not appear during 6 years of follow-up. This case indicates that CMV infection can affect the posterior segment of the eye while leaving the anterior segment normal and that aqueous humor testing is not a reliable basis for a final diagnosis. XLRS could be controlled by IVR or IVG without serious adverse events. Timely and effective measures for complications are required. Genetic testing, combining with optical coherence tomography (OCT) and ERG, is the most definitive evidence for diagnosis of XLRS. However, two patients in our study had none RS1gene mutations. To the best of our knowledge, it is not easy to distinguish atypical XLRS from familial exudative vitreoretinopathy (FEVR), rod-cone degeneration disease and other diseases. In a clinical study by Dhananjay Shukla, severe dragging of the macula was the most striking clinical finding in five XLRS patients 23 . Extensive mottling and atrophy of the retinal pigment epithelium at the posterior pole was presented in three XLRS patients. Differential diagnosis for such disease includes both infectious and noninfectious conditions like ROP, FEVR, toxocariasis, toxoplasmosis, tuberculosis, Coats disease, etc. Therefore, using varies kinds of tests skillfully and being familiar with the clinical features of various diseases for clinician will help to avoid misdirected investigations and interventions for mimicking diseases. Genead MA found that patients with XLRS had the potential to experience a beneficial effect from sustained treatment with 2% dorzolamide, as significant improvement in the foveal zone thickness was observed 24 . The result suggests that a mechanism independent of retinoschisin itself is responsible for the improvement in the degree of cystic spaces 25 . However, the case series by Khandhadia S found that this did not necessarily correlate with improvement in VA and that genotypic information did not predict the response to this treatment 26 . Nevertheless, it might have a role to play in combination with gene or surgical therapies by quickly reducing cavity size, which would aid the healing process by bringing tissue together and by setting the stage for inducing RS protein expression by gene transfer to stabilize the retina over a longer term 27 .
Several mouse models of XLRS have been created, mainly by genetic engineering, and remarkable progress has been achieved recently. Yong Zeng et al. created an XLRS mouse model (RS1-KO) by substituting a neomycin resistance cassette for exon 1 and 1.6 kb of intron 1 of Rs1h, the murine orthologue of the human RS-1 gene 28 . The mouse mimics structural features of human XLRS with dissection through, and disorganization of, multiple retinal layers and implicates a synaptic transmission deficit. Replacement therapy by supplementing normal Rs1h protein induced retinoschisin expression and restored the normal ERG configuration. Another study demonstrated that restoration of RS1 via retina-specific delivery of adeno-associated virus type 8-RS1 (AAV8-RS1) vector rescued molecular pathology and restored function in adult Rs1h-KO animals 29 . Abu E. Bashar delivered syngeneic adipose-derived mesenchymal stem cells (MSCs) that were genetically modified to secrete human RS1 to the retina of the Rs1h knockout mouse model by intravitreal injection 30 . RS1-expressing MSCs were found mainly in the inner retinal layers. Four months later, the schisis cavity areas were reduced by 78%, more photoreceptor nuclei were present, and the ERG b-wave was significantly improved after multiple injections. Studies in rabbits and Rs1h-KO mice of a recombinant AAV8 vector for X-linked retinoschisis gene therapy showed retinal detachment without other ocular or systemic adverse events 31 . Although biodistribution analysis detected vector genome in extraocular tissues of Rs1h-KO mice, no evidence of damage was found in organs or tissues. On this basis, a human trial was carried out recently with the AAV8-RS1 vector to probe synaptic plasticity in XLRS patients 29 .
Safe and effective gene therapy in an animal model eventually provides a basis for gene-directed therapy. However, it is unknown how the mutant protein would interact with normal retinoschisin protein or at what level it would be expressed inside the cell. Studies are under way to gain insight into this. As we know, gene therapy could not deal with the patients with serious complications mentioned above. Restore the relative normal structure before the gene therapy may be a preferred option to cure this disease. Our study provided abundant clinical and genetic data to the comprehensive treatment with gene therapy. However, we still need to address the possible challenges of gene therapy, including the timing of gene intervention, immune response control, complication management, interaction with other treatments and ocular and systemic adverse events in the long run.
In conclusion, RS1 gene mutation is a major cause of XLRS. OCT and ERG tests typically feature in the diagnostic process. Molecular genetic examination confers advantages in diagnosis and differential diagnosis. XLRS complications are variable in clinical practice. Timely and targeted measures should be considered when managing the complications to prevent progressive visual deterioration and improve visual prognosis.

Methods
The research protocol was approved by the ethics review board of the Peking University School of Medicine. The study procedures were carried out in accordance with institutional guidelines and the Declaration of Helsinki. Informed consent was obtained from all patients after a full explanation of the procedures.
Thirty suspected male patients were enrolled in this study from 2004 to 2016. Patients underwent a comprehensive ophthalmological examination including slit-lamp biomicroscopy, funduscopic examinations, measurements of best corrected visual acuity (VA) with a logarithmic VA chart, OCT examinations (Spectralis; Heidelberg Engineering, Heidelberg, Germany) and ERG tests (RETIport, ROLAND CONSULT, Germany). Clinical therapies were recorded during the follow-up visit. Molecular genetic examinations of the RS1 gene were performed in the 30 males, 51 parents, and 100 control subjects (normal vision and without any eye diseases). Genomic DNA was extracted using an Agilent SureSelect Target Enrichment System Kit (Agilent, USA). Polymerase chain reaction (PCR) was performed to amplify six exons of the RS1 gene. Samples were sequenced directly by loading the sequencing reaction product onto an NEXTSEQ500 (Illumina, USA). T test was carried out on the process of clinical data comparison (SPSS 16.0). P value of 0.05 was considered to be statistically significant.