Comprehensive analysis of genetic and clinical characteristics of 30 patients with X‐linked juvenile retinoschisis in China

Abstract Purpose To provides the clinical and genetic characteristics of a series of Chinese patients with X‐linked juvenile retinoschisis (XLRS) through multimodal imaging and next‐generation sequencing. Methods Thirty patients (60 eyes) from 29 unrelated families of Chinese origin with XLRS were screened using multigene panel testing, and underwent a complete clinical evaluation. All variants identified in this study and reported in the Human Gene Mutation Database were analysed. Results Twenty‐five distinct variants in the retinoschisin gene were identified, of which eight were novel, and one was de novo. Missense mutations were the most prevalent type, and mutation hot spot was localized in the discoidin domain. The mean Snellen best‐corrected visual acuity was 0.28 ± 0.17. Of all eyes presenting with schisis, 92.86% had lamellar schisis and 62.5% had peripheral schisis. Schisis changes mostly involved inner and outer nuclear layers. X‐linked juvenile retinoschisis (XLRS) patients had a high incidence of complications, and peripheral schisis was a risk factor for it. No obvious genotype–phenotype association was observed. Conclusion This study provides comprehensive analyses of the genetic and clinical characteristics of XLRS in a cohort of Chinese patients. The fourth de novo mutation in RS1 was identified. And we show that XLRS has a wide spectrum of clinical characteristics; hence, molecular diagnosis is crucial for its diagnosis, differential diagnosis and genetic counselling. Peripheral schisis is a risk factor for the high incidence of complications, and no clear genotype–phenotype correlations were found.


Introduction
X-linked juvenile retinoschisis (XLRS; OMIM 312700) is a common earlyonset X-linked recessive disease, with a prevalence ranging between 1 in 5000 and 1 in 25 000 (Molday et al. 2012). It is characterized by bilateral macular involvement with onset in the first decade of life, sometimes as early as 3 months of age (Renner et al. 2008). The schisis primarily occurs in the retinal nerve fibre layer, including the inner nuclear layer (INL), outer nuclear layer (ONL) and ganglion cell layer (GCL) (Gehrig et al. 1999a;Gehrig et al. 1999b;Li et al. 2007). X-linked juvenile retinoschisis (XLRS) is categorized into four types: foveal schisis (FS) patients have FS only; foveo-lamellar schisis (FLS) patients have FS and lamellar schisis (LS), but not peripheral schisis (PS); complex type disease (FLPS) includes patients with foveal, lamellar and PS; and foveo-peripheral type patients (FPS) have foveal and PS, without LS (Prenner et al. 2006). The penetrance of XLRS is almost complete, but clinical expression is highly variable (Molday et al. 2012), resulting in an unpredictable disease progression and severity even within families. Visual acuity varies greatly among affected males, typically from 20/60 to 20/120, and remains relatively stable until the patient reaches their fifth decade (Pimenides et al. 2005). During the course of the disease, secondary complications including retinal detachment (RD), macular holes (MH), vitreous haemorrhage (VH) and neovascular glaucoma may occur, leading to a poor outcome (Ip et al. 1999;Pimenides et al. 2005). Swept-domain optical coherence tomography (SD-OCT) is the main diagnostic technique for this disease (Renner et al. 2008;Yu et al. 2010).
The retinoschisin gene (RS1; OMIM 300839), located on Xp22.13, is the only gene known to be associated with XLRS. It contains six exons and encodes a 224 amino acid protein.
Retinoschisin protein is organized into four distinct regions, including an Nterminal signal sequence characteristic of secreted proteins (exons 1 and 2; 23 amino acids), a retinoschisin domain (exon 3; 39 amino acids), a well-conserved discoidin domain (exons 4-6; 157 amino acids) implicated in cell adhesion, and a C-terminal segment (end of exon 6; five amino acids) (Gehrig et al. 1999a;Gehrig et al. 1999b). Retinoschisin protein is predominantly expressed and secreted from photoreceptor cells and most bipolar cells (Molday et al. 2001). It is an extracellular adhesion protein that plays a crucial role in maintaining the structural integrity of the retina (Molday et al. 2001).
A broad spectrum of clinical phenotypes has been reported for XLRS patients (Ip et al. 1999;Molday et al. 2012), making clinical diagnosis challenging, especially for young patients. Genetic testing is an effective method for providing timely, rapid and accurate diagnoses (Splinter et al. 2018), and multimodal imaging is an excellent technique for confirming phenotypes and helping clinicians understand the disease (Shinohara et al. 2018). The present study was aimed to report and analyse the genetic and clinical characteristics of XLRS in a cohort of Chinese patients, to advance our understanding of the disease. These data facilitate genetic counselling and the selection of patients who are eligible for gene augmentation.

Subjects and ethics statement
Participants were screened in the eye genetic disease clinic of the Eye and ENT Hospital of Fudan University (Shanghai, China) between January 2017 and March 2019. This study was approved by the Ethics Committee of the Eye and ENT Hospital of Fudan University, and adhered to the tenets of the Declaration of Helsinki. All patients or their guardians provided written informed consent.

Genetic analysis and variant assessment
Next-generation sequencing was performed in all patients and available family members. Genomic DNA was extracted from whole peripheral blood using the FlexiGene DNA Kit (Qiagen, Venlo, the Netherlands) according to the manufacturer's protocol. All participants underwent sequencing with a 762 gene panel (including genes involved in common inherited eye diseases; Table S1; BGI, Shenzhen, China) using BGISEQ-2000. We aligned sequence reads to the reference human genome (UCSC hg 38) with the Burrows-Wheeler Aligner version 0.7.10 . The bioinformatics pipeline was applied as previously reported Gao et al. 2019a;Gao et al. 2019b). Previously reported variants were determined using the Human Gene Mutation Database (HGMD, professional version 2019.1). Variants were classified according to the American College of Medical Genetics (Stenson et al. 2012). Sanger sequencing was performed to evaluate the presence of variants within other members in the family.

Cohort characteristics
Accurate genetic diagnosis was made after pedigree cosegregation and phenotypic confirmation. A total of 30 patients from 29 unrelated families of Chinese origin with a genetic diagnosis of XLRS were recruited. All patients were male, and both eyes were affected. The mean age was 15.73 AE 12.13 years (range, 4-48 years; median, 10 years; Table 1).
No patients had any pathogenic or likely pathogenic mutations associated with other inherited eye diseases.

Revision and refinement of initial clinical diagnosis and risk to family members
After genetic analysis, 30 patients received a definite diagnosis of XLRS. We successfully revised the clinical diagnosis of three patients (10%), who were initially diagnosed with macular hole, cystoid macular oedema and macular dystrophy. Of all the patients in this study, 76.67% (n = 23) had no known family history, and 23.33% (n = 7) had at least one other affected relative: four were presumed to be inherited in an autosomal recessive manner, while three were presumed to show X-linked recessive inheritance. However, the revised distribution of inheritance modes showed that all 30 patients had X-linked recessive inheritance, of whom 29 had inherited the mutation from their mothers and one carried a de novo mutation.

Ocular comorbidities
Complications seen in this cohort of patients are shown in Tables 1 and 2 and Fig. 1D. Vascular abnormalities (VAAs; Fig. 3A 3D) were found in four eyes (6.67%) of three patients (FLPS); one patient (age 25 years) also had PCAG, one (age 9 years) had RD, and one (age 9 years) had RD, pigment changes, exudation and VAAs. Retinal detachment (RD; Fig. 3E) occurred in two eyes (3.33%) from two patients, one aged 9 and one aged 25, both of whom were FLPS. Inner MH (IMH; Fig. 3F) was identified in two eyes (FLPS) from two patients (age 10 and 40 years). Vitreous haemorrhage (VH; Fig. 3G) was identified in two eyes (FLPS) from two patients, both of whom also had RD. Retinal white flecks (Fig. 3C) were found in one eye (FLPS) of a 9-year-old boy who also had VAAs, RD and VVs. In total, 94.29% (n = 33) of eyes with PS (n = 35) showed different degrees and types of complications, while the other two patients were only aged 4 and 6 years. However, of the 21 eyes without PS, only two eyes (9.52%) from two patients had complications (IMH). No significant difference was found between the ages of the patients and the degrees or types of complications (Table 1).

Genotype-phenotype correlations
To determine whether these clinical findings were associated with specific mutations, or types or locations of mutations, we further analysed the relationship between structural changes and ocular complications with specific mutations (Table 1). We found that the clinical findings did not correlate with specific RS1 variant or mutation types. Different individuals, even with the same mutation, were found to show completely different structural changes and complications (e.g. F2 and F8, F3 and F4, F12 and F19, and F16 and F27).

Discussion
We herein provide detailed and comprehensive reports of the genetic and clinical characteristics of a cohort of 30 unselected Chinese male patients with XLRS. We identified 25 distinct RS1 variants, of which eight were novel. The large proportion of novel mutations and variable types of causative variants suggest that the RS1 mutational spectrum in the Chinese population is distinct from that in other ethnic groups. Perhaps there may be many RS1 mutations remain to be identified and confirmed in different races, which will be vital for pathogenicity judgement and genetic counselling. A de novo mutation in RS1 was first identified by Gehrig et al. (1999a) and Gehrig et al. (1999b) (c.608C>T Pro203Leu), and then, another two de novo mutations were reported in 2009: c.238C>T p.Gln80Ter and c.2I4G>A p.Glu72Lys (Riveiro-Alvarez et al. 2009). It has been suggested that de novo mutations mainly occur in CG dinucleotides (Riveiro-Alvarez et al. 2009). Consistent with this, we identified the fourth de novo mutation in   (12) RS1, c.626G>T p.Arg209Leu, which is also located within CG dinucleotides, and representing a frequency of 3.33% (1/30). These results indicate that de novo mutations play a vital role in XLRS, and attention should be paid to them because they may have important implications in obtaining a correct genetic diagnosis and in enabling more accurate genetic counselling. By analysing the reported variants and those identified in the present study, we found that most mutations were localized in the discoidin domain and that missense mutations were the most common type (Hu et al. 2017;Chen et al. 2020). Of all reported variants, 11.76% are gross deletions or insertions, which may not be detectable by NGS. This has important guidance for the selection of sequencing methods used to diagnose XLRS, especially when small mutation analyses show negative results.
The mean BCVA of patients younger than 20 years was 0.33 AE 0.17, and it is better than that of patients older than 20 years. These data indicate that the optimal intervention window for a possible therapeutic approach is within the first 2 decades of life. We report PS in 58.33% (35/60) of eyes in our cohort, which is similar to previously reported (ranged from 43% to 79%) (Prenner et al. 2006;Wang et al. 2015;Chen et al. 2020). The rate was generally higher than the rate (48%) reported recently by Chen et al. (2020), who studied a large series of Chinese patients affected by XLRS. One possible reason is that ultra-widefield fundus imaging was used for retinal examination in this study, which enabled a more detailed identification of the peripheral retinal changes, and for identification of unnoticed peripheral retinoschisis (Leshno et al. 2019;Cicinelli et al. 2020). The most common type of XLRS in this study was FLPS, followed by FLS, then FS and FPS, which is similar to earlier findings (Ores et al. 2018). The mechanisms that leading to the occurrence of different types of schisis are not well understood, and we propose the possible reasons are the distribution and secretion of retinoschisin protein. We identified schisis changes mainly in the INL and ONL, followed by the GCL and OPL, which is in agreement with previous studies (Ores et al. 2018). These data indicate that retinoschisin is widely distributed in the retina. However, further analysis revealed no significant correlations between schisis type, layer involvement, patient age and mutations, suggesting the existence of more complex regulatory mechanisms responsible for schisis variations.
X-linked juvenile retinoschisis (XLRS) patients have a high incidence of various complications, with VAAs, VVs and pigment changes the most common in the present study, and the incidence is higher than reported in Caucasians (Pennesi et al. 2018). This may be attributed to the higher rate of PS in our study (58.33% versus 25%). Optical coherence tomography (OCT) angiography-based assessment previously found that microvascular changes were mainly located in the deep capillary plexus and that enlargement of the foveal avascular zone area correlated with BCVA decline (Romano et al. 2019). However, it remains unclear whether this is the cause (originating directly from a mutation) or the outcome (caused by other lesions such as cystic impairment). Vitreous veils (VVs) are common in XLRS and result from the occurrence of larger atrophic inner breaks in the thin inner schisis cavities (Campbell et al. 2015), which are often associated with VH, RD and VAAs (Ambler & Gutman 1991). Exudation and IRB are not rare in XLRS and may occur from the wrinkling or traction of macular detachment, abnormal vascular proliferation, or traction of VVs (Hu et al. 2017;Wood et al. 2019). Neovascular glaucoma has also previously been reported in XLRS (Ando et al. 2000), and with some studies attributing it to RD or extensive schisis (Manschot 1972), while others report that ischaemia in ocular tissues is the main cause (Ando et al. 2000). Although we did not observe neovascular glaucoma in our study, we detected two cases with POAG and one with PCAG; current information is insufficient to determine whether they are complication or simply coexist with XLRS. Macular atrophy (MA) was found in four eyes of three patients who were older than other patients; it was associated with a thinner ONL, which is consistent with previous results (Ores et al. 2018). Retinal detachment (RD) was only found in two eyes from two patients, one aged 9 and the other aged 25; the rate reported here is lower than the 5.5%-16% documented in previous reports (Fahim et al. 2017). Both IMH and VH were identified in two eyes from two patients in our study. The occurrence of VH at a young age is consistent with the report of its development before the age of 10 (Kellner et al. 1990). However, IMH can occur at any stage. Retinal white flecks in the posterior pole were seen in one eye in our study, and have also been reported in previous studies (Fahim et al. 2017). No second pathogenic variants associated with inherited eye diseases, including those in RDH5, RHO, RLBP1 and PRPH2, were found in the affected patient. This unusual fundus finding could be a rare characteristic of XLRS, though larger samples are needed for further verification. The prevailing viewpoint is that PS increases the likelihood of these complications occurring (Fahim et al. 2017;Wood et al. 2019). Our data showed that 94.29% of eyes with PS had different degrees and types of complications, while only 9.52% of eyes without PS had other abnormalities. However, future complications should not be ruled out in these patients with PS because they are still young (aged 4 and 6 years). Therefore, close follow-up is needed. It should also be noted that morning glory disc anomaly and POAG occurred as abnormalities in five eyes, although they may simply coexist with FLS.
Limitations of this study include the small size of our cohort, and the lack of a longitudinal follow-up to investigate morphofunctional correlations and changes. However, because XLRS is relatively rare, it is difficult to collect a large study group, and multimodal images are hard to acquire, especially from young patients.

Conclusions
This study provides comprehensive analyses of the genetic and clinical characteristics of XLRS in a cohort of Chinese patients. We show that the most prevalent mutation type is missense mutation, and the hot spot is localized in the discoidin domain. The fourth de novo mutation in RS1 was identified. X-linked juvenile retinoschisis (XLRS) has a wide spectrum of clinical characteristics; hence, molecular diagnosis is crucial for its diagnosis and genetic counselling. Peripheral schisis (PS) is a risk factor for the high incidence of complications, and closer follow-up may be beneficial. And no clear genotype-phenotype correlations were found. Our results not only have guiding value for selecting sequencing methods used for XLRS patients and genetic counselling, but also have an important clinical impact that should be considered in the perspective of gene therapy for XLRS patients.

Data Availability Statement
All data relevant to the study are included in the article or uploaded as supplementary information.

Supporting Information
Additional Supporting Information may be found in the online version of this article:  Table S1. Details of the 762 genes in the panel. Table S2. Analysis of the potential pathogenicity of the novel variants. Table S3. Schisis localization of studied eyes in this cohort of patients. Table S4. Electrophysiological characteristics of X-linked juvenile retinoschisis patients.