Whole exome sequencing identifies novel USH2A mutations and confirms Usher syndrome 2 diagnosis in Chinese retinitis pigmentosa patients

Retinitis pigmentosa (RP) is a common phenotype in multiple inherited retinal dystrophies (IRD). Disease gene identification can assist the clinical diagnosis of IRD patients for better clinical management, treatment and counseling. In this study, we aimed to delineate and characterize the disease-causing mutations in Chinese familial and sporadic patients with initial diagnosis of RP. Four unrelated Chinese families and 118 sporadic RP patients were recruited for whole exome sequencing analysis. A total of 5 reported and 3 novel USH2A mutations were identified in four Chinese probands. The probands and their family members showed typical RP features and mild to severe hearing impairment, confirming the diagnosis of Usher syndrome 2 (USH). Moreover, 11 sporadic RP patients were identified to carry the compound heterozygous mutations in the USH2A gene, confirming the diagnosis of USH2. The patients carried the truncating mutations had a younger age of first visit than the patients carried only the missense mutations (p = 0.017). In summary, this study revealed 8 novel USH2A variants in Chinese familial and sporadic RP patients, assuring that whole exome sequencing analysis is an adequate strategy to facilitate the clinical diagnosis of USH from the sporadic RP patients.

autosomal recessive RP, apart from the typical USH2 16,17 . Precise diagnosis can be difficult, especially with atypical fundus appearance 18 . Disease gene identification can greatly assist the clinical diagnosis of individual RP patient for better clinical management, treatment and counseling.
In this study, we applied the whole exome sequencing (WES) analysis to hunt for the disease-causing mutations in four unrelated Chinese families and 118 sporadic patients with initial diagnosis of RP. In addition, the clinical correlation of the identified mutations was also analyzed.

Results
Whole exome sequencing analysis in familial RP patients. Four unrelated Chinese families initially diagnosed with RP were recruited, following the recessive inheritance (Fig. 1). The proband of each recruited family was subjected to WES analysis. Each WES analysis resulted in a total of 70GB of sequence data, and 95.5% of sequence reads were originated from the exons, with a mean coverage of 100-fold. The total numbers of variants (SNPs and indels) of exons and splice sites identified were 23,774 for F1-II-6, 24,006 for F2-II-11, 23,499 for F3-II-6 and 23,911 for F4-II-15 ( Supplementary Fig. 1). After filtering the synonymous, intergenic, intronic and common variants, the candidate variants of known RP genes were reduced to 2 for F1-II-6, 2 for F2-II-11, 5 for F3-II-6 and 3 for F4-II-15 (Table 1).
For the family F2, two heterozygous variants in USH2A gene were identified in the two patients (F2-II-10 and the proband: F2-II-11), a non-synonymous variant c.5581 G > A (p.G1861S) in exon 28 and a splice site variant c.9570 + 1 G > A in intron 48 (Table 1). These two variants have been previously reported 20,21 . The c.5581 G > A (p.G1861S) and c.9570 + 1 G > A variants should be the causative mutations for the USH2 family F2.
For the family F4, two heterozygous variants in USH2A gene were identified in the proband (F4-II-15), a non-synonymous variant c.2802 T > G (p.C934W) in exon 13 and a splice site variant c.5858-1 G > A in intron 30 (Tables 1 and 2). Both variants have been previously reported 16,17 . The unaffected children (F4-III-34 and F4-III-35) only carried the splice site variant c.5858-1 G > A. Besides, non-synonymous variant in CNGB1 gene (c.367 C > T; p.R123W) was also identified in the proband (F4-II-15) and his unaffected son (F4-III-35), but the CNGB1 variant did not co-segregate with the phenotype in the pedigree. The c.2802 T > G (p.C934W) and c.5858-1 G > A variants should be the causative mutations for the USH2 family F4.
Clinical characteristics of the USH2 families. All the 4 recruited families initially diagnosed with RP followed the pattern of autosomal recessive inheritance (Fig. 1). For the four-generation family F1, the three affected patients (F1-II-2, F1-II-4 and the proband: F1-II-6) presented with the onset of night blindness at teenage and typical RP features. The fundus photographs showed the attenuation of retinal arteries, the bone-spicule pigmentation of the retina and the waxy optic disc (Fig. 3A). Patchy hypo-autofluorescence was observed in the mid-peripheral retina and hyper-autofluorescence in the macula by auto-fluorescence imaging (Fig. 3B). Moreover, optical coherence tomography (OCT) imaging displayed significant reduction in retinal thickness and extensive disruption in ellipsoid zone (Fig. 3C). The representative flattening in the rod and cone ERG response was detected (Fig. 3D). The pure-tone audiometry (PTA) results showed bilateral down-sloping severe sensorineural hearing loss (Fig. 3E), indicating the diagnosis of USH2. From the medical history, the affected members did not exhibit the delayed gait development, poor speech acquisition and balance disturbance or unstable walking. Their unaffected brother (F1-II-7) showed normal gross motor development, speech ability and eyesight; yet, he has bilateral mild sensorineural hearing loss and mild reduction in cone and rod response ( Table 2). The remaining unaffected subjects (F1-III-5, F1-III-8, F1-III-9, F1-III-11) did not show any symptoms or abnormalities in the clinical examinations.
The three probands (F2-II-11, F3-II-6 and F4-II-15) from the other three recruited families showed progressive night blindness and mild to severe hearing impairment, which indicated the diagnosis of USH2. The fundus abnormalities, the flattening in the rod and cone responses ERG and the reduced photoreceptor layer thickness in OCT examination supported the typical RP features in these patients.
USH2A mutation analysis in sporadic RP patients. To Figure 1. Pedigrees of the four recruited Chinese retinitis pigmentosa families. All the 4 recruited families initially diagnosed with RP followed the pattern of autosomal recessive inheritance. Squares and circles represent men and women respectively. The asterisk represents the participants' blood were collected and the arrow indicates proband. The slash represents deceased person. Black: the affected members; White: the unaffected members.
www.nature.com/scientificreports www.nature.com/scientificreports/ and c.15178 T > C (p.S5060P). All of the amino acids are conserved in USH2A protein across 8 different species (Fig. 4), except G3754 (partially conserved), indicating that the missense mutations could affect the USH2A protein function or structure.
Based on the changes in protein properties by the mutations, we divided the 14 USH2 and sporadic RP patients into the missense mutation only group and the truncating mutation group. The patients carrying 1 or 2 nonsense, indel or splice site mutations in the truncating mutation group have a younger age of first visit (39.75 ± 16.34 years) than the patients only carrying the missense mutations (59.17 ± 10.76 years, p = 0.017; Fig. 5), implying that the nonsense, indel and splice site mutations could lead to earlier onset of disease or more severe disease progression compared to the missense mutations.

Discussion
USH is usually discovered first by RP in the initial screening, and ophthalmic examination alone might not easily discover other invisible dysfunctions. Genetic diagnosis, in this scenario, can facilitate the precise clinical diagnosis of complicated diseases for better management and counseling. In this study, we identified 10 sporadic RP patients, out of 118 recruited subjects, carrying at least 2 compound heterozygous USH2A mutations (Table 3), confirming the diagnosis of USH2. This indicates that 8.47% of sporadic RP patients could belong to USH in Southern Chinese population.
Previous studies in Chinese populations have reported the identification of USH2A and ADGRV1 mutations for USH2 in Chinese populations 20,[22][23][24] . The USH2A gene, located on chromosome 1q41, is consisted of 72 exons, whereas the ADGRV1 gene, located on chromosome 5q14.3, is consisted of 90 exons. It is not efficient and time-consuming to screen the disease-causing mutations in these genes by traditional Sanger sequencing. Target exome sequencing has been applied to map the disease-causing gene for USH 20,[22][23][24] . Yet, the WES analysis, based on the general enrichment and capturing platform, can efficiently delineate the variants in all genes with a reasonable price and processing time. It is a favorable sequencing analysis for the disease-causing mutation screening in the inherited retinal dystrophies 20,25,26 .
USH is an autosomal recessive disorder characterized by retinitis pigmentosa (RP) and bilateral sensorineural hearing loss. In family F1, the USH2A mutation carrier (F1-II-7; c.187 C > T p.R63*) had best corrected visual acuity of 1.0, showed normal gross motor development as well as speech ability and did not complained of nyctalopia; yet, he has bilateral mild sensorineural hearing loss and slightly reduction in rod and cone ERG responses ( Fig. 3 and Table 2). The bilateral mild sensorineural hearing loss on PTA could probably be due to aging. However, whether the single nonsense mutation would lead to ERG response reduction or a milder symptoms of USH requires further investigations. Yet, we identified 21 sporadic RP patients carrying single USH2A variant, and whether carrying single USH2A variant contributes to the RP development is also unknown. www.nature.com/scientificreports www.nature.com/scientificreports/ Usherin, coded by the USH2A gene, is a basement membrane protein with 5202 amino acids. The first 31 amino acids are responsible for the signal peptide, and the rest of the protein is composed of the N-terminal laminin domain, the laminin EGF-like domains, the laminin G-like domains, the fibronectin type-III domains, a transmembrane domain and a cytoplasmic tail. In this study, we identified 30 mutations in the USH2A gene, of which p.C934W and c.9570 + 1 G > A in 5 patients (18.52%) and p.S5060P in 4 patients (14.81%), indicating that these mutations have higher frequencies in Chinese populations and a potential of founder effect. In the 14 USH2 and sporadic RP patients, we identified 15 missense mutations from by the WES analysis (Tables 1  and 3). One mutation (p.C934W) is located in the laminin EGF-like domain, 3 mutations (p.G1526R, p.G1723E and p.W1800R) in the laminin G-like domains and 7 mutations (p.E1985K, p.V2333A, p.Y3745H, p.G3754V, p.P4444T, p.P4660L and p.Y4673H) in the fibronectin type-III domains and 3 mutations (p.N1494I, p.G1861S and p.S5060P) in the linker regions, indicating that they could induce structural or functional changes to the usherin protein. Moreover, the p.C934W mutation also belong to the site for disulphide bond formation. The amino acids of these 14 missense mutations, except p.G3754V (partially conserved), are highly conserved across 8 different species (Fig. 4), reassuring the importance of these amino acid residues for usherin protein function.   Fig. 5), implying that the nonsense, indel and splice site mutations could lead to earlier onset of disease or more severe disease progression compared to the missense mutations. Similarly, it was reported that USH2 patients carrying the p.Glu767Serfs*21 USH2A mutation can have earlier diagnosis of disease, onset of night blindness, onset of visual field loss, diagnosis of cataract and onset of hearing loss than those carrying the p.Cys759Phe mutation 27 . Moreover, patients with two truncating USH2A mutations develop significantly more severe hearing impairment throughout life than patients with two non-truncating mutations 28 . Yet, how these mutations cause visual defects and hearing impairment requires further functional and mechanism investigations.
In summary, this study revealed 11 novel variants in USH2A gene from four unrelated Chinese USH2 families and 23 sporadic RP patients, expanding the mutation spectrums of USH2A gene. Ten sporadic RP patients were diagnosed as USH2 after WES. Genetic diagnosis is an adequate strategy to aid the clinical diagnosis for better clinical management and counseling. It should be recommended as a routine examination for inherited retinal dystrophies.

Study subjects and clinical examinations. Four unrelated Chinese families initially diagnosed with RP
were recruited from the Joint Shantou International Eye Center of Shantou University and the Chinese University of Hong Kong, including seven patients and nine unaffected family members (Fig. 1). The audiological criteria to consider a USH2 is mild to moderate sensorineural hearing loss 29 . Comprehensive ophthalmic examinations, including best-corrected visual acuity (BCVA), fundus photography, visual field tests, OCT, electroretinogram (ERG) and auto-fluorescence, were performed for all patients and some of the unaffected family members. The severity of hearing loss was evaluated by PTA and classified as mild (25-40 dB), moderate (41-70 dB), severe (71-90 dB), profound (>90 dB). Vestibular function was recorded from the detailed medical history, including motor development, speech acquisition, stability and balance.
For the validation analysis, 118 sporadic RP patients and 200 unrelated control subjects were also enrolled. The sporadic RP patients generally showed pubertal night blindness, restricted peripheral vision, progressive vision loss, overall bone-spicule pigmentation of the retina, attenuation of retinal vessels, and the flattening of the rod and cone ERG responses. The control subjects, aged 60 years or above, did not have any family history or symptom of inherited retinal diseases, hearing loss, or any other major eye diseases except mild senile cataracts and low myopia.
Peripheral blood was collected from the study subjects, and the genomic DNA was extracted by TIANGEN DNA blood Kit DP318 (TIANGEN, Beijing, China), and stored at −80 °C freezer before sequencing analysis.
This study protocol was approved by the Ethics Committee on Human Research at the Joint Shantou International Eye Center of Shantou University and the Chinese University of Hong Kong, which is in accordance with the tenets of the Declaration of Helsinki. Written informed consents were obtained from all study subjects or their guardians after explanation of the nature and possible consequences of the study.

Sanger sequencing validation.
The potential causative gene variants identified by WES in the Chinese RP families were validated by polymerase chain reaction (PCR) and Sanger sequencing with specific primers   www.nature.com/scientificreports www.nature.com/scientificreports/ (AIJI Biotechnology Company, Guangdong, China; Supplementary Table 1) in the ABI 3730XL Genetic Analyzer (Applied Biosystems, USA) to confirm the co-segregation pattern. The sequencing results were analyzed by NOVOSNP (http://www.molgen.vib-ua.be/bioinfo/novosnp/) with the alignment to the reference DNA sequence. In addition, the identified variants were also screened in 200 control subjects.