Secondary Logo

Journal Logo

Research Articles

Whole exome sequencing identifies mutations of multiple genes in a Chinese cohort of 95 sporadic probands with presumptive retinitis pigmentosa

Huang, Lulina,b,c; Yang, Jialianga; Xu, Shiyaoa; Mao, Yaoa; Lee, Dean Yaoa; Yang, Jiyuna,b,c; Qu, Chaod; Li, Yange,∗; Yang, Zhenglina,b,c,∗

Author Information
doi: 10.1097/JBR.0000000000000021

Abstract

Introduction

Retinitis pigmentosa (RP) is the most common untreatable inherited eye disease. Thus far, 77 genes have been associated with non-syndromic RP, including 26 autosomal dominant genes, 55 autosomal recessive genes, and 3 X-linked genes (https://sph.uth.edu/retnet/sum-dis.htm). Mutations in 6 genes, including BEST1, NR2E3, NRL, RHO, RP1, and RPE65, can cause RP in both autosomal dominant and autosomal recessive models.[1,2] Because RP is highly heterogeneous, both in clinical presentation and genetic etiology,[1,3–5] an accurate method to screen for mutations would be highly beneficial with respect to improvement of clinical diagnosis and gene-specific treatment of patients with RP.[2] A particular challenge in the molecular diagnosis of RP is that the ethnicity-specific mutation databases, which contain both clinical and genetic information, are largely insufficient.

The traditional approach for molecular diagnosis of RP comprised an evaluation that consisted of targeted single gene or gene panel testing.[6–9] However, due to the genetic heterogeneities inherent in RP, molecular diagnosis is very limited when using these approaches. In recent years, targeted next-generation sequencing methods have been used for the molecular diagnosis of retinal diseases.[9–13] The primary disadvantages of this type of sequencing are as follows: 1) it is not convenient for identification of newly updated mutations in RP genes, and 2) it is not convenient for discovery of new disease genes, which may be present in approximately 40% of patients with RP.[1]

Whole exome sequencing (WES) represents a considerable advancement in the discovery of novel mutations, and is a powerful tool for analyses of the etiologies of retinal disorders.[14] This high-throughput technique can sequence all exons in the human genome. Therefore, it is capable of identifying single-gene mutations in Mendelian disorders. Furthermore, because of its efficiency and cost-effectiveness, WES is a promising approach for the discovery of novel mutations in retinal diseases, particularly for sporadic cases, because of its high-throughput nature and ability to control standard quality, as well as its increasing affordability.[15]

We previously described the molecular diagnosis by WES of retinal disease in probands, primarily within families. However, molecular diagnosis is particularly challenging for sporadic RP probands, who constitute a majority of RP cases.[16,17] In these patients, the mutation detection rate might approach 36%.[13] In this study of sporadic probands, we screened for mutations of RP by using WES to analyze 95 sporadic probands who were initially diagnosed with RP (ie, presumptive RP).

Participants and methods

Collection of RP samples

A total of 95 sporadic RP probands were collected from January 2008 to December 2012, among patients who presented to the ocular fundus diseases clinic at the Eye Center of Beijing Tongren Hospital, Capital Medical University, China and Department of Ophthalmology (83 probands), and Sichuan Province People's Hospital, China (12 probands) for etiological evaluation of retinal disorders. The diagnosis of RP was made in all patients and family members when possible on the basis of ophthalmic examination, including a visual test, slit-lamp examination, ophthalmoscopy, fundus photography, and International Society for Clinical Electrophysiology of Vision electroretinograms.[18] The families received formal genetic counseling, including a discussion of the risks and benefits of WES. Clinical data of each patient were included in this analysis. After diagnosis of the proband, consultation was then provided regarding their family disease history and their relatives’ visual function. In many cases, additional family members were invited to participate in the study at the time of the original clinic visit, or at a later visit. If there was an indication of disease among their relatives, further confirmatory ophthalmic examinations were completed in all family members when possible. All probands were initially clinically diagnosed with non-syndromic RP. This study was approved by the Institutional Review Boards of the Beijing Tongren Hospital and Sichuan Province People's Hospital, China. Venous blood (5 mL) was collected from the patients and used for DNA extraction with the GentraPuregene Blood DNA kit (China), in accordance with the instructions in the GentraPuregene Handbook. All patients received and signed informed consent forms.

WES data filtering and mutation interpretation

In this study, WES was performed on genomic DNA obtained from each proband; the workflow for this process is shown in Figure 1. First, DNA samples were prepared as an Illumina sequencing library; in the second step, the sequencing libraries were enriched for the desired target, using the Illumina Exome Enrichment protocol (TreSeqtechnology, http://www.illumina.com/science/education/truseq.html). The captured libraries were then sequenced using the Illumina HiSeq 2500 sequencer. The targeted information obtained through WES is shown in Figure 1A.

Figure 1
Figure 1:
Diagram of whole exome sequencing and data analysis procedure in this study. (A) Summary of Illumina TruSeq Exome sequencing chip used in this study. (B) Sequencing and data alignment pipeline. (C) Variant annotation reference genome and tools. (D) SNP databases for variant filtering. (E) Prioritization and determination steps for pathogenicity variants.

After the sequencing platform generated the sequencing images, clean FASTQ data were obtained by using CASAVA 1.82 (Illumina) (Fig. 1B). The clean reads were then mapped against UCSC hg19 (http://genome.ucsc.edu/) by using BWA (http://bio-bwa.sourceforge.net/). The SNPs and Indels were detected by SAMTOOLS (http://samtools.sourceforge.net/) and annotated by ANNOVAR (Fig. 1C).[19] The following 6 databases were used for variant filtering: dbSNP138, 1000 Genomes (http://browser.1000genomes.org/index.html), NHLBI Exome Sequencing Project (ESP) (http://evs.gs.washington.edu/EVS/, ESP6500), the ExAC Browser (Beta) (http://exac.broadinstitute.org/), gnomAD browser beta (http://gnomad.broadinstitute.org/) and 1877 in-house non-RP controls (Fig. 1D). The strategies for prioritization and determination of pathogenicity variants are shown in Figure 1E): (1) variants had total read depth > 5 X and SNP quality score > 50; (2) variants had minor allele frequency < 0.1% in all 6 variant databases for recessive genes, or minor allele frequency < 0.01% in all 6 variant databases for dominant genes; (3) variants were SNVs (stoploss, stopgain, nonsynonymous, nonframeshift_substitution, nonframeshift_insertion, nonframeshift_deletion, frameshift_insertion, frameshift_deletion) or splice site variants (splicing within 2-bp of a splicing junction); (4) variants were consistent with the known pattern of inheritance models (ie, homozygous/compound heterozygous for recessive genes, or heterozygous for dominant genes); (5) variants were predicted to be pathogenic by at least 2 predictors; and (6) variants were confirmed by sanger sequencing and segregation analysis. Predictions of potential functional consequences of variants were conducted using SIFT/PROVEAN (http://sift.jcvi.org/www/SIFT_chr_coords_submit.html), Mutation Taster (http://mutationtaster.org/) and Polymorphism Phenotyping v2 (PolyPhen-2, http://genetics.bwh.harvard.edu/pph2/).

Polymerase chain reaction and direct Sanger sequencing for variant confirmation

Sanger sequencing and segregation analysis were used for mutation confirmation. Primers were designed to perform polymerase chain reaction amplification on the 500–600-bp region flanking the mutation. To ensure high-quality Sanger sequencing, amplifications were designed with a boundary at least 150 bp from the mutation base. Amplifications then underwent Sanger sequencing on an ABI 3730 capillary sequencer. The Sanger sequencing results were analyzed with the Applied Biosystems Sequencer software. Variants were defined as “compound heterozygous” in a patient when the patient's father and mother each carried a heterozygous mutation, or when the direct relatives or siblings without RP only carried a heterozygous mutation.

Results

Variant spectrum of RP genes in WES

In this study, WES with the Illumina HiSeq 2500 platform provided a mean sequence coverage of more than 60×, with more than 91% of the target bases exhibiting at least 10× coverage. Approximately 70,000 variants were identified in each patient by WES. We first focused on 77 disease-causing genes for non-syndromic RP. In total, we obtained 32,106 variants of these 77 genes in 95 RP probands, of which 10,524 variants were in the coding region, and 5462 variants were predicted to change the amino acid sequence. No variants were found in EMC1 or KIZ. The spectrum of variants found among these 75 RP genes is shown in Figure 2.

Figure 2
Figure 2:
Spectrum of variants found among 75 retinitis pigmentosa genes.

Pathogenic mutations in known causative RP genes were identified in 29 probands

After determination of pathogenicity variation, we identified 44 mutations in 19 RP genes among 29 of the 95 sporadic probands, including 4 recurring mutations (Table 1). The clinical information for these probands is shown in Additional Table 1; http://links.lww.com/JR9/A2. The overall presumptive diagnostic rate for this Chinese cohort of sporadic RP probands was 30.5% (Fig. 3 and Table 1). Seventeen probands carried mutations within autosomal recessive RP genes that appeared to be homozygous or compound heterozygous. A total of 11 probands carried mutations within autosomal dominant RP genes; two of these probands, for whom parental data were not available, carried mutations that had previously been reported in HGMD. Additionally, two probands carried mutations within X-linked RP genes.

Table 1
Table 1:
List of 44 mutations identified in 95 sporadic probands
Figure 3
Figure 3:
Pie chart based on 95 probands analyzed for the presence or absence of mutations in the retinal genes (percentage).

Information regarding the mutations detected in RP genes is listed in Table 1; relevant clinical information is listed in Additional Table 1; http://links.lww.com/JR9/A2. In particular, the EYS gene demonstrated high mutation frequency: 6 probands carried compound heterozygous mutations in the EYS gene. In addition to these mutations in EYS, three probands carried mutations in RP1. Additionally, two probands carried mutations in ABCA4, PDE6A, and RHO genes. Probands 19093 and 19179 carried compound heterozygous mutations in the ABCA4 gene. These two probands were reassessed and determined to have Stargardt macular dystrophy, following molecular diagnosis. Mutations for other genes, including C2orf71, CA4, CNGA1, CRX, IMPG2, MERTK, RP1L1, SEMA4A, CYP4V2, RPE65, RPGR, SNRNP200, and PDE6B, were detected in only one proband. Patient 19095 carried pathogenic compound heterozygous mutations, c.1091-2A>G and p.Ser482Leu, in the CYP4V2 gene, and was re-diagnosed with Bietti crystalline dystrophy.

Among the remaining 66 probands, 10 (10.5% of the entire cohort) exhibited unique variants in known RP genes that could not be confirmed at the current stage (Table 2). This may be because the variants were neutral or tolerated, or because they were pathogenic compound heterozygous variants, but showed slightly higher frequencies (0.1% < minor allele frequency < 5%) in our control databases; they may also fail to conform to the inheritance models of the reported RP genes (Table 2).

Table 2
Table 2:
List of the uncertain RP variants

Pathogenic mutations in other retinal genes were identified in 23 probands

In 23 (24.2% of the entire cohort) of the remaining probands, we found mutations in 18 other retinal disease genes (Table 3), including mutations in AHI1, AIPL1, ALMS1, CACNA1F, CDHR1, CHM, COL11A1, CRB1, HMCN1, KCNJ13, KIF11, NPHP1, NPHP4, PDE6B, PITPNM3, RS1, TIMP3, and VCAN. Among these genes, the most enriched phenotypes were macular dystrophy and cone or cone-rod dystrophy.

Table 3
Table 3:
List of variants detected in other retinal genes

Discussion

In sporadic probands with heterogeneous diseases, it is particularly difficult to identify the disease genes, due to the lack of family genetic background and the effects of environmental factors.[20] Furthermore, the mutation spectrum may change significantly on the basis of geographical or ancestral backgrounds. For instance, a previous WES study involving 157 families revealed mutations in known RP genes in 50% of the cohort.[14] Another WES study involving 47 Chinese families with cone-rod dystrophy identified mutations in approximately 21% of the families.[21] In our sporadic RP cohort, 30.5% of the probands exhibited mutations in known RP genes. Forty-four mutations were identified in 19 RP genes, among which 40 mutations were novel. Eleven probands carried mutations in autosomal dominant genes (38.0%), 16 probands carried mutations in autosomal recessive genes (55.2%), and two probands carried mutations in X-linked genes (6.9%). Twenty-eight mutations in 18 other retinal disease genes were also identified in 23 probands. Overall, mutations were identified in 52 probands (54.7%). Notably, these reported mutations may support the provision of personalized medicine and development of future gene therapy for RP; this information may be critical in future studies of the mechanisms of RP.

Due to the rapid development of sequencing technologies and their various uses, it has become increasingly easy to identify new disease genes. In recent years, there has been increasing use of WES to perform mutation screening for molecular diagnosis of RP. Consequently, reverse clinical phenotyping may become increasingly relevant after mutation identification.[15] It will then be the responsibility of individual ophthalmologists to match each mutation carried by a patient with a particular phenotype, and then to conduct the necessary diagnosis and disease management.[22]

Despite these advantages of WES for use in evaluation of sporadic RP mutations, interpretation of the mutation data may be difficult due to the large number of variations and ethnic diversity involved in RP. Careful analysis of the potential impact of the detected variants, as well as clinical correlations of genotype and phenotype, may help to clarify the roles of these variants. In this study, we employed 6 databases to filter the data, among which the 1877 in-house controls, obtained using the same sequencing platform and population as that of the RP cohort, provided very powerful support for data filtering, interpretation, and lowering of the false genotype rate.[9]

This study has thoroughly tested the power of WES to perform molecular diagnostic analysis for patients with sporadic RP. We identified 44 mutations requiring genetic counseling among these sporadic RP probands, of which 40 mutations are novel. However, the detection rate of RP genes was 30.5%, which is lower than in other reported Chinese cohorts.[14] This is likely because of the misdiagnosis of RP at the time of initial diagnosis, based on clinical presentation similar to that of other retinal diseases, due to the highly heterogeneous phenotype (ie, other retinal diseases with mutations identified in this study also have some characteristics of RP) or doctors’ lack of experience. A total of 24 probands had mutations in other retinal genes, with the cone-rod dystrophy and macular dystrophy genes most prevalent. We invited these patients to return to the hospital and thoroughly reviewed their phenotypes after gene screening; then, we provided a corrected clinical diagnosis. This study provides additional experience and knowledge for retinal disease diagnosis in the future. We also found USH2A gene mutations in 6 probands (6.3% of the whole cohort); because we previously performed Sanger sequencing of whole exons of this gene in the present cohort combined with Usher syndrome patients, we did not list those mutations in this report. Taken together, our gene testing output was 61.0% for this cohort. Another reason for the low detection rate is the presence of high G-C content in a portion of exon 15 in RPGR, which occurs at a high frequency in patients with RP; notably, this could not be detected by WES.[9] Furthermore, some RP genes might harbor variants in deep intronic areas that would be missed by WES; these may not be detected in this study.

Acknowledgments

We would like to thank all the RP patients and their families for participating in this study.

Author contributions

ZY designed the study. YL, Jialiang Yang, Jiyun Yang, SX, CQ, YL and ZY recruited the participants. LH and YM performed the genotyping and the statistical analysis. LH wrote the initial draft, with edits from ZY and DYL. YL corrected the spelling and grammar. All authors critically revised and gave final approval of this manuscript.

Financial support

This study was supported by the National Key Scientific Research Program (No. 2016YFC0905200, to ZY), by the National Natural Science Foundation of China (No. 81170883, 81790643 and 81430008 (to ZY), 81300802 and 81670895 (to LH), 81271048 (to JY), 81570848 and 81100693 (to CQ)), by the Department of Science and Technology of Sichuan Province, China (No. 2014SZ0169, 2015SZ0052 (to ZY), 2015JQO057 (to LH), 2016HH0072 (to LH), 2017JQ0024 (to LH), 2015SZ0060 (to YL), 2013JY0195 (to LH), 2015SZ0060 (to YL), 2014FZ0124 (to DYL) and 2015JZ0004 (to CQ)) and by High-level Talents Program of UESTC Y03001023601021016 (to LH).

Institutional review board statement

The study was approved by the Institutional Review Boards of the Beijing Tongren Hospital and Sichuan Province People's Hospital, China, and performed in accordance with the principles of the Declaration of Helsinki.

Conflicts of interest

The authors declare that they have no conflicts of interest.

References

1. Hartong DT, Berson EL, Dryja TP. Retinitis pigmentosa. Lancet 2006; 368:1795–1809.
2. Ferrari S, Di Iorio E, Barbaro V, et al. Retinitis pigmentosa: genes and disease mechanisms. Curr Genomics 2011; 12:238–249.
3. Krawczyński MR, Pecold K. Genetic heterogeneity of retinitis pigmentosa. Klin Oczna 1994; 96:24–29.
4. Goldberg MF. Molecular heterogeneity in retinitis pigmentosa. More mutations. Ophthalmic Genet 1994; 15:47–50.
5. Parmeggiani F, Sato G, De Nadai K, et al. Clinical and rehabilitative management of retinitis pigmentosa: up-to-date. Curr Genomics 2011; 12:250–259.
6. Vervoort R, Wright AF. Mutations of RPGR in X-linked retinitis pigmentosa (RP3). Hum Mutat 2002; 19:486–500.
7. Martínez-Gimeno M, Gamundi MJ, Hernan I, et al. Mutations in the pre-mRNA splicing-factor genes PRPF3, PRPF8, and PRPF31 in Spanish families with autosomal dominant retinitis pigmentosa. Invest Ophthalmol Vis Sci 2003; 44:2171–2177.
8. Wada Y, Tamai M. Molecular genetic analysis for Japanese patients with autosomal dominant retinitis pigmentosa. Nippon Ganka Gakkai Zasshi 2003; 107:687–694.
9. Stone EM, Andorf JL, Whitmore SS, et al. Clinically focused molecular investigation of 1000 consecutive families with inherited retinal disease. Ophthalmology 2017; 124:1314–1331.
10. González-del Pozo M, Borrego S, Barragán I, et al. Mutation screening of multiple genes in Spanish patients with autosomal recessive retinitis pigmentosa by targeted resequencing. PLoS One 2011; 6:e27894.
11. Eisenberger T, Slim R, Mansour A, et al. Targeted next-generation sequencing identifies a homozygous nonsense mutation in ABHD12, the gene underlying PHARC, in a family clinically diagnosed with Usher syndrome type 3. Orphanet J Rare Dis 2012; 7:59.
12. Wang X, Wang H, Sun V, et al. Comprehensive molecular diagnosis of 179 Leber congenital amaurosis and juvenile retinitis pigmentosa patients by targeted next generation sequencing. J Med Genet 2013; 50:674–688.
13. Neveling K, Collin RW, Gilissen C, et al. Next-generation genetic testing for retinitis pigmentosa. Hum Mutat 2012; 33:963–972.
14. Xu Y, Guan L, Xiao X, et al. Mutation analysis in 129 genes associated with other forms of retinal dystrophy in 157 families with retinitis pigmentosa based on exome sequencing. Mol Vis 2015; 21:477–486.
15. Srivastava S, Cohen JS, Vernon H, et al. Clinical whole exome sequencing in child neurology practice. Ann Neurol 2014; 76:473–483.
16. van den Born LI, Bergen AA, Bleeker-Wagemakers EM. A retrospective study of registered retinitis pigmentosa patients in The Netherlands. Ophthalmic Paediatr Genet 1992; 13:227–236.
17. Neidhardt J, Glaus E, Lorenz B, et al. Identification of novel mutations in X-linked retinitis pigmentosa families and implications for diagnostic testing. Mol Vis 2008; 14:1081–1093.
18. Marmor MF, Fulton AB, Holder GE, et al. ISCEV Standard for full-field clinical electroretinography (2008 update). Doc Ophthalmol 2009; 118:69–77.
19. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 2010; 38:e164.
20. Shastry BS. Signal transduction in the retina and inherited retinopathies. Cell Mol Life Sci 1997; 53:419–429.
21. Huang L, Zhang Q, Li S, et al. Exome sequencing of 47 Chinese families with cone-rod dystrophy: mutations in 25 known causative genes. PLoS One 2013; 8:e65546.
22. Goldzweig CL, Rowe S, Wenger NS, et al. Preventing and managing visual disability in primary care: clinical applications. JAMA 2004; 291:1497–1502.
Keywords:

gene mutation; mutation; proband; retinitis pigmentosa; whole exome sequencing

Supplemental Digital Content

Copyright © 2018 The Chinese Medical Association, Published by Wolters Kluwer Health, Inc. under the CCBY-NC-ND license.