If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Identification of false-negative mutations missed by next-generation sequencing in retinitis pigmentosa patients: a complementary approach to clinical genetic diagnostic testing
Division of Ophthalmic Genetics, Laboratory for Stem Cell and Retinal Regeneration, The Eye Hospital of Wenzhou Medical University, Wenzhou, ChinaThe State Key Laboratory Cultivation Base and Key Laboratory of Vision Science, Ministry of Health, Wenzhou, China
Division of Ophthalmic Genetics, Laboratory for Stem Cell and Retinal Regeneration, The Eye Hospital of Wenzhou Medical University, Wenzhou, ChinaThe State Key Laboratory Cultivation Base and Key Laboratory of Vision Science, Ministry of Health, Wenzhou, China
Division of Ophthalmic Genetics, Laboratory for Stem Cell and Retinal Regeneration, The Eye Hospital of Wenzhou Medical University, Wenzhou, ChinaThe State Key Laboratory Cultivation Base and Key Laboratory of Vision Science, Ministry of Health, Wenzhou, China
Division of Ophthalmic Genetics, Laboratory for Stem Cell and Retinal Regeneration, The Eye Hospital of Wenzhou Medical University, Wenzhou, ChinaThe State Key Laboratory Cultivation Base and Key Laboratory of Vision Science, Ministry of Health, Wenzhou, China
Division of Ophthalmic Genetics, Laboratory for Stem Cell and Retinal Regeneration, The Eye Hospital of Wenzhou Medical University, Wenzhou, ChinaThe State Key Laboratory Cultivation Base and Key Laboratory of Vision Science, Ministry of Health, Wenzhou, China
Retinitis pigmentosa (RP) is a major cause of heritable human blindness with extreme genetic heterogeneity. A large number of causative genes have been defined by next-generation sequencing (NGS). However, due to technical limitations, determining the existence of uncovered or low-depth regions is a fundamental challenge in analyzing NGS data. Therefore, undetected mutations may exist in genomic regions less effectively covered by NGS.
Methods
To address this problem, we tested a complementary approach for identifying previously undetected mutations in NGS data sets. The strategy consisted of coverage-based analysis and additional target screening of low-depth regions. Fifty RP patients were analyzed, and none of the mutations found had previously been identified by NGS.
Results
Coverage-based analysis indicated that, because of a highly repetitive sequence, the RPGR open reading frame (ORF)15 was located in an uncovered or low-depth region. Through additional screening of ORF15, we identified pathogenic mutations in 14% (7/50) of patients, including four novel mutations first described herein.
Conclusion
In brief, we support the need for a complementary approach to identify mutations undetected by NGS, underscoring the power and significance of combining coverage-based analysis with additional target screening of low-depth regions in improving diagnosis of genetic diseases. In addition to its usefulness in RP, this approach is likely applicable to other Mendelian diseases.
Nonsyndromic RP is characterized by symptoms including night blindness and progressive vision field constriction, and can lead to eventual blindness caused by the death or dysfunction of photoreceptors.
RP is inherited as a Mendelian trait in autosomal dominant, autosomal recessive, and X-linked manners. The disorder exhibits enormous heterogeneity in genetic etiology, and thus far more than 3,000 distinct mutations in more than 100 disease-causing genes have been identified as causative of RP (RetNet; http://www.sph.uth.tmc.edu/retnet/home.htm).
Despite the large number of mutations and disease-related genes identified, however, a large proportion of patients with RP seem to lack genetic defects in all known genes.
Current advances in next-generation sequencing (NGS) have revolutionized Mendelian disorder analysis and greatly improved the effectiveness of molecular diagnosis of RP. The three major NGS-based methodologies are targeted exome sequencing (TES), whole-exome sequencing (WES), and whole-genome sequencing (WGS). TES is an accurate, rapid, and cost-effective method for genetic screening,
Identity-by-descent-guided mutation analysis and exome sequencing in consanguineous families reveals unusual clinical and molecular findings in retinal dystrophy.
WGS analyzes the entire genome and is especially useful for detecting mutations in noncoding regions, structural variations, and copy-number variations.
Both TES and WES are commonly used to detect pathogenic mutations in patients with RP; WGS is less widely used than TES or WES, primarily because of its high cost. Although numerous disease-causing genes and mutations in RP have recently been identified by TES and WES, a considerable proportion of RP cases seem to lack these pathogenic mutations. We postulate that the major underlying reason is failure to capture certain coding regions in predesigned targeted panels due to technical limitations, particularly in highly repetitive and GC-rich regions.
Open reading frame 15 (ORF15) is a large exon in the 3′ terminus of the RPGR gene in which nearly 60% of disease-causing RPGR mutations are located. This exon contains highly repetitive sequences and purine-rich regions. As a result, ORF15 is typically difficult to capture using conventional NGS; only a few mutations in the nonrepetitive regions are identified.
Detailed information about the poor performance of ORF15 on NGS platforms has not been ascertained yet.
In this study we tested a complementary approach including coverage-based analysis and additional target screening of low-depth regions. This strategy was validated in 50 unrelated patients with either simplex RP or XLRP, all of whom previously tested negative for pathogenic mutations by exome sequencing.
Huang XF, Huang F, Wu KC, et al. Genotype-phenotype correlation and mutation spectrum in a large cohort of inherited retinal dystrophy patients revealed by next-generation sequencing. Genet Med; e-pub ahead of print 6 November 2014.
We assessed and compared the sequencing quality of entire targeted regions and ORF15, which demonstrated that ORF15 lies in a region of very low sequencing depth. Subsequent screening of ORF15 directly by Sanger sequencing resulted in the identification of pathogenic mutations in 7 of 50 patients (14%). Overall, this study supports the hypothesis that a complementary approach can reveal false-negative mutations in RP patients that had previously been missed by NGS analysis, demonstrating the power and significance of combining coverage-based analysis and additional targeted screening of low-depth regions in clinical genetic diagnostic testing of Mendelian disorders such as RP.
Materials and Methods
Patient recruitment
We initially collected a cohort of 179 families (71% of which are simplex cases) and performed TES in the probands. Using a capture panel including 164 known retinal disease–causing genes, we previously identified pathogenic mutations in 99 patients.
Huang XF, Huang F, Wu KC, et al. Genotype-phenotype correlation and mutation spectrum in a large cohort of inherited retinal dystrophy patients revealed by next-generation sequencing. Genet Med; e-pub ahead of print 6 November 2014.
Among the remaining 80 unrelated patients with unknown genetic causes, 49 male patients with either simplex RP or XLRP were included in this study. Another patient (A-1; Table 1) was previously screened using WES, which failed to discover any disease-causing variants. In total, 50 unrelated patients, none of whom had mutations identified by TES (49/50) or WES (1/50), were tested. This study was approved by the ethics committee of the Eye Hospital of Wenzhou Medical University, and the study protocol was administered in accordance with the tenets of the Declaration of Helsinki.
Table 1Patients with RPGR open reading frame 15 (ORF15) mutations
TES was performed using a panel containing 164 known RP-related genes according to previously described protocols. WES was enriched using the Exome Enrichment V5 Kit (Agilent Technologies, Santa Clara, CA). Both targeted sequencing and WES were performed using the HiSeq 2000 platform (Illumina, San Diego, CA). Sequencing quality of the entire targeted regions and ORF15 was evaluated by coverage-based analysis.
ORF15 screening
The entire ORF15 sequence (1,706 bp) was amplified from patients as a 1,786-bp polymerase chain reaction product using the following primers: forward, 5′-GTATGATTTTAAA-TGTGATCGCTTGTCAGAG-3′; reverse, 5′-AAGGCATTTAAATTGTCTGACT-GGCCATAATC-3′. Each 50-μl polymerase chain reaction contained 100 ng DNA, 1 mmol/l forward and reverse primers, 25 µl Gflex Buffer (Takara, Dalian, China), and 0.5 µl Gflex DNA Polymerase. Polymerase chain reaction was amplified by an initial denaturation at 98 °C for 5 min, followed by 35 cycles of 98 °C for 10 s, 60 °C for 15 s, and 72 °C for 90 s, with a final extension at 72 °C for 5 min. Polymerase chain reaction products were cloned into the pEASY-Blunt Simple Cloning vector (TransGen Biotech, Beijing, China). The cloned ORF15 insert was then sequenced using primers described previously,
Among the 50 unrelated patients involved in this study, 6 (12%) were diagnosed with XLRP, whereas 44 (88%) were diagnosed with simplex RP. All probands complained of progressive night blindness. Subsequently, the best-corrected visual acuity became decreased and electroretinography testing showed extinguished response of photoreceptors. The fundus displayed a typical RP phenotype including bone-spicule hyperpigmentation and attenuated arteries (Supplementary Figure S1a online). Patients also presented with visual field defects (Supplementary Figure S1b online) and thinning of the retinae (Supplementary Figure S1c online). Taken together, the clinical evidence and family histories suggested diagnoses of simplex RP or XLRP.
TES of 164 known retinal disease–causing genes was performed on 49 patients (Supplementary Table S1 online), and 1 patient was previously screened using WES.
Huang XF, Huang F, Wu KC, et al. Genotype-phenotype correlation and mutation spectrum in a large cohort of inherited retinal dystrophy patients revealed by next-generation sequencing. Genet Med; e-pub ahead of print 6 November 2014.
No deleterious mutations were found in any of the 164 known genes, however, suggesting either that pathogenic mutations might have been missed by exome sequencing or that the clinical phenotype was caused by previously undescribed, novel RP genes.
Analysis of low-depth regions
To test the hypothesis that disease-causing mutations were present but had been missed by TES and WES, we prioritized analysis of the sequencing quality of ORF15. ORF15 is a highly repetitive region, mutations in which are responsible for a considerable proportion of RP cases. We selected the most repetitive region of ORF15 (Supplementary Figure S2 online) for quality analysis. Interestingly, the results indicated that the sequencing quality of ORF15 was significantly lower compared with that of the entire target region (Supplementary Figure S3 online). Statistical analysis included coverage in 1×, 4×, 10×, 20×, and the average depth, based on the raw data of all 50 samples. Therefore, it is possible that disease-causing mutations in ORF15 were missed during exome sequencing.
Discovery of missed mutations
Screening of 50 male patients with simplex RP or XLRP revealed ORF15 mutations in 7 patients (14%) from four XLRP pedigrees and three simplex families (Table 1, Supplementary Figure S2 online). Five unique ORF15 mutations were detected, including four novel mutations and one previously reported mutation.
Notably, the four mutations first described in this study are all frameshift mutations (c.3146delA, c.2485_2488delGAAG, c.2892_2893delGG, and c.2233_2236delAGAG). All of the causative mutations were segregated in the family. Detailed coverage distribution in the highly repetitive region indicated an ~1,000-bp segment missing from capture exome sequencing (Figure 1). Interestingly, seven mutations were located in a region that was completely uncovered because of its highly repetitive sequence. Another mutation was sequenced for seven reads because of its proximity to the 3′ terminus and to a less repetitive region. Taken together, of the 50 individuals diagnosed with simplex RP or XLRP, 7 (14%) harbored a disease-causing mutation that was not detected using traditional NGS approaches.
Figure 1Identification of mutations in low-depth regions. The x-axis indicates the genomic location, and the y-axis indicates the sequencing depth. The line indicates the sequencing depth of the target regions; arrows indicate the position of identified mutations.
particularly inherited degenerative retinopathies. As expected, given the extreme genetic heterogeneity of RP, numerous disease-related mutations have been discovered by TES or WES.
However, almost all current reports focus on single-nucleotide polymorphisms or insertions/deletions detected by NGS, thereby discarding information in uncovered or low-depth regions. Several previous studies of RP briefly reported this phenomenon without detailed investigation or deep discussion. For example, O’Sullivan et al.
mentioned that single-nucleotide polymorphisms identified by Sanger sequencing that were not identified using an NGS assay were in ORF15, but that study principally demonstrated the power and effectiveness of NGS.
In the current study, we investigated the feasibility of identifying genetic alterations in uncovered or low-depth regions in which mutations were not identified using NGS. We support the need for a complementary approach that combines coverage-based analysis with targeted gene screening in analysis of these regions in order to capture single-nucleotide polymorphisms, insertions, and deletions. Our results revealed pathogenic mutations in 14% (7/50) of male patients with simplex RP or XLRP, for which mutations in known RP- or XLRP-related genes were not detected by TES or WES. To the best of our knowledge, this study is the first to demonstrate that mutations missed by NGS could be discovered in the same samples by complementary sequencing and analytical approaches.
WES is highly effective at revealing rare exonic variations; however, the current detection rate for WES is far below the expected rate of genetic driver mutations in coding regions. WES analysis of one proband with XLRP initially suggested the absence of candidate variants, whereas only seven reads of additional target sequencing revealed a frameshift mutation in ORF15 located in a low-depth segment(Figure 1). TES is widely used to screen target regions in known genes and is both accurate and cost-effective. A capture panel including 164 known RP-related genes was used for TES in this study. Of the 49 patients in whom TES identified no mutations, 7 patients contained disease-causing mutations in ORF15, which were included in the designed panel. Interestingly, all seven mutations were located in uncovered regions, specifically in a fragment ~1,000 bp in size (Figure 1). In total, seven cases lacking candidate mutations based on TES or WES analysis were confirmed to harbor pathogenic mutations in the target region.
One of the primary reasons for failure of NGS to capture these variations is their localization within target regions containing a highly repetitive sequence. The NGS platform most commonly used worldwide is the HiSeq 2500, which produces maximum read lengths of 2 × 125 bp; however, this technical limitation can result in alignment errors. Notably, none of the current aligning tools can detect variations in large stretches of repetitive genomic sequence. For example, ORF15 contains a highly repetitive and purine-rich sequence more than 1,000 bp in length that comprises ~75% of the entire ORF (Supplementary Figure S2 online). The most advanced sequencing system, HiSeq X Ten, was recently shown to sequence more than 18,000 genomes per year, with read lengths of 2 × 150 bp. Although this new sequencing platform increases the feasibility of WGS analysis, it does not resolve the limitation. The seven cases in which existing mutations were not identified by TES or WES also could not be resolved by WGS. The study presented here illustrates this phenomenon in an RP cohort and provides an analysis-based solution to this limitation. Notably, in our next study we aimed to investigate this approach in patients with Usher syndrome (an autosomal recessive disorder), all of whom previously tested negative for pathogenic mutations by NGS. Five patients are being tested currently. Crucially, we identified a heterozygous variant in low-depth regions of the USH2A gene (Supplementary Table S2 online). Although the single heterozygous variant could not explain the phenotype, the result illustrated the opportunity to identify false-negative variants in other Mendelian diseases. Extended tests should be performed in a larger patient cohort with Usher syndrome or other monogenic disorders. Taken together, our results also suggest the likelihood that patients with other Mendelian disorder express mutations not previously identified by NGS analysis.
In summary, our study supports a complementary approach that is necessary to discover pathogenic mutations previously missed by NGS (TES, WES, or WGS) in RP patients. This approach consists of coverage-based analysis and additional target screening of low-depth regions. Using this comprehensive approach, we successfully identified disease-causing mutations in 14% (7/50) of RP patients who were undetected by TES or WES. Our results demonstrate that applying this method to an RP cohort, in which traditional NGS failed to reveal genetic predispositions, provides a valuable opportunity to identify novel disease-causing mutations both in RP and in other Mendelian diseases.
Disclosure
The authors declare no conflict of interest.
Acknowledgements
We appreciate all patients and family members for their participation in this study. This study was supported by the National Key Basic Research Program (2013CB967502 to Z.-B.J.), the Zhejiang Provincial Natural Science Foundation of China (LR13H120001 to Z.-B.J.) Major Health Science Project supported by NHFPC and Zhejiang Province (to Z.-B.J.), and the National Natural Science Foundation of China (81371059 to Z.-B.J.).
Identity-by-descent-guided mutation analysis and exome sequencing in consanguineous families reveals unusual clinical and molecular findings in retinal dystrophy.
Huang XF, Huang F, Wu KC, et al. Genotype-phenotype correlation and mutation spectrum in a large cohort of inherited retinal dystrophy patients revealed by next-generation sequencing. Genet Med; e-pub ahead of print 6 November 2014.
03192-0&id=gr1.jpg){kind=link}