Deep intronic GPR143 mutation in a Japanese family with ocular albinism

Deep intronic mutations are often ignored as possible causes of human disease. Using whole-exome sequencing, we analysed genomic DNAs of a Japanese family with two male siblings affected by ocular albinism and congenital nystagmus. Although mutations or copy number alterations of coding regions were not identified in candidate genes, the novel intronic mutation c.659-131 T > G within GPR143 intron 5 was identified as hemizygous in affected siblings and as heterozygous in the unaffected mother. This mutation was predicted to create a cryptic splice donor site within intron 5 and activate a cryptic acceptor site at 41nt upstream, causing the insertion into the coding sequence of an out-of-frame 41-bp pseudoexon with a premature stop codon in the aberrant transcript, which was confirmed by minigene experiments. This result expands the mutational spectrum of GPR143 and suggests the utility of next-generation sequencing integrated with in silico and experimental analyses for improving the molecular diagnosis of this disease.

There is no obvious explanation for this low rate of detecting mutations. The possibility that mutated genes other than GPR143 are responsible for OA1 seems unlikely, because linkage analysis of families without GPR143 mutations point to the OA1 locus (Xp22.3) 5,10-14 . Another possibility is thatGPR143 mutations occur outside the coding region or include larger insertions or deletions. Most intronic mutations occur within or very close to conserved splice sites, and are readily detected using Sanger sequencing of polymerase chain reaction (PCR) amplicons generated from exons. We are aware of only one deep intronic mutation (intron 7) of GPR143 15 . This mutation creates an aberrantly spliced mRNA by generating a de novo splicing enhancer motif that is similar to mutations detected by analysis of mRNAs of other genes [16][17][18] . However, genes expressed by specific cell lineages that are difficult to isolate, including pigment cells that express GPR143 19 , are difficult to identify through analysis of mRNAs.
Here we applied whole-exome sequencing (WES) to detect mutations and copy number variations to study one Japanese family with two male siblings affected by nystagmus and ocular albinism. Although mutations in coding regions nor copy number alterations of GPR143 or other candidate genes were not detected, detailed analysis of intronic sequences flanking target exons presented here show that ocular albinism in this family may be caused by the deep intronic mutation c.659-131 T > G in GPR143 (NM_000273.2) intron 5 that creates a cryptic donor splice site. Moreover, this mutation causes a 41-bp insertion in the transcript by creating a cryptic exon within intron 5.

Results
Clinical description. The proband (II:1) was a 7-year-old boy who visited our hospital because of congenital nystagmus since birth, with poor visual activity and tilting of the head to the right (Table 1). Therefore, we recommended that he wear glasses to correct astigmatism. He is the first child of non-consanguineous parents. Ophthalmological examination revealed hypopigmentation of the fundus and macular hypoplasia in both eyes, but his irises were normal ( Fig. 1, Table 1). His skin and hair, as well as his physical growth and development, were normal. II:2 is the younger brother (23 months) of II:1. He exhibited congenital nystagmus since birth, but his visual acuity was not affected. Ophthalmological examination revealed hypopigmentation of the fundus and macular hypoplasia in both eyes ( Fig. 1, Table 1). His iris, skin and hair were normal as were his physical growth and development.
Candidate mutations within the coding sequences and exon-intron boundaries corresponding to an autosomal recessive model, including compound heterozygosity and X-linked recessive models, are shown in Table S1. Although we evaluated the pathogenic significance of each variant, we did not detect mutations in genes possibly responsible for ocular albinism. Furthermore, we searched for mutations within flanking intronic sequences of targeted coding exons of candidate genes, which were obtained using WES, to identify pathogenic variants within non-coding sequences including introns, UTRs and regulatory regions. Through this analysis, we identified one hemizygous mutation NM_000273.2:c.659-131 T > G (NC_000023.10:g.9711844 A > C) within GPR143 intron 5 in both affected individuals (Fig. 2a). This mutation was confirmed using Sanger sequencing, and the heterozygous GPR143 c.659-131 T > G mutation was detected in their mother's DNA (Fig. 2b). This mutation is not present in dbSNP138, 1000 Genomes Project, ESP6500, and HGVD. We searched for potential splice sites using three different algorithms, such as MaxEntScan, NNSPLICE 0.9 and Human Splice Finder Version 2.4.1, and found that the mutation created a potential cryptic splice donor site and an out-of-frame 41-bp pseudoexon that generates a frameshift (Fig. 2c). Therefore, we speculated that the GPR143 c.659-131 T > G mutation affects the splicing of the GPR143 primary transcript and may cause ocular albinism. The mutation is not recorded in databases of genetic mutations responsible for human inherited diseases, such as the HGMD, ClinVar (http://www.ncbi.nlm.nih.gov/clinvar/), and the Locus Specific Mutation Databases of the Human Genome Variation Society (http://www.hgvs.org/locus-specific-mutation-databases/).

Minigene assay.
Because GPR143 is exclusively expressed by pigment cells 19 , e.g. melanocytes from skin that were not available from each affected patient, we performed a minigene assay to analyse the effect of the c.659-131 T > G mutation on the splicing of GPR143 transcript (Fig. 3a) 20 .    the mutant protein, preceded by a mutant sequence of 10 amino acid residues (p.V221Gfs10*, Fig. 3c). These findings strongly support the conclusion that the intronic mutation identified in GPR143 has pathogenic effects.

Discussion
Approximately 10% of mutations in disease-associated genes listed in the HGMD affect splicing, and most occur within or close to conserved consensus splice sites and are readily detected by Sanger sequencing of PCR amplicons targeting coding exons. However, deep intronic mutations associated with aberrant splicing of target genes are detected with increasing frequency in the genomes of patients with diseases, such as HPRT deficiency, Usher syndrome type 2, cystic fibrosis and megaloencephalic leukoencephalopathy with subcortical cysts type 1 [21][22][23][24] . Evaluating such splicing defects of GPR143 may be important for patients with OA1 because GPR143 mutations are not present in all patients with OA1. Mutations in GPR143 that affect splicing are present mainly in authentic splice sites (HGMD and ClinVar). Only one deep intronic mutation, which generates a consensus binding motif for the splicing factor enhancer and activates a cryptic donor-splice site that inserts an intronic fragment as a cryptic exon in the transcript, was identified through analysis of transcripts expressed by skin melanocytes obtained from patients with OA1 15 . In the present study, we report a novel intronic mutation that creates a putative cryptic splice donor site in GPR143. Although it was not possible to detect aberrant transcripts of GPR143 because expression of this gene is limited to pigmented cells 19 not available from the affected patients, in silico analysis and the minigene experiment demonstrated that the c.659-131 T > G mutation creates a cryptic splice that generates an aberrant transcript containing a 41-bp insertion with a premature stop codon. Ocular albinism encompasses a group of genetic disorders, in which reduced eye pigmentation is associated with decreased visual acuity, nystagmus, strabismus and photophobia. Although OA1 or Nettleship-Falls type ocular albinism, is the most common form of ocular albinism, these symptoms are not exclusive to this disease. Moreover, the diagnosis of OA1 in black or East Asian patients in contrast to Caucasian patients is difficult; this is because former patients with OA1 often have brown irises with little or no translucency and varying degrees of fundus hypopigmentation (non-albinotic fundus) 25,26 .
Here the two male siblings were affected by ocular albinism and congenital nystagmus, suggesting that X-linked (XLOA) or autosomal recessive (AROA) diseases caused by mutations should be considered. Furthermore, exome sequence analysis detects genomic copy number abnormalities that cause loss-of-function, although only enriched regions of interest can be evaluated 27 . Therefore, we applied WES to analyse this family because the mutations in GPR143 associated with OA1 are highly heterogeneous. Because a significant portion of the reads acquired using exome sequencing originate outside of the designed target regions 28 , exome data provide information useful for identifying mutations in the sequences around the target exons. For example, 27% and 24% of the single nucleotide polymorphisms (SNPs) identified using exome sequencing occur in the flanking ( ≤ 200 bp) sequences or in sequences >200 bp away from the target regions 29 , respectively.
The unusual feature of the GPR143 mutation described here is that it generates a new donor splice site within intron 5 and activates a cryptic acceptor splice site 41-bp upstream. Although the sequence c.659-131 T is evolutionarily conserved in mammals, such as Pan troglodytes, Macaca mulatta, Mus musculus, Rattus norvegicus, Canis familiaris, Bos taurus and Equus caballus, compared with Homo sapiens, the sequence around the cryptic acceptor splice site 41-bp upstream is conserved only in P. troglodytes and M. mulatta (Fig. 4). Therefore, it is unlikely that the mutation occurred around an evolutionary remnant or a 'leftover' of an exon within intron 5, although two of three splicing prediction algorithms yielded a relatively high score for the cryptic acceptor splice site in the human genome (Fig. 2c). Because the sequence around the mutation is not so highly conserved, it is also unlikely that the mutation occurred within consensus sequences of intronic cis-regulatory elements to direct transcription levels, such as enhancers, silencers, and insulators.
One limitation of this study is that the endogenous aberrant transcript containing the 41-bp insertion of the cryptic exon was not investigated in RNA samples from the two affected patients. Analyses of GPR143 and its mRNA in patients were hampered by the restricted expression of GPR143 in pigmented cells of the retinal pigmented epithelium and skin 19 , and such samples were unavailable from affected siblings and their mother. Therefore, we were unable to establish whether the phenotypes of our patients were most likely explained by the detected GPR143 intronic mutation. Vetrini et al. 15 analysed a pure culture of melanocytes derived from skin biopsies of patients with OA1 and detected an endogenous, aberrantly spliced mRNA with an anomalous cryptic exon, which was expressed from an intronic sequence, because a point mutation generated a consensus binding motif for the splicing factor enhancer. In their analysis, the quantities of the mutant mRNA and protein were very low, possibly due to nonsense-mediated mRNA decay caused by the premature translation-termination codon in the mutant mRNA 30,31 . Therefore, it may be difficult to detect expression of endogenous aberrant transcripts to diagnose OA1, even if samples suitable for expression analysis of GPR143 are available. There is the possibility that a significant number of non-exonic GPR143 mutations contributes to OA1 1,5-9 . Therefore, integrated analyses using bioinformatics tools and experiments using cells transfected with genomic DNA fragments may determine the functional significance of mutations and facilitate early diagnosis, appropriate early therapy and genetic counselling for this OA1.
In summary, we used data obtained by WES to identify a novel deep intronic mutation c.659-131 T > G in GPR143 in a Japanese family with male siblings affected by ocular albinism and congenital nystagmus. This mutation was predicted to create a cryptic splice donor site and an out-of-frame 41-bp pseudoexon, which creates a frameshift and a premature stop codon in the transcript, within intron 5. This in silico prediction was confirmed by the minigene experiments. To the best of our knowledge, this is the first report of an intronic mutation that creates a cryptic splice-donor site in GPR143 of patients with OA1, expanding the spectrum of GPR143 mutations. This study further illustrates the value of conducting WES analysis integrated with in silico and experimental analyses for molecular diagnosis of this disease.

Methods
Subjects. The Japanese family with two male siblings affected by congenital nystagmus (Fig. 1, Table 1) was identified and enrolled at the Eye Department of Osaka Medical Center and Research Institute for Maternal and Child Health. Clinical and ophthalmological examinations were performed in the hospital. Peripheral blood was collected from the two siblings and their parents. This study complied with the tenets of the Declaration of Helsinki. The Ethics Committee of Osaka Medical Center and Research Institute for Maternal and Child Health (Osaka, Japan) and Tokushima University (Tokushima, Japan) approved this study and informed consent was obtained from the parents before testing commenced.
WES and Sanger Sequencing. Genomic DNA was extracted from peripheral blood leukocytes using standard methods, enriched using the SureSelect Human All Exon Kit V5 (Agilent Technologies, Santa Clara, CA) and sequenced using a HiSeq 1000 platform (Illumina, San Diego, CA) with 101-bp paired-end reads. Image analysis and base calling were performed using sequence-control software real-time analysis (Illumina) and CASAVA software v1.8 (Illumina). Reads were aligned to the human genome sequence assembly hg19 (GRCh37) using the Burrows-Wheeler Alignment tool (BWA) 0.7.10. (http://bio-bwa.sourceforge.net/). The alignments were converted from a sequence alignment map (SAM) format to sorted and indexed binary alignment map (BAM) files (SAMtools version 0.1.19). Duplicate reads were removed using Picard 1.118 (http://broadinstitute.github.io/picard/). Local realignments around insertions or deletions (indels) and base quality score recalibration were performed using the Genome Analysis Toolkit (GATK) 3.2− 2 (https://www.broadinstitute.org/gatk/). Single nucleotide variants and small indels were identified using the GATK UnifiedGenotyper and filtered according to the Broad Institute's best-practice guidelines (https://www.broadinstitute.org/gatk/guide/best-practices). Variants that passed the filters were annotated using ANNOVAR (http://www.openbioinformatics.org/ annovar/). After merging the VCF files of all members of the family, we filtered variants with MAF > 0.01 from the 1000 Genomes Project databases (http://www.1000genomes.org/, 1000g2012April_all version), NHLBI GO Exome Sequencing Project (ESP6500, http://evs.gs.washington.edu/EVS/, esp6500si_all version) and the Human Genetic Variation Database (HGVD, http://www.genome.med.kyoto-u.ac.jp/ SnpDB/). A summary of WES performance is shown in Table S2. To complement the SNV and indel analyses, an algorithm that detects copy number alteration (CNA) was applied to the alignment map files to identify large CNAs as follows: 1) BAM files were converted into files covering target regions using GATK, 2) log coverage ratios and Z-scores were calculated to compare the case samples with others and 3) regions with CNAs were detected using circular binary segmentation with DNAcopy (R/ Bioconductor) 32 . PCR and direct Sanger sequencing using a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) were conducted with a 3130 Genetic Analyzer (Applied Biosystems) and each variant was identified in the sequences of both strands (Table S3). Computer-assisted splice-site prediction was accomplished using MaxEntScan (http://genes.mit.edu/ burgelab/maxent/Xmaxentscan_scoreseq.html), NNSpice 0.9 version (http://www.fruitfly.org/seq_tools/ splice.html) and Human Splicing Finder Version 2.4.1 (http://www.umd.be/HSF/).
In vitro Minigene assay. We amplified a 3,318-bp genomic DNA fragments by PCR, including intron 5 and flanking sequences of exons 5 and 6 of GPR143 from subject II:1 carrying the hemizygous G variant and his healthy father I:1 carrying the wild-type T. PCR primers containing the appropriate restriction enzyme sites were used to amplify these DNAs (Table S3). PCR products were cloned into an Exontrap Cloning Vector pET01 (MoBiTec GmbH, Göttingen Germany) for the minigene assay. The sequences of wild-type (pET01-INT5-WT) and mutant (pET01-INT5-MUT) plasmids were validated using Sanger sequencing.
The wild-type and mutant transcripts were produced by transient transfection of HCT116 or HEK293 cells with each plasmid using the X-tremeGENE HP DNA Transfection Reagent (Roche, USA) according to the manufacturer's instructions. Total RNA was extracted from cells 48 h after transfection using an RNAiso Plus kit (Takara, Otsu, Japan) and reverse-transcribed using ReverTra Ace reverse transcriptase (Toyobo, Osaka, Japan). Aberrantly spliced transcripts were identified by sequencing their RT-PCR products (Table S3) that were isolated after electrophoresis through 1.5% agarose gels.