Frameshift variant in MITF gene in a large family with Waardenburg syndrome type II and a co-segregation of a C2orf74 variant

Waardenburg syndrome (WS) is a hereditary disorder affecting the auditory system and pigmentation of hair, eyes, and skin. Different variants of the disease exist with the involvement of mutation in six genes. The aim of the study is to identify the genetic defects underlying Waardenburg syndrome in a large family with multiple affected individuals. Here, in this study, we recruited a large family with eleven affected individuals segregating WS type 2. We performed whole genome SNP genotyping, whole exome sequencing and segregation analysis using Sanger approach. Whole genome SNP genotyping, whole exome sequencing followed by Sanger validation of variants of interest identified a novel single nucleotide deletion mutation (c.965delA) in the MITF gene. Moreover, a rare heterozygous, missense damaging variant (c.101T>G; p.Val34Gly) in the C2orf74 has also been identified. The C2orf74 is an uncharacterized gene present in the linked region detected by DominantMapper. Variants in MITF and C2orf74 follows autosomal dominant segregation with the phenotype, however, the variant in C2orf74 is incompletely penetrant. We proposed a digenic inheritance of variants as an underlying cause of WS2 in this family.


Introduction
Waardenburg syndrome (WS) is a group of rare hereditary disorders. It is characterized by pigmentary defects of hair (white forelock), eyes (heterochromia iridis) and skin (hypo-pigmented skin), abnormalities in the inner ear (bilateral sensorineural hearing impairment), and dystopia canthorum (lateral displacement of the inner canthi of the eyes). WS syndrome has been categorized into four major types (WS1, WS2, WS3 and WS4). Further subtypes do exist though additional clinical manifestations are required for differential diagnosis of subtypes. WS1 (OMIM: 193500) and WS3 (OMIM: 148820) are characterized by the occurrence of dystopia canthorum. Absence of dystopia canthorum and other midface defects are the  [1][2][3]. Clinically WS1 and WS2 subtypes are the most common syndromes.
The clinical features of WS are not completely penetrant and highly varied expression has been observed [4]. This makes the differential diagnosis challenging [5]. Genetic diagnosis can help in identifying the type of WS in an individual patient. Mutations in at least six genes have been reported as an underlying cause of WS. Genotype-phenotype correlation between a causative gene mutation and various clinical features in WS are largely unclear. Until recently, mutations in EDN3 (20q13.32), EDNRB (13q22.3), MITF (3p14p13), PAX3 (2q36.1), SNAI2 (8q11.21), and SOX10 (22q13.1) have been reported in isolated patients as well as in families segregating WS [6][7][8]. For an instance, mutations of PAX3 gene is recognized as a key indicator for the clinical features of WS1 and WS3, and mutations in MITF, SOX10, and SNAI2 genes are identified in WS2. Moreover, genetic mutations in EDN3, SOX10, and EDNRB have been implicated in WS4 [9][10][11][12][13][14]. The majority of WS1 cases and some moderate WS3 cases are due to heterozygous mutations in the PAX3 gene. Heterozygous mutations in MITF and SOX10 are estimated to account for 30% of WS2 cases, and only 5% of the cases are attributed to the heterozygous mutations in EDNRB and EDN3, and homozygous mutations in SNAI2 gene. In WS4, approximately 50% of cases have heterozygous mutations in SOX10 and homozygous or heterozygous mutations in EDNRB and EDN3 genes lead to WS4 in 20-30% of patients [15]. WS2 is phenotypically well characterized. It is further sub-divided into at least five types, WS2A (OMIM: 600193), WS2B (OMIM: 606662), WS2C (OMIM: 193510), WS2D (OMIM: 611584), and WS2E (OMIM: 608890).
We pursued our previous findings of identification of common genomic regions in multiple WS2 patients [16] and performed whole exome sequencing in two DNA samples of individuals from an extended family. We identified pathogenic variants in MITF and C2orf74 genes as an underlying cause of WS2 phenotype.

Ethical approval and sample collection
This study was approved by the institutional ethics review committee at the College of Medicine, Taibah University (study approval ID 051-02-2017) and the collaborating hospital. Written informed consents were obtained from all individuals included in the study or from their legal guardians. All experiments on human DNA were carried out in accordance with the declaration of Helsinki. A large six generation pedigree was drawn by interviewing the elders of the family (Fig 1). All subjects were examined at Magrabi Eye and Ear Hospital Almadinah Almunawwarah, Saudi Arabia. Detailed clinical history was recorded. During clinical phenotyping, colour of skin, hair and iris were noted. Moreover, special attention was given to observe dystopia canthorum and other anomalies such as medial eyebrow flare, limb defects, and Hirschsprung disease.

Extraction of DNA and genetic analysis
Genomic DNA was extracted from the peripheral blood samples of all nineteen individuals using the QIAamp DNA Blood Mini Kit (QIAGEN GmbH-Germany). DNA was quantified using micro-volume spectrophotometer (MaestroGen, Hsinchu City 30091, Taiwan). Whole genome SNP genotyping was performed as described elsewhere [17,18] to identify the shared chromosomal segments in affected individuals. DNA from two affected individuals (VI:1, VI:2) were used to sequence the complete coding region of the human genome (whole exome sequencing). Whole exome was captured by Agilent SureSelect Target Enrichment Kit (v6) and sequenced by Macrogen Inc., using Illumina NovaSeq 6000 sequencing instrument (Illumina Inc, California, US). The length of the paired end reads were 150 bps, and the average coverage of the captured region was approximately 100x. Reads were aligned to the human reference genome (GRCh38/hg38). Genome Analysis Toolkit (GATK v3.7) was used to call the variants and variants were annotated using Illumina VariantStudio software. EVS (Exome Variant Server evsClient-v.0.0.16), GnomAD (Genome Aggregation Database v2.2.1), dbSNP (database of Single Nucleotide Polymorphisms dbSNP 2.0 Build 153), ExAC (Exome Aggregation Consortium), PolyPhen2 (Polymorphism Phenotyping v2), SIFT (Sorting Intolerant from Tolerant) and ClinVar were used to annotate, filter and prioritize the variants. Variants of interest were Sanger sequenced in both affected individuals by designing variant-specific primers. Primer3 tool was used for primer designing. Validated variants were screened in the DNA of all available individuals followed by segregation analysis.

Clinical description of cases
Ocular hypopigmentation was present in all affected individuals except IV:5, IV:7, V:2, V:8, VI:2. All affected individuals have profound to mild hearing impairment (HI). W index was less than 1.95 mm for each affected individual, therefore, dystopia canthorum was considered as absent. This feature helped us in excluding WS type 1 (Fig 2). Moreover, no integumentary hypopigmentation in the form of white forelock or leukoderma was observed in any of the individual. Furthermore, synophrys, broad nasal root, hypoplasia of alae nasi or premature greying of the hair were not observed. Detailed examination ruled out the presence of upper limb malformations and chronic constipation. This ruled out WS type III and WS type IV in this family. Based on the clinical phenotyping, the individuals in this study segregates WS type II in an autosomal dominant manner. Detail clinical analysis of each individual can be found elsewhere [16] (Fig 2).

Genetic analysis identified a novel deletion mutation in the MITF gene and a missense variant in the C2orf74 gene
Whole exome sequencing (WES) of two affected sisters (VI:1 and VI:2) produced~62 million raw reads, with~60 million high-quality reads aligned to the GRCh38/hg38 reference genome. Overall, a total of 101,648 high-quality variants were identified and 95.5% of these variants are present in the dbSNP142. Coding missense and indel variants and variants in the flanking intronic regions (within 10 bps) accounted for 48,934 variants, while 12,733 variants were synonymous. Various filter were applied to search for potential candidate variants including depth of coverage (DP >10), genotype quality (GQ >20), and unreported SNP. Only variants with low allelic frequency (below 1% in 1000G, ExAC) were taken into consideration. These settings yielded 67 rare coding variants (S1 Table). Variants in all known WS candidate genes (EDN3, EDNRB, MITF, PAX3, SOX10, SNAI2, and TYRO3) were searched and a novel rare heterozygous deletion mutation (c.965delA; p.Asn322fs) was identified in the MITF gene in both patients. Moreover, heterozygous missense variants in SNAI3 (c.607C>T; p.Arg203Cys) and TYRO3 (c.1037T>A; p.Ile346Asn) gene was identified in the exome data of both patients. Variant in SNAI3 (c.607C>T; p.Arg203Cys) gene is rare in population and is probably damaging and deleterious as predicted by PolyPhen2 and SIFT, respectively. Variant in TYRO3 (c.1037T>A; p.Ile346Asn) gene is present in population databases with high frequency (0.22 MAF) and is benign and tolerated as predicted by PolyPhen2 and SIFT, respectively.
Studies have shown that WNT pathway genes including LEF-1 may modulate the WS2 phenotype in WS2 cases with MITF mutation [19,20]. Therefore, exome data was searched for variants in WNT pathway genes (LEF-1, RNF43, APC, ZNRF3, LRP4, LRP5, LRP6, ROR1, ROR2, GSK3, CK1, APC, BCL9, and BCL9L) as well. No potentially pathogenic rare variant was identified. In order to identify variant(s) in other genes which might influence the expressivity of WS phenotype in our cases, exome data was filtered by using an unbiased and hypothesis-free approach. A rare missense variant (c.101T>G; p.Val34Gly) in the C2orf74 gene was identified in both affected individuals. This variant is possibly damaging and deleterious as predicted by PolyPhen2 and SIFT, respectively. Our previous genetic work on this family identified regions on chromosome 2p16.3-p15 and 18q21.33-q22.1 segregating with WS2 phenotype [16]. Interestingly, C2orf 74 gene lies in the region on chromosome 2.

Segregation analysis confirmed that the variants in MITF and C2orf74 segregate with the disease phenotype in the family
The relevant coding exon of four genes (C2orf74, MITF, SNAI3, and TYRO3) were sequenced in all nineteen available members of a family including 11 affected individuals. Sanger sequencing validates the exome discovered mutations in both affected individuals and confirmed the segregation of MITF mutation in all family members as per autosomal dominant inheritance. Moreover, variant in C2orf74 segregates with the phenotype in all family member except V:1 and V:2. V:1 is an unaffected individual carrying heterozygous variant while V:2 is an affected individual with wild type sequence (Fig 3). Variants in SNAI2 and TYRO3 are not segregating in the family (Fig 1).

Discussion
Waardenburg syndrome (WS) is clinically characterized by occurrence of hearing impairment (HI) along with pigmentation abnormalities, including patchy hypopigmentation of the skin and hair and heterochromia iridis. In a subset of patients some other clinical features like Hirschsprung disease, dystopia canthorum, limb defects or neurological abnormalities are present. These features are often used to classify WS in four types (WS1, WS2, WS3, and WS4). WS is generally inherited in an autosomal dominant manner with variable penetrance. Inter-and intrafamilial variability in the expression of symptoms have also been observed. In some cases of WS, biallelic mutations have also been reported suggesting an autosomal recessive inheritance [8,[21][22][23][24][25][26][27].
Patients with Waardenburg syndrome type 2 (WS2) shows congenital sensorineural HI with depigmentation of hair, skin, and eyes. Dystopia canthorum is absent in WS2. It is a highly heterogeneous syndrome in terms of the underlying genetics. WS2 can be subdivided into five different types including WS2A (193510) caused by mutations in the MITF gene [4, 8,

PLOS ONE
12, 13,20,[27][28][29][30][31][32][33][34][35][36][37], WS2B (600193) mapped to chromosome 1p21-p13.3, WS2C (606662) mapped to chromosome 8p23, WS2D (608890) caused by mutation in the SNAI2 gene [38], and WS2E caused by mutations in the SOX10 gene [39][40][41][42][43][44][45][46]. It is thought that haploinsufficiency is the underlying causative mechanism for WS2 [47]. Although, mutations in three genes (MITF, SOX10, and SNAI2) have been shown to cause WS2 subtypes, a number of cases remain unexplained at the molecular level. Moreover, inter-and intrafamilial phenotypic variability cannot be explained due to mutations in a single gene. Digenic inheritance or modulation of phenotype by modifier genes could possibly explain the variability and expressivity. Therefore, we embarked a genetic study using a large family segregating WS2. SNP genotyping followed by detection of shared regions identified two regions on chromosome 2 and 18 [16]. In this study, we sequenced complete exome in two affected individuals and identified candidate variants in MITF (c.965delA), SNAI2 (c.607C>T) and C2orf74 (c.101T>G) genes. Variant in SNAI2 is not segregating with the disease phenotype therefore it was excluded as an underlying cause of WS2 in the family. MITF variant is perfectly segregating in all the affected members (Fig 3). Segregation analysis of C2orf74 variant revealed that the variant segregates with the WS2 phenotype in all family members except V:1 and V:2. V:1 is an unaffected individual carrying heterozygous variant while V:2 is an affected individual with wild type sequence (Fig 1). Presence of heterozygous variant in an unaffected individual (V:1) might be due to incomplete expression of the phenotype.
Mutations in melanocyte inducing transcription factor (MITF), coding for a basic helixloop-helix (BHLH) leucine zipper protein, are known to cause the WS2 phenotype due to defects in survival, proliferation, and migration of melanocytes [13,28]. The deletion mutation (c.965delA) identified in this study lies in the BHLH domain and predicted to cause frameshift (p.Asn322fs) and stop codon seven amino acids downstream (Asn322Metfs � 7). The missense variant (c.101T>G) in the C2orf74 gene changes the conserved amino acid Valine to Glycine (p.Val34Gly). C2orf74 is an uncharacterized gene and no functional data is available, however, Expression Atlas detected the expression of the gene in the eye (https://www.ebi.ac.uk/gxa/ home). The gene C2orf74 is present in the genomic region shared by all affected individuals [16] and therefore, we consider this as a candidate gene for WS2 phenotype. C2orf74 gene might interact with MITF gene product and give rise to the spectrum of phenotype varying from severe phenotype with complete penetrance to partial features.

Conclusion
In this study, we analysed a large family segregating Waardenburg syndrome type 2 to identify the underlying genetic defects. Whole genome SNP genotyping, whole exome sequencing and segregation analysis using Sanger approach was performed and a novel single nucleotide deletion mutation (c.965delA) in the MITF gene and a rare heterozygous, missense damaging variant (c.101T>G; p.Val34Gly) in the C2orf74 was identified. Both variants follow autosomal dominant segregation with the phenotype, however, the variant in C2orf74 is not completely penetrant. We proposed a digenic inheritance of variants as an underlying cause of WS2 in this family.
Supporting information S1