Exome sequencing revealed a novel loss‐of‐function variant in the GLI3 transcriptional activator 2 domain underlies nonsyndromic postaxial polydactyly

Abstract Background Polydactyly is a common genetic limb deformity characterized by the presence of extra fingers or toes. This anomaly may occur in isolation (nonsyndromic) or as part of a syndrome. The disease is broadly divided into preaxial polydactyly (PPD; duplication of thumb), mesoaxial polydactyly (complex polydactyly), and postaxial polydactyly (PAP: duplication of the fifth finger). The extra digits may be present in one or both the limbs. Heterozygous variants in the GLI3, ZRS/SHH, and PITX1 have been associated with autosomal dominant polydactyly, while homozygous variants in the ZNF141, IQCE, GLI1, and FAM92A have been associated with autosomal recessive polydactyly. Pathogenic mutations in the GLI3 gene (glioma‐associated oncogene family zinc finger 3) have been associated with both nonsyndromic and syndromic polydactyly. Methods Here, we report an extended five generation kindred having 12 affected individuals exhibiting nonsyndromic postaxial polydactyly type A condition. Whole‐exome sequencing followed by variant prioritization, bioinformatic studies, Sanger validation, and segregation analysis was performed. Results Using exome sequencing in the three affected individuals, we identified a novel heterozygous frameshift variant (c.3567_3568insG; p.Ala1190Glyfs*57) in the transcriptional activator (TA2) domain of the GLI3 encoding gene. Conclusion To the best of our knowledge, the present study reports on the first familial case of nonsyndromic postaxial polydactyly due to the GLI3 variant in Pakistani population. Our study also demonstrated the important role of GLI3 in causing nonsyndromic postaxial polydactyly.


| INTRODUCTION
Polydactyly or hexadactyly is a common congenital limb de formity evident prenatally or instantly after birth. Polydactyly has a general population incidence of approximately 1.6−10.7/1,000 in live births (Malik, 2014;Umair, Ullah, Abbas, et al., 2018). To date, more than 300 syndromic forms of polydactyly have been characterized. Polydactyly is classi fied into three main types, postaxial polydactyly (PAP; ulnar) where the extra digit is located along fifth digit, preaxial (PPD; radial side) when the extra digit is located along the thumb or great toe, and complex polydacyly or mesoaxial polydactyly, when the extra digit originate between the second, third, and fourth digits (Umair, Ullah, Abbas, et al., 2018).
In this study, we have ascertained an extended Pakistani family exhibiting typical features of nonsyndromic PAP types A. Whole-exome sequencing followed by Sanger sequencing revealed a novel frameshift variant in the GLI3 located on chromosome 7p14.1.

| Ethical approval and family recruitment
Ethical approval for conducting the present study was ob tained from Institutional Review Board of Quaid-i-Azam University Islamabad, Pakistan. The family was recruited from the Punjab province of Pakistan. Affected and unaf fected individuals of the family were briefed about aims and objectives of the project in local language. All the partici pating individuals signed the informed consent and approved publication of data, photographs, and radiographs. A total of 16 affected individuals including six females and 10 males were observed in five generations with no generation skip ping exhibiting autosomal dominant inheritance pattern ( Figure 1a).

| Blood collection and extraction of genomic DNA
Pedigree was drawn according to the standard instructions and detailed interview with family elders clearly depicting autosomal dominant inheritance. Blood samples were col lected from thirteen affected and five normal individuals (represented by asterisks; Figure 1a) in the EDTA containing vacutainer sets (BD, Franklin Lakes, New Jersey). Genomic DNA was extracted from blood samples using commercially available kit following standard protocols.

| Whole-exome sequencing
DNA of three affected individual (V-2, V-7, V-13) was exome sequenced using Illumina HiSeq 2500 (Illumina, San Diego, CA) and libraries were prepared using the Agilent SureSelect Target Enrichment Kit as described earlier (Umair, Ullah, Abbas, et al., 2018). Owing to the dominant inheritance pat tern observed in the pedigree (Figure 1a), only heterozygous variants were filtered and further validated using Sanger sequencing.

| Clinical description
The postaxial polydactyly type A (PAPA) was observed as the hallmark feature in all affected individuals, while most of the affected individuals had undergone surgical procedure. The PAP type A was observed in both hands and feet, mostly affecting all four autopods (13/16). Affected individuals presented PAP with well-developed extra figure, while syn dactyly of the fourth and fifth digits/toe and polysyndactyly of fifth-sixth figures/digits was also observed in most of the affected individuals (9/16) (Figure 1b,c). There were no neurological, craniofacial, cardiovascular, obesity, ophthal mological abnormalities observed in the affected and normal individuals of the family. The extra digits were nonfunctional and presented fixed flexion deformity and caused difficulty in daily life, thus surgically removed in most of the cases (Figure 1d). Radiographical examination of the hands of af fected individuals showed normal presentation after surgery while the feet showed duplication at metatarsal level in the left foot, while duplication in right foot occurs at the fifth  (Figure 1e). Features such as height, limb length, head shape and circumference, facial dysmorphism, throat (epiglottis), and deafness (audio gram; pure-tone hearing test), were also examined in order to rule out different syndromic defects.

| WES and Sanger sequencing
DNA of three affected individual (V-2, V-7, V-13) was subjected to exome sequencing using Illumina HiSeq 2500 (Umair, Ullah, Abbas, et al., 2018). After WES, filters were applied for screening different variants. As the pedigree demonstrated autosomal dominant inheritance pattern, thus only heterozygous variants were given priority. We obtained 61 common heterozygous variants in the exome data of all three affected individuals. Further filtration identified a novel frameshift variant (c.3567_3568insG; p.Ala1190Glyfs*57) in exon 15 of the GLI3. Using Sanger sequencing approach, we have demonstrated that the variant is perfectly segregat ing with the disease phenotype in all members of the fam ily. The mutation (p.Ala1190Glyfs*57) is present in a highly conserved transcriptional activation (TA2) domain ( Figure  1j). The identified variant was not observed in the ExAC browser (http://exac.broadinstitute.org/), gnomAD (http:// gnomad.broadinstitute.org/), 1,000 Genomes, Pakistan Genetic Mutation database (Qasim et al., 2018) and in 135 in-house exomes (Pakistani exomes).

| DISCUSSION
Pathogenic sequence variants in the GLI3 have been asso ciated with both syndromic and nonsyndromic polydactyly. To date, 236 different mutations in the GLI3 gene has been listed. This included only eight variants causing nonsyndro mic polydactyly phenotype. Most of the mutations in the GLI3 result in syndromic forms including GCPS and PHS.
Here, using WES and Sanger sequencing, we have identi fied a novel frameshift variant (c.3567_3568insG) in exon 15 of the GLI3 (NM_000168.6). The variant (c.3567_3568insG) successfully co-segregated with the polydactyly phenotype in the family. The identified variant resulted in a frameshift and created a premature stop codon 57 amino acids downstream of the site of mutation (p.Ala1190Glyfs*57). It is highly likely that this mutation might result either in truncated GLI3 protein or complete loss of transcript through nonsense me diated mRNA decay.
The GLI3 functions as an important tissue patterning and developmental regulator. It is one of the three GLI transcrip tion factors (GLI1, GLI2, GLI3) playing important role in the canonical Hedgehog (HH) signaling pathway (Hui & Angers, 2011). Human GLI3 is constructed with 15 exons encoding 1,580 amino acids protein. The GLI3 functional domains comprises an N-terminal transcriptional repressor, five zinc finger (mediate DNA binding), protease cleavage site, CBP-binding regions (TA/CBP), two C-terminal tran scriptional activation (TA2 and TA1) and an α-helical region (Figure 1g). The CBP-binding region expressed ubiquitously and functions as transcriptional co-activator. The α-helical acts as an activation domain (Naruse, Ueta, Sumino, Ogawa, & Ishikiriyama, 2010). The mutation identified in the present study is located in the conserved transcriptional activation 2 (TA2) domain, which is predicted to result in loss of the TA1 and the α-helical region thus resulting in a shorter GLI3 protein.
Several studies have elucidated genotype-phenotype cor relation in terms of location of the mutation in particular domain and the resulting phenotype (Ni et al., 2018;Wang et al., 2014). Currently, most GLI3 mutations causing syn dromic and nonsyndromic phenotypes are loss-of-function variants (Ni et al., 2018). Mutations in N-terminal and the Cterminal regions are mostly associated with the GCPS, while the PHS phenotype mostly results due to mutations in the central part of the protein (Demurger et al., 2015;Jamsheer et al., 2012). The GLI3 mutation causing isolated polydactyly is not restricted to one specific domain. Our data also supports Wang et al. (2014) observation, that GLI3 mutations caus ing nonsyndromic polydactyly are located in all functional domains except the TA/CBP domain. Identification of more nonsyndromic polydactyly cases due to GLI3 mutations will help to further outline such association and present proper genotype-phenotype correlations.
In conclusion, we have reported first study of a novel lossof-function variant in the GLI3 responsible for nonsyndromic PAP type A in a Pakistani family. The present study increase the mutation spectrum of GLI3 associated pathogenesis and also addresses a thought whether a specific mutation type is a separate identity, which might lead to different digit/limb deformities.

ACKNOWLEDGMENTS
We highly appreciate cooperation and participation of the family members in this study.