Background

Brachydactyly (BD) is an autosomal dominant hand and foot malformation that is characterized by abnormally shortened or missing phalanges and/or metacarpals/metatarsals [1]. It may occur as an isolated trait or as part of a complex malformation syndrome [2]. According to the Bell classification, heritable isolated BDs have been classified into subtypes A to E according to their patterns of skeletal involvement [3]. Clinical features and pathogenic genes of isolated BDs are listed in Table 1 [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]. Brachydactyly type B1 (BDB1, OMIM 113,000) is the most severe form of brachydactyly and is characterized by shortening or hypoplasia of the distal and middle phalanges of digits 2 through 5 with or without nail dysplasia, fusion of the middle and distal phalanges, variable degrees of distal and proximal symphalangism, and a broad or bifid thumb. The feet are similarly but less severely affected [13,14,15].

Table 1 Types of isolated brachydactyly
Full size table

BDB1 is usually caused by a heterozygous variant in the receptor tyrosine kinase (RTK)-like orphan receptor 2 (ROR2) gene (OMIM 602,337), which is located on chromosome 9q22.31 [14, 21]. The ROR2 gene contains nine exons and spans a genomic length of approximately 228 kb. The ROR2 protein contains 943 amino acids and belongs to the ROR family of RTKs. It consists of an extracellular region, a transmembrane section, and an intracellular region [22]. Heterozygous variants that truncate the intracellular portion of ROR2, either the N-terminal or C-terminal of the TK domain, are the leading causes of BDB1 [22, 23]. In addition, homozygous or compound heterozygous variants in the ROR2 gene are responsible for recessive Robinow syndrome [22].

In recent years, whole-exome sequencing has become a routine strategy for discovering potential causal variants in inherited Mendelian disorders. It provides a cost-effective, fast-track approach to variant discovery and considerably increases the overall diagnostic rate. In this study, we identified the disease-causing variant in a five-generation Chinese family with BDB1 using whole-exome sequencing and Sanger sequencing. Pathogenicity was inferred by bioinformatic analysis.

Case presentation

A five-generation Chinese family with BDB1 was recruited at Central Hospital Affiliated to Shenyang Medical College. Clinical information of all family members and blood specimens from the proband and her parents were obtained. Among the 13 affected individuals, 10 had brachydactyly, and the remaining three had brachydactyly with syndactyly (Fig. 1). The 1-year-old female proband was physically examined and received a radiographic examination of the right hand before surgery. She exhibited bilateral shortening and hypoplasia of the distal and middle phalanges of digits 2 to 5 with cutaneous syndactyly of right fingers 3 to 4. The proband underwent digital separation on her right hand. Her mother underwent digital separation on both hands before enrolment in this study. Therefore, her phenotypic information was determined after reviewing her medical records and performing a physical examination. She had shortening/hypoplasia of the distal and middle phalanges of digits 2 to 5 with cutaneous syndactyly of fingers 3 to 4 on both hands, and she also had a broad hallux on both feet with nail dysplasia (Fig. 2). The phenotypic information of the other family members was provided by the proband’s mother.

Fig. 1
figure 1

Pedigree of the family with BDB1. Black shading on the left half represents brachydactyly, and grey shading on the right half represents syndactyly. The proband is marked by an arrow

Full size image
Fig. 2
figure 2

Phenotype of the proband and her mother. a A photograph of the proband's right hand before surgery. b An X-ray of the proband’s right hand before surgery. c A picture of the proband’s hands after the operation. d A photograph of the proband's feet. e A photograph of the mother’s hands after the operation. f A photograph of the mother’s feet

Full size image

Genomic DNA was extracted from the peripheral venous blood of the proband using the QIAamp DNA Blood Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s protocol and the AI Exome V1 Kit (iGeneTech, AIEV1). The extracted DNA was used for whole-exome capture and enrichment. Exome sequencing was performed on the NovaSeq 6000 sequencing platform (Illumina). Gene variants were initially annotated using the wANNOVAR tool (http://wannovar.wglab.org/). Those variants with minor allele frequencies greater than 0.01 according to public genomic databases, such as 1000Genomes, dbSNP, Exome Variant Server, gnomAD, and the Exome Aggregation Consortium were filtered out. The c.1320dupG, p.(Arg441Alafs*18) variant was identified in exon 8 of ROR2 by whole-exome sequencing and Sanger sequencing. This variant results in a frameshift at amino acid position 441. This frameshift creates a polypeptide chain of 17 new amino acids and a premature termination codon at amino acid position 458. To read the duplication sequencing chromatogram more clearly, the purified PCR fragment of the proband was inserted into the pMD-18 T vector. The selected clones were sequenced using the universal primers M13F and M13R (Fig. 3a). At the protein level, the c.1320dupG variant is predicted to produce a truncated protein that lacks tyrosine kinase and serine/threonine- and proline-rich structures, resulting in a loss of the whole intracellular region (Fig. 3b). Three-dimensional structures of wild-type and mutant ROR2 proteins were generated by the Robetta online server (http://robetta.org/) and validated using different parameters, such as the Ramachandran plot, ERRAT score, confidence and P value [24, 25]. The first two parameters were calculated using Structure Analysis and Verification Server (SAVES) version 6.0 (https://saves.mbi.ucla.edu/), and the last two parameters were determined using the ModFOLD Model Quality Assessment Server (https://www.reading.ac.uk/bioinf/ModFOLD/) [26,27,28,29]. Illustrations were prepared using VMD (www.ks.uiuc.edu/Research/vmd/). Molecular modelling shows that p.(Arg441Alafs*18) severely affects the tertiary structure of the remaining peptide chain of ROR2 (Fig. 4). The Ramachandran plot shows that the total number of residues in the most favoured regions for the wild-type and mutant ROR2 proteins was 688 (85.9%) and 330 (86.2%), respectively, and the corresponding number of residues in disallowed regions was 5 (0.6%) and 4 (1.0%), respectively. The ERRAT scores of wild-type and mutant ROR2 proteins were 89.732 and 86.9863, respectively. The confidence scores and P values of the wild-type (0.04122) and mutant proteins (0.05723) were less than 0.1 (Additional file 1). The above results indicate that the three-dimensional structures of the wild-type and mutant ROR2 proteins we predicted were reliable.

Fig. 3
figure 3

Gene variant analysis of a family with BDB1. a The novel heterozygous variant c.1320dupG, p.(Arg441Alafs*18) in ROR2 was verified by Sanger sequencing. The variant is marked by a red arrow. CS: Clone sequencing. b A schematic diagram showing the encoded domain structure of the ROR2 gene. A recurrent variant is marked in black, and the novel variant identified in this study is highlighted in red. This ROR2 variant results in the loss of the whole intracellular region

Full size image
Fig. 4
figure 4

Three-dimensional structures of the wild-type and variant ROR2 proteins. Molecular modelling by Robetta showing that p.(Arg441Alafs*18) remarkably altered the tertiary structures of the remaining peptide chain of the ROR2 protein. a Wild-type; b Mutant type

Full size image

Discussion and conclusion

ROR2 is a single-pass transmembrane protein. The domain structure of ROR2 is composed of an extracellular immunoglobulin-like domain (Ig, aa residues 55–145), frizzled-like cysteine-rich domain (FRZ or CRD, aa residues 169–303), Kringle domain (Kr, aa residues 316–394), transmembrane domain (TM, aa residues 404–424), and intracellular tyrosine kinase (TK, aa residues 473–746), and serine/threonine-rich (ST1, aa residues 753–782; ST2, aa residues 859–882) and proline-rich (PRD, aa residues 784–857) structures [22]. The extracellular region interacts with Wnt5a, which binds to the CRD region of ROR2 [30]. Both canonical and noncanonical Wnt5a/Ror2 signalling play an important role in limb skeletal development and morphogenesis [30,31,32]. The intracellular region of ROR2 is associated with many factors, such as 14–3-3 protein beta/alpha, nuclear receptor subfamily 2 group C member 2, proto-oncogene tyrosine-protein kinase Src, Wilms tumour protein 1-interacting protein, and SHC-transforming protein 1 [23, 30, 33,34,35,36]. Activation of ROR2 kinase requires the intracellular C-terminal region for recruitment of the nonreceptor kinase Src [34, 37].

Up to the time of this article’s publication, a total of 15 different variants in the ROR2 gene were found to be associated with BDB1 (Table 2). All the documented BDB1-causing ROR2 variants were nonsense or frameshift variants and were clustered in the last two exons (exons 8 and 9) or in the last intron (intron 8). Thus, the mutant mRNAs were able to escape degradation by nonsense-mediated decay, resulting in a partial or complete truncation of the intracellular portion of the ROR2 protein [38]. These mutations are believed to have a specific gain-of-function effect, not a simple haploinsufficiency [39]. The BDB1-related variants can be divided into distal and proximal variants. The distal variant that is located after TK produces a truncated protein that lacks the ST1, ST2, and PRD domains, leading to a more severe phenotype [14, 40]. In contrast, the proximal variant results in a loss of the whole intracellular region, causing a less severe but more variable phenotype [37].

Table 2 Clinical manifestation of ROR2 causing BDB1
Full size table

In this study, the novel proximal frameshift variant c.1320dupG, p.(Arg441Alafs*18) in the ROR2 gene was identified in a Chinese family with BDB1. This variant is predicted to produce a premature termination codon at amino acid position 458 and a new polypeptide consisting of 17 amino acids at the variant position. The location of this variant was very similar to the locations of c.1321_1325delCGGCG, p.(Arg441Thrfs*16), and c.1324C > T, p.(Arg442*), but their phenotypes were different. Individuals heterozygous for c.1321_1325delCGGCG, p.(Arg441Thrfs*16) exhibited bilateral hypoplasia of the distal and middle phalanges of the fingers and toes 2 to 5 to varying degrees, with or without hypoplasia of the nails. In contrast, individuals homozygous for c.1321_1325delCGGCG, p.(Arg441Thrfs*16), showed severe skeletal defects that primarily affected the distal limbs and the spine [14]. Interestingly, a heterozygous c.1324C > T, p.(R442*) carrier had a normal limb phenotype, and a homozygous c.1324C > T, p.(R442*) individual exhibited features of recessive Robinow syndrome in conjunction with severe recessive brachydactyly [39]. The c.1320dupG, p.(Arg441Alafs*18) variant in the ROR2 gene reported by our study caused bilateral or unilateral shortening/hypoplasia of the distal and middle phalanges of digits 2 to 5 with or without cutaneous syndactyly of fingers 3 to 4. The relationship between the ROR2 gene variants and phenotypes is not completely clear. Previous studies suggest that BDB1 variants result in a gain in function, whereas Robinow syndrome variants result in a loss of function [14, 49]. Further research should investigate the reasons for the different phenotypes resulting from these similar gene variants.

In conclusion, we report the novel variant c.1320dupG, p.(Arg441Alafs*18) in the ROR2 gene in a Chinese family with BDB1. Our study extends the gene variant spectrum of BDB1 and provides information for the genetic counselling of family members.