Clin Genet 2014 Printed in Singapore. All rights reserved
© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd CLINICAL GENETICS doi: 10.1111/cge.12401
Original Article
De novo WNT5A-associated autosomal dominant Robinow syndrome suggests specificity of genotype and phenotype Roifman M., Marcelis C.L.M, Paton T., Marshall C., Silver R., Lohr J.L., Yntema H.G., Venselaar H., Kayserili H., van Bon B., Seaward G., FORGE Canada Consortium, Brunner H.G., Chitayat D. De novo WNT5A-associated autosomal dominant Robinow syndrome suggests specificity of genotype and phenotype. Clin Genet 2014. © John Wiley & Sons A/S. Published by John Wiley & Sons Ltd, 2014 Robinow Syndrome (RS), a rare skeletal dysplasia syndrome, is characterized by dysmorphic features resembling a fetal face, mesomelic limb shortening, hypoplastic external genitalia in males, and renal and vertebral anomalies. Both autosomal dominant and autosomal recessive patterns of inheritance have been reported. Since the description of autosomal dominant Robinow Syndrome (ADRS; OMIM 180700) in 1969 by Meinhard Robinow and colleagues, the molecular etiology remained elusive until only recently. WNT5A was proposed to be the candidate gene for ADRS, as mutations were found in two affected families, one of those being the originally described index family. We report three families with RS caused by novel heterozygous WNT5A mutations, which were confirmed in the first family by whole exome sequencing, and in all by Sanger sequencing. To our knowledge, this is the largest number of published families with ADRS in whom a WNT5A mutation was identified. Families 1 and 2 are the first cases showing de novo inheritance in the affected family members and thus strengthen the evidence for WNT5A as the causative gene in ADRS. Finally, we propose WNT5A mutation specificity in ADRS, which may affect interactions with other proteins in the Wnt pathway. Conflict of interest
The authors declare no conflict of interest.
M. Roifmana,b,† , C.L.M. Marcelisc,† , T. Patond , C. Marshalld , R. Silvera , J.L. Lohre , H.G. Yntemac , H. Venselaarc , H. Kayserilif , B. van Bonc , G. Seawardg , FORGE Canada Consortium, H.G. Brunnerc and D. Chitayata,b a The
Prenatal Diagnosis and Medical Genetics Program, Department of Obstetrics and Gynecology, Mount Sinai Hospital, b Division of Clinical and Metabolic Genetics, Department of Paediatrics, The Hospital for Sick Children, University of Toronto, Toronto, ON, Canada, c Department of Human Genetics, Radboud University Medical Center, Nijmegen, The Netherlands, d The Centre for Applied Genomics and Program in Genetics and Genome Biology, The Hospital for Sick Children, Toronto, ON, Canada, e Lillehei Heart Institute, University of Minnesota, Minneapolis, MN, USA, f Medical Genetics Department, I˙stanbul Medical Faculty, Istanbul University, Istanbul, Turkey, and g Department of Obstetrics and Gynecology, Mount Sinai Hospital, University of Toronto, Toronto, ON, Canada † These
authors contributed equally.
Key words: autosomal dominant – Robinow syndrome – skeletal dysplasia – WNT5A Corresponding authors: Dr David Chitayat, Department of Obstetrics and Gynecology, The Prenatal Diagnosis and Medical Genetics Program, Room 3292, The Ontario Power Generation Building, 700 University Avenue, M5G 1Z5 Toronto, ON, Canada. Tel.: +416 586 4523; fax: +416 586 4723;
1
Roifman et al. e-mail: [email protected] and Dr Han Brunner, Radboud University Medical Center, Department of Human Genetics, 20 Geert Grooteplein, Nijmegen, Gelderland 6525GA, The Netherlands. Tel.: +312 436 14017; fax: +312 436 68752; e-mail: [email protected] Received 26 January 2014, revised and accepted for publication 7 April 2014
Robinow Syndrome (RS), a rare skeletal dysplasia, is characterized by dysmorphic features resembling a fetal face, mesomelic limb shortening, hypoplastic external genitalia, and renal and vertebral anomalies. Two types of RS have been reported and are distinguished by the severity of their signs and symptoms and by their pattern of inheritance; the severe form of RS has an autosomal recessive mode of inheritance and the milder form is inherited in an autosomal dominant fashion. Since the description of autosomal dominant Robinow Syndrome (ADRS; OMIM 180700) in 1969 by Robinow and colleagues (1), the molecular etiology remained elusive until only recently (2). Person et al (2) proposed the WNT5A gene as the candidate gene for ADRS, with mutations identified in two affected families including the originally described index family. However, the segregation of the gene mutation with the condition in an affected family, although suggestive, could not serve as a proof that this is the causative gene. WNT5A is a member of the Wnt family of proteins, which regulate critical morphogenic events, including embryonic patterning, cell differentiation, growth and migration (3–6). Mutations in the more widely published ROR2 gene, a co-receptor of WNT5A, have been implicated in autosomal recessive RS and Brachydactyly type B. Many cases of ROR2-associated RS have been described (7–9). Since Person et al.’s (2) identification of WNT5A mutations in two cases of ADRS, there has been no reported confirmation of this association. We report three additional families with WNT5A-associated ADRS, which prove that this is the causative gene in ADRS. Clinical reports Family 1
The proband (the mother) presented to the Prenatal Genetics clinic at the age of 37 in her first pregnancy. She was born to 35-year-old parents who were healthy, had normal stature and were non-consanguineous, of German-British-Italian descent. At birth, she was noted to have facial dysmorphism and short upper and lower limbs with normal weight and head circumference. Skeletal X-rays at the age of 4 months showed that all long bones were short and the radial heads were
2
dislocated. The type of her skeletal dysplasia could not be delineated. She was later noted to have dental problems, including persistent primary teeth requiring extraction at 18 years of age, and subsequent misalignment of permanent teeth. She had normal development and cognition. Her pregnancy was initially uncomplicated and there was no history of maternal exposures. NT was measured at 11.5-weeks gestation and was 1.0 mm and screening for Down syndrome and open neural tube defects was screen negative. A detailed fetal ultrasound at 21 weeks gestation suspected intrauterine growth restriction (IUGR) showing that all long bones measured below −2 SD (2 weeks behind gestational age), suggestive of a skeletal dysplasia. The bones appeared well mineralized and there were no fractures. The couple was counseled that the baby had a non-lethal skeletal dysplasia, likely the same diagnosis as his mother, and elected to continue the pregnancy with no invasive testing. The couple is non-consanguineous. The father of this pregnancy is a healthy 37-year-old man who was adopted out after delivery. He is Caucasian but of unknown descent. The mother (i.e. proband) measured 150 cm tall (−2.5 SD) and the father measured 170 cm tall (10th–25th centile). The baby was born at term via Caesarian section for breech presentation. Birth weight was 2360 g (3rd centile), length was 43 cm (−2.5 SD; 7 weeks behind gestational age), and head circumference was 33.2 cm (10th centile). Dysmorphic features were noted including, dolichocephaly, high forehead with frontal bossing, hypoplasia of the supraorbital ridges, prominent eyes, hypertelorism, short flat nose, anteverted nares, low set ears and micrognathia. Many of these features were shared with his mother (Fig. 1). All the four limbs were short; mesomelia was most prominent with upper arm, forearm and hand measurements plotting 1, 12 and 8 weeks behind gestational age, respectively. There was bilateral cryptorchidism and hypoplastic scrotal tissue. Although the penis appeared small due to its embedded position within the scrotum, it measured within normal limits (2.7 cm). Upper limb X-rays revealed short long bones and dislocated radial heads bilaterally (Fig. 2). On the basis of the baby’s combined features of dysmorphism resembling a fetal face, mesomelic limb shortening and genital
De novo WNT5A-associated autosomal dominant Robinow syndrome
Fig. 1. Family 1: The mother in infancy (top row) and her son (middle and bottom rows). Note the high forehead, frontal bossing, hypertelorism, hypoplasia of the supraorbital ridges, depressed nasal bridge, broad nasal root and tip, hypoplastic malar regions, micrognathia and low set ears. Skeletal features include short stature (height G; p.Tyr86Cys) in the WNT5A gene that results in an amino acid exchange predicted to disrupt protein folding and affect Wnt function. This mutation was later confirmed with direct WNT5A gene sequencing. The maternal grandparents did not show the mutation detected in their daughter and
3
Roifman et al.
Fig. 2. Upper limb X-rays of the proband in Family 1 showing short long bones and dislocated radial heads bilaterally.
He was seen again at the age of 11 months and on examination, his length was 73 cm (25th centile) and he was noted to have relatively short limbs (mesomelic). His facial dysmorphism and genital abnormalities remained the same. At the age of 2, he had bilateral orchidopexy. A follow-up at 8 years showed normal development and academic achievements. His height was 120 cm (3rd centile), his weight was 22.2 kg (10th–25th centile) and his head circumference was 49 cm (just below the 3rd centile). His arm span was 111 cm and his span/height was 0.93 (normal 0.99). Facial dysmorphism remained unchanged with a flat face, hypertelorism, prominent eyes, broad mouth and full cheeks. His arms were short with short hands [total length 13 cm (−2.5SD)] and brachydactyly [3rd finger length 5 cm (A, p.Cys69Tyr). Family 3
Fig. 3. Family 1 pedigree showing the affected proband and her son (shaded in black). Whole exome sequencing was performed on affected mother and son and unaffected maternal grandparents (*).
grandson. This confirms that the mutation arose de novo in the proband. Family 2
The proband is the second child of healthy non-consanguineous parents of Dutch–Spanish descent. He has a healthy older sister and the parents and sister are of average stature with normal development and cognition. The pregnancy was uneventful and delivery was at term and uncomplicated; his birth weight was 3530 g (50th centile), and his head circumference was 35.5 cm (25th–50th centile). Although birth length was not measured, length was 51 cm (10th centile) at the age of 4 weeks. At birth, he was noted to have facial dysmorphism with a high and broad forehead with a naevus flammeus, hypertelorism, large prominent eyes, a depressed nasal bridge, short flat nose with anteverted nares, macrostomia and gingival hypertrophy (Fig. 5). He had a micropenis and bilateral cryptorchidism. The differential diagnosis included Opitz G-BBB syndrome and RS but molecular analysis of the MID1 and ROR2 genes showed no detectable mutation.
4
The proband is the first child born to nonconsanguineous parents of Turkish descent. The pregnancy was complicated by short long bones indicating growth restriction, but with normal placental function, identified in the second trimester (precise gestational age unknown). The proband was born at term with a birth weight of 2450 g (3rd centile). On examination by a clinical geneticist at the age of 13 days, her weight was 2700 g (3rd–10th centile), length was 44 cm (3rd centile), head circumference was 32.7 cm (10th centile) and upper/lower segment ratio was 1.31 (G; p.Tyr86Cys) in the WNT5A gene that resulted in an amino acid exchange predicted to disrupt protein folding and affect Wnt function. Top: Graph of filtering criteria used to identify causative variant. Bottom: Integrative Genomics Viewer (14) image showing WNT5A heterozygous variant identified in the affected son (1st row) and his mother (2nd row), but not in the unaffected maternal grandparents (3rd and 4th rows). This mutation was later confirmed with Sanger sequencing. [*Population allele frequencies based on data from 1000 Genomes Project and National Heart, lung and blood institute (NHLBI).]
mutational analysis of the WNT5A gene was performed by Sanger sequencing on the proband and her father and revealed a novel heterozygous mutation c.257A>G (p.Tyr86Cys) in both. Wnt5a homology model
We sought to evaluate whether specificity exists among the mutations identified in all five WNT5Aassociated ADRS-affected families. The threedimensional-structure of the Wnt5a protein is still unknown. Therefore, we built a homology model using the experimentally solved structure of Wnt8 (PDB file
4f0a) from Xenopus Leavis as a template. This structure and human Wnt5 share 44% sequence identity over 300 residues. The YASARA & WHAT IF Twinset was used for modeling and subsequent visualization and analysis (10–12). Although the first 65 residues are lacking, the model covers the four mutated positions. Strikingly, all four mutations seem to be located on one side of the protein (Fig. 7). Discussion
Heterozygous WNT5A mutations have been reported previously in two families with ADRS (2). A comparison
5
Roifman et al.
Fig. 5. Family 2. The proband at different ages showing short stature, high forehead, frontal bossing, hypertelorism, hypoplasia of the supraorbital ridges, depressed nasal bridge, broad nasal root and tip, low set ears and genital hypoplasia.
Fig. 7. Wnt5a homology model. Mutations found in all WNT5Aassociated ADRS cases seem to be located on one side of the protein and may affect interactions with other proteins in the Wnt pathway (10–12).
Fig. 6. Radiographic image of the proband from Family 3 at the age of 13 days.
of their findings along with these cases is summarized in Table 1. WNT5A is a member of the Wnt family of proteins, which regulate critical morphogenic events, including embryonic patterning and cell differentiation,
6
growth and migration (3–5). Wnt5a and Ror2 have been shown to interact physically and function together to activate the noncanonical Wnt signaling pathway (6). Both Wnt5a-null mice and Ror2-null mice exhibit a similar RS-like phenotype, including shortened anterior–posterior axis, dysmorphic facial features and genital hypoplasia, but have a more severe phenotype than human ADRS and include cardiac defects, rib fusions and perinatal lethality (4–6, 13). These findings are more similar to the human ARRS phenotype
Autosomal dominant
Presumed de novo dominanta Autosomal dominant
Short long bones
Short long bones
3 (2 underwent confirmatory molecular testing)
1
8 (7 underwent confirmatory molecular testing)
Current report, Family 3
Person et al. (2010)
Robinow et al. (1969) and Person et al. (2010)
ADRS, autosomal dominant Robinow Syndrome. a Mutation analysis of unaffected parents was not performed to confirm de novo dominant inheritance.
1 Current report, Family 2
Short stature, mesomelic limb shortening, hypertelorism, mandibular hypoplasia, irregular dental alignment, genital hypoplasia.
Short long bones
c.206G>A, p.Cys69Tyr heterozygous c.257A>G, p.Tyr86Cys heterozygous Autosomal dominant
c.248G>C, p.Cys83Ser; heterozygous c.544–545 CT>TC, p.Cys182Arg; heterozygous
c.257A>G, p.Tyr86Cys heterozygous Autosomal dominant
Short long bones; Bilateral radial head dislocation Not performed 2 Current report, Family 1
Mesomelic short stature, limb shortening, malar hypoplasia, hypertelorism, short flat nose, low set ears, limited suppination, persistent primary teeth, brachydactyly, bilateral cryptorchidism and genital hypoplasia (n = 1) Short stature with limb shortening, hypertelorism, short flat nose, brachydactyly, bilateral cryptorchidism and genital hypoplasia Short stature with mesomelic limb shortening, brachydactyly, large anterior fontanel, hypertelorism, prominent eyes with bilateral epicanthic folds, wide down-slanting palpebral fissures, flat midface, short upturned nose, broad nasal bridge, anteverted nares, long philtrum, gingival hyperplasia, short oral frenulum, posteriorly rotated ears, micrognathia., genital hypoplasia, sacral dimple, hairy sacral patch Short stature, mesomelic limb shortening, marked hypertelorism, short nose
Pattern of inheritance Radiographical features Clinical features Number of family members affected Case
Table 1. A summary of the clinical, radiological and molecular analysis findings of our cases and the two other reported cases with WNT5A-associated ADRS
WNT5A molecular diagnosis
De novo WNT5A-associated autosomal dominant Robinow syndrome (OMIM 268310). Humans with homozygous WNT5A mutations have not been described and may be embryonically lethal. The heterozygous WNT5A mutations identified in two families with ADRS by Person et al. (2) were thought to act as hypomorphic alleles, based on functional expression studies. Although heterozygous WNT5A mutations were identified in all affected members of the two families previously described by Person et al. (2), the cases reported inherited the mutation by descent and thus lacked the proof that the gene mutation was the cause of the condition. Families 1 and 2 are the first reports of de novo WNT5A mutations associated with the ADRS in the absence of other likely candidate genes identified by whole exome sequencing. This confirms the WNT5A gene to be the causative gene in ADRS. Interestingly, the probands in Families 2 and 3 exhibited acquired microcephaly, a feature not previously described in RS. In our Wnt5a homology model, all four known ADRS mutations appear to be located on the same side of the protein. It is possible that mutations in these residues change the surface structure and thereby affect interactions with other proteins in the Wnt pathway. Residues C69 and C83 are modeled as free cysteines. Their specific role in multimerization and/or stabilization of the complex is unknown. However, mutation of these residues into Tyrosine and Serine, respectively, could affect this putative activity. Residue Y86 is also located on the surface. Its side chain is big and aromatic and could play a role in complex formation. The mutation converts this residue into Cysteine, which is smaller and cannot make the same interactions. Besides the loss of the Tyrosine side chain, this mutation might also introduce a new cysteine-bond with the free Cysteine at position 83. In our model, residue C182 makes a disulfide bond with residue C164. This will stabilize the local domain and the interaction surface. Mutation C182R will disturb this disulfide bond. In addition, a bigger side chain is introduced at this position, which will change the surface of the protein. Conclusions
We report three families with WNT5A-associated ADRS, caused by novel heterozygous WNT5A mutations confirmed, in the first family by whole exome, and in all by Sanger sequencing. Families 1 and 2 are the first cases showing de novo inheritance in the affected family members and thus strengthen the evidence for WNT5A as the causative gene in ADRS. We propose a specificity to the ADRS WNT5A genotype since all five families reported to date have missense mutations affecting the same part of the Wnt5a protein, and two of these harbor the identical nucleotide change. Finally, we note that all RS families with WNT5A mutations have a classical AD Robinow phenotype, whereas several less typically affected patients (with short-limbed dwarfism without a mesomelic pattern, with less typical facial dysmorphism and/or absence of genital hypoplasia) tested in our diagnostic laboratory do not show such mutations. Thus,
7
Roifman et al. both genotype and phenotype of WNT5A-associated RS are likely to be highly specific. Acknowledgements Whole exome sequencing was funded by the Government of Canada through Genome Canada, the Canadian Institutes of Health Research and the Ontario Genomics Institute (OGI-049). Additional funding was provided by Genome Quebec, Genome British Columbia, and the McLaughlin Centre. Sanger sequencing was performed by the Oxford Medical Genetics Laboratory at the Churchill Hospital in Oxford, UK in Family 1, and by the Genome Diagnostics laboratory of the Department of Human Genetics in Nijmegen, The Netherlands in Families 2 and 3.
References 1. Robinow M, Silverman FN, Smith HD. A newly recognized dwarfism syndrome. Am J Dis Child 1969: 117 (6): 645–651. 2. Person AD, Beiraghi S, Sieben CM et al. WNT5A mutations in patients with autosomal dominant Robinow syndrome. Dev Dyn 2010: 239 (1): 327–337. 3. Nusse R. Wnt signaling in disease and in development. Cell Res 2005: 15 (1): 28–32. 4. Yamaguchi TP, Bradley A, McMahon AP, Jones S. A Wnt5a pathway underlies outgrowth of multiple structures in the vertebrate embryo. Development 1999: 126 (6): 1211–1223.
8
5. Schleiffarth JR, Person AD, Martinsen BJ et al. Wnt5a is required for cardiac outflow tract septation in mice. Pedriatr Res 2007: 61 (6): 386–391. 6. Oishi I, Suzuki H, Onishi N et al. The receptor tyrosine kinase Ror2 is involved in non-canonical Wnt5a/JNK signalling pathway. Genes Cells 2003: 8 (7): 645–654. 7. van Bokhoven H, Celli J, Kayserili H et al. Mutation of the gene encoding the ROR2 tyrosine kinase causes autosomal recessive Robinow syndrome. Nat Genet 2000: 25 (4): 423–426. 8. Afzal AR, Rajab A, Fenske CD et al. Recessive Robinow syndrome, allelic to dominant brachydactyly type B, is caused by mutation of ROR2. Nat Genet 2000: 25 (4): 419–422. 9. Afzal AR, Jeffrey S. One gene, two phenotypes: ROR2 mutations in autosomal recessive Robinow syndrome and autosomal dominant brachydactyly type B. Hum Mutat 2003: 22 (1): 1–11. 10. Krieger E, Koraimann G, Vriend G. Increasing the precision of comparative models with YASARA NOVA–a self-parameterizing force field. Proteins 2002: 47 (3): 393–402. 11. Vriend G. WHAT IF: a molecular modeling and drug design program. J Mol Graph 1990: 8 (1): 52–56. 12. Krieger E, Joo K, Lee J et al. Improving physical realism, stereochemistry, and side-chain accuracy in homology modeling: four approaches that performed well in CASP8. Proteins 2009: 77 (9): 114–122. 13. Schwabe GC, Trepczik B, Süring K et al. Ror2 knockout mouse as a model for the developmental pathology of autosomal recessive Robinow syndrome. Dev Dyn 2004: 229 (2): 400–410. 14. Thorvaldsdottir H, Robinson JT, Mesirov JP. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform 2013: 14 (2): 178–192.
© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd CLINICAL GENETICS doi: 10.1111/cge.12401
Original Article
De novo WNT5A-associated autosomal dominant Robinow syndrome suggests specificity of genotype and phenotype Roifman M., Marcelis C.L.M, Paton T., Marshall C., Silver R., Lohr J.L., Yntema H.G., Venselaar H., Kayserili H., van Bon B., Seaward G., FORGE Canada Consortium, Brunner H.G., Chitayat D. De novo WNT5A-associated autosomal dominant Robinow syndrome suggests specificity of genotype and phenotype. Clin Genet 2014. © John Wiley & Sons A/S. Published by John Wiley & Sons Ltd, 2014 Robinow Syndrome (RS), a rare skeletal dysplasia syndrome, is characterized by dysmorphic features resembling a fetal face, mesomelic limb shortening, hypoplastic external genitalia in males, and renal and vertebral anomalies. Both autosomal dominant and autosomal recessive patterns of inheritance have been reported. Since the description of autosomal dominant Robinow Syndrome (ADRS; OMIM 180700) in 1969 by Meinhard Robinow and colleagues, the molecular etiology remained elusive until only recently. WNT5A was proposed to be the candidate gene for ADRS, as mutations were found in two affected families, one of those being the originally described index family. We report three families with RS caused by novel heterozygous WNT5A mutations, which were confirmed in the first family by whole exome sequencing, and in all by Sanger sequencing. To our knowledge, this is the largest number of published families with ADRS in whom a WNT5A mutation was identified. Families 1 and 2 are the first cases showing de novo inheritance in the affected family members and thus strengthen the evidence for WNT5A as the causative gene in ADRS. Finally, we propose WNT5A mutation specificity in ADRS, which may affect interactions with other proteins in the Wnt pathway. Conflict of interest
The authors declare no conflict of interest.
M. Roifmana,b,† , C.L.M. Marcelisc,† , T. Patond , C. Marshalld , R. Silvera , J.L. Lohre , H.G. Yntemac , H. Venselaarc , H. Kayserilif , B. van Bonc , G. Seawardg , FORGE Canada Consortium, H.G. Brunnerc and D. Chitayata,b a The
Prenatal Diagnosis and Medical Genetics Program, Department of Obstetrics and Gynecology, Mount Sinai Hospital, b Division of Clinical and Metabolic Genetics, Department of Paediatrics, The Hospital for Sick Children, University of Toronto, Toronto, ON, Canada, c Department of Human Genetics, Radboud University Medical Center, Nijmegen, The Netherlands, d The Centre for Applied Genomics and Program in Genetics and Genome Biology, The Hospital for Sick Children, Toronto, ON, Canada, e Lillehei Heart Institute, University of Minnesota, Minneapolis, MN, USA, f Medical Genetics Department, I˙stanbul Medical Faculty, Istanbul University, Istanbul, Turkey, and g Department of Obstetrics and Gynecology, Mount Sinai Hospital, University of Toronto, Toronto, ON, Canada † These
authors contributed equally.
Key words: autosomal dominant – Robinow syndrome – skeletal dysplasia – WNT5A Corresponding authors: Dr David Chitayat, Department of Obstetrics and Gynecology, The Prenatal Diagnosis and Medical Genetics Program, Room 3292, The Ontario Power Generation Building, 700 University Avenue, M5G 1Z5 Toronto, ON, Canada. Tel.: +416 586 4523; fax: +416 586 4723;
1
Roifman et al. e-mail: [email protected] and Dr Han Brunner, Radboud University Medical Center, Department of Human Genetics, 20 Geert Grooteplein, Nijmegen, Gelderland 6525GA, The Netherlands. Tel.: +312 436 14017; fax: +312 436 68752; e-mail: [email protected] Received 26 January 2014, revised and accepted for publication 7 April 2014
Robinow Syndrome (RS), a rare skeletal dysplasia, is characterized by dysmorphic features resembling a fetal face, mesomelic limb shortening, hypoplastic external genitalia, and renal and vertebral anomalies. Two types of RS have been reported and are distinguished by the severity of their signs and symptoms and by their pattern of inheritance; the severe form of RS has an autosomal recessive mode of inheritance and the milder form is inherited in an autosomal dominant fashion. Since the description of autosomal dominant Robinow Syndrome (ADRS; OMIM 180700) in 1969 by Robinow and colleagues (1), the molecular etiology remained elusive until only recently (2). Person et al (2) proposed the WNT5A gene as the candidate gene for ADRS, with mutations identified in two affected families including the originally described index family. However, the segregation of the gene mutation with the condition in an affected family, although suggestive, could not serve as a proof that this is the causative gene. WNT5A is a member of the Wnt family of proteins, which regulate critical morphogenic events, including embryonic patterning, cell differentiation, growth and migration (3–6). Mutations in the more widely published ROR2 gene, a co-receptor of WNT5A, have been implicated in autosomal recessive RS and Brachydactyly type B. Many cases of ROR2-associated RS have been described (7–9). Since Person et al.’s (2) identification of WNT5A mutations in two cases of ADRS, there has been no reported confirmation of this association. We report three additional families with WNT5A-associated ADRS, which prove that this is the causative gene in ADRS. Clinical reports Family 1
The proband (the mother) presented to the Prenatal Genetics clinic at the age of 37 in her first pregnancy. She was born to 35-year-old parents who were healthy, had normal stature and were non-consanguineous, of German-British-Italian descent. At birth, she was noted to have facial dysmorphism and short upper and lower limbs with normal weight and head circumference. Skeletal X-rays at the age of 4 months showed that all long bones were short and the radial heads were
2
dislocated. The type of her skeletal dysplasia could not be delineated. She was later noted to have dental problems, including persistent primary teeth requiring extraction at 18 years of age, and subsequent misalignment of permanent teeth. She had normal development and cognition. Her pregnancy was initially uncomplicated and there was no history of maternal exposures. NT was measured at 11.5-weeks gestation and was 1.0 mm and screening for Down syndrome and open neural tube defects was screen negative. A detailed fetal ultrasound at 21 weeks gestation suspected intrauterine growth restriction (IUGR) showing that all long bones measured below −2 SD (2 weeks behind gestational age), suggestive of a skeletal dysplasia. The bones appeared well mineralized and there were no fractures. The couple was counseled that the baby had a non-lethal skeletal dysplasia, likely the same diagnosis as his mother, and elected to continue the pregnancy with no invasive testing. The couple is non-consanguineous. The father of this pregnancy is a healthy 37-year-old man who was adopted out after delivery. He is Caucasian but of unknown descent. The mother (i.e. proband) measured 150 cm tall (−2.5 SD) and the father measured 170 cm tall (10th–25th centile). The baby was born at term via Caesarian section for breech presentation. Birth weight was 2360 g (3rd centile), length was 43 cm (−2.5 SD; 7 weeks behind gestational age), and head circumference was 33.2 cm (10th centile). Dysmorphic features were noted including, dolichocephaly, high forehead with frontal bossing, hypoplasia of the supraorbital ridges, prominent eyes, hypertelorism, short flat nose, anteverted nares, low set ears and micrognathia. Many of these features were shared with his mother (Fig. 1). All the four limbs were short; mesomelia was most prominent with upper arm, forearm and hand measurements plotting 1, 12 and 8 weeks behind gestational age, respectively. There was bilateral cryptorchidism and hypoplastic scrotal tissue. Although the penis appeared small due to its embedded position within the scrotum, it measured within normal limits (2.7 cm). Upper limb X-rays revealed short long bones and dislocated radial heads bilaterally (Fig. 2). On the basis of the baby’s combined features of dysmorphism resembling a fetal face, mesomelic limb shortening and genital
De novo WNT5A-associated autosomal dominant Robinow syndrome
Fig. 1. Family 1: The mother in infancy (top row) and her son (middle and bottom rows). Note the high forehead, frontal bossing, hypertelorism, hypoplasia of the supraorbital ridges, depressed nasal bridge, broad nasal root and tip, hypoplastic malar regions, micrognathia and low set ears. Skeletal features include short stature (height G; p.Tyr86Cys) in the WNT5A gene that results in an amino acid exchange predicted to disrupt protein folding and affect Wnt function. This mutation was later confirmed with direct WNT5A gene sequencing. The maternal grandparents did not show the mutation detected in their daughter and
3
Roifman et al.
Fig. 2. Upper limb X-rays of the proband in Family 1 showing short long bones and dislocated radial heads bilaterally.
He was seen again at the age of 11 months and on examination, his length was 73 cm (25th centile) and he was noted to have relatively short limbs (mesomelic). His facial dysmorphism and genital abnormalities remained the same. At the age of 2, he had bilateral orchidopexy. A follow-up at 8 years showed normal development and academic achievements. His height was 120 cm (3rd centile), his weight was 22.2 kg (10th–25th centile) and his head circumference was 49 cm (just below the 3rd centile). His arm span was 111 cm and his span/height was 0.93 (normal 0.99). Facial dysmorphism remained unchanged with a flat face, hypertelorism, prominent eyes, broad mouth and full cheeks. His arms were short with short hands [total length 13 cm (−2.5SD)] and brachydactyly [3rd finger length 5 cm (A, p.Cys69Tyr). Family 3
Fig. 3. Family 1 pedigree showing the affected proband and her son (shaded in black). Whole exome sequencing was performed on affected mother and son and unaffected maternal grandparents (*).
grandson. This confirms that the mutation arose de novo in the proband. Family 2
The proband is the second child of healthy non-consanguineous parents of Dutch–Spanish descent. He has a healthy older sister and the parents and sister are of average stature with normal development and cognition. The pregnancy was uneventful and delivery was at term and uncomplicated; his birth weight was 3530 g (50th centile), and his head circumference was 35.5 cm (25th–50th centile). Although birth length was not measured, length was 51 cm (10th centile) at the age of 4 weeks. At birth, he was noted to have facial dysmorphism with a high and broad forehead with a naevus flammeus, hypertelorism, large prominent eyes, a depressed nasal bridge, short flat nose with anteverted nares, macrostomia and gingival hypertrophy (Fig. 5). He had a micropenis and bilateral cryptorchidism. The differential diagnosis included Opitz G-BBB syndrome and RS but molecular analysis of the MID1 and ROR2 genes showed no detectable mutation.
4
The proband is the first child born to nonconsanguineous parents of Turkish descent. The pregnancy was complicated by short long bones indicating growth restriction, but with normal placental function, identified in the second trimester (precise gestational age unknown). The proband was born at term with a birth weight of 2450 g (3rd centile). On examination by a clinical geneticist at the age of 13 days, her weight was 2700 g (3rd–10th centile), length was 44 cm (3rd centile), head circumference was 32.7 cm (10th centile) and upper/lower segment ratio was 1.31 (G; p.Tyr86Cys) in the WNT5A gene that resulted in an amino acid exchange predicted to disrupt protein folding and affect Wnt function. Top: Graph of filtering criteria used to identify causative variant. Bottom: Integrative Genomics Viewer (14) image showing WNT5A heterozygous variant identified in the affected son (1st row) and his mother (2nd row), but not in the unaffected maternal grandparents (3rd and 4th rows). This mutation was later confirmed with Sanger sequencing. [*Population allele frequencies based on data from 1000 Genomes Project and National Heart, lung and blood institute (NHLBI).]
mutational analysis of the WNT5A gene was performed by Sanger sequencing on the proband and her father and revealed a novel heterozygous mutation c.257A>G (p.Tyr86Cys) in both. Wnt5a homology model
We sought to evaluate whether specificity exists among the mutations identified in all five WNT5Aassociated ADRS-affected families. The threedimensional-structure of the Wnt5a protein is still unknown. Therefore, we built a homology model using the experimentally solved structure of Wnt8 (PDB file
4f0a) from Xenopus Leavis as a template. This structure and human Wnt5 share 44% sequence identity over 300 residues. The YASARA & WHAT IF Twinset was used for modeling and subsequent visualization and analysis (10–12). Although the first 65 residues are lacking, the model covers the four mutated positions. Strikingly, all four mutations seem to be located on one side of the protein (Fig. 7). Discussion
Heterozygous WNT5A mutations have been reported previously in two families with ADRS (2). A comparison
5
Roifman et al.
Fig. 5. Family 2. The proband at different ages showing short stature, high forehead, frontal bossing, hypertelorism, hypoplasia of the supraorbital ridges, depressed nasal bridge, broad nasal root and tip, low set ears and genital hypoplasia.
Fig. 7. Wnt5a homology model. Mutations found in all WNT5Aassociated ADRS cases seem to be located on one side of the protein and may affect interactions with other proteins in the Wnt pathway (10–12).
Fig. 6. Radiographic image of the proband from Family 3 at the age of 13 days.
of their findings along with these cases is summarized in Table 1. WNT5A is a member of the Wnt family of proteins, which regulate critical morphogenic events, including embryonic patterning and cell differentiation,
6
growth and migration (3–5). Wnt5a and Ror2 have been shown to interact physically and function together to activate the noncanonical Wnt signaling pathway (6). Both Wnt5a-null mice and Ror2-null mice exhibit a similar RS-like phenotype, including shortened anterior–posterior axis, dysmorphic facial features and genital hypoplasia, but have a more severe phenotype than human ADRS and include cardiac defects, rib fusions and perinatal lethality (4–6, 13). These findings are more similar to the human ARRS phenotype
Autosomal dominant
Presumed de novo dominanta Autosomal dominant
Short long bones
Short long bones
3 (2 underwent confirmatory molecular testing)
1
8 (7 underwent confirmatory molecular testing)
Current report, Family 3
Person et al. (2010)
Robinow et al. (1969) and Person et al. (2010)
ADRS, autosomal dominant Robinow Syndrome. a Mutation analysis of unaffected parents was not performed to confirm de novo dominant inheritance.
1 Current report, Family 2
Short stature, mesomelic limb shortening, hypertelorism, mandibular hypoplasia, irregular dental alignment, genital hypoplasia.
Short long bones
c.206G>A, p.Cys69Tyr heterozygous c.257A>G, p.Tyr86Cys heterozygous Autosomal dominant
c.248G>C, p.Cys83Ser; heterozygous c.544–545 CT>TC, p.Cys182Arg; heterozygous
c.257A>G, p.Tyr86Cys heterozygous Autosomal dominant
Short long bones; Bilateral radial head dislocation Not performed 2 Current report, Family 1
Mesomelic short stature, limb shortening, malar hypoplasia, hypertelorism, short flat nose, low set ears, limited suppination, persistent primary teeth, brachydactyly, bilateral cryptorchidism and genital hypoplasia (n = 1) Short stature with limb shortening, hypertelorism, short flat nose, brachydactyly, bilateral cryptorchidism and genital hypoplasia Short stature with mesomelic limb shortening, brachydactyly, large anterior fontanel, hypertelorism, prominent eyes with bilateral epicanthic folds, wide down-slanting palpebral fissures, flat midface, short upturned nose, broad nasal bridge, anteverted nares, long philtrum, gingival hyperplasia, short oral frenulum, posteriorly rotated ears, micrognathia., genital hypoplasia, sacral dimple, hairy sacral patch Short stature, mesomelic limb shortening, marked hypertelorism, short nose
Pattern of inheritance Radiographical features Clinical features Number of family members affected Case
Table 1. A summary of the clinical, radiological and molecular analysis findings of our cases and the two other reported cases with WNT5A-associated ADRS
WNT5A molecular diagnosis
De novo WNT5A-associated autosomal dominant Robinow syndrome (OMIM 268310). Humans with homozygous WNT5A mutations have not been described and may be embryonically lethal. The heterozygous WNT5A mutations identified in two families with ADRS by Person et al. (2) were thought to act as hypomorphic alleles, based on functional expression studies. Although heterozygous WNT5A mutations were identified in all affected members of the two families previously described by Person et al. (2), the cases reported inherited the mutation by descent and thus lacked the proof that the gene mutation was the cause of the condition. Families 1 and 2 are the first reports of de novo WNT5A mutations associated with the ADRS in the absence of other likely candidate genes identified by whole exome sequencing. This confirms the WNT5A gene to be the causative gene in ADRS. Interestingly, the probands in Families 2 and 3 exhibited acquired microcephaly, a feature not previously described in RS. In our Wnt5a homology model, all four known ADRS mutations appear to be located on the same side of the protein. It is possible that mutations in these residues change the surface structure and thereby affect interactions with other proteins in the Wnt pathway. Residues C69 and C83 are modeled as free cysteines. Their specific role in multimerization and/or stabilization of the complex is unknown. However, mutation of these residues into Tyrosine and Serine, respectively, could affect this putative activity. Residue Y86 is also located on the surface. Its side chain is big and aromatic and could play a role in complex formation. The mutation converts this residue into Cysteine, which is smaller and cannot make the same interactions. Besides the loss of the Tyrosine side chain, this mutation might also introduce a new cysteine-bond with the free Cysteine at position 83. In our model, residue C182 makes a disulfide bond with residue C164. This will stabilize the local domain and the interaction surface. Mutation C182R will disturb this disulfide bond. In addition, a bigger side chain is introduced at this position, which will change the surface of the protein. Conclusions
We report three families with WNT5A-associated ADRS, caused by novel heterozygous WNT5A mutations confirmed, in the first family by whole exome, and in all by Sanger sequencing. Families 1 and 2 are the first cases showing de novo inheritance in the affected family members and thus strengthen the evidence for WNT5A as the causative gene in ADRS. We propose a specificity to the ADRS WNT5A genotype since all five families reported to date have missense mutations affecting the same part of the Wnt5a protein, and two of these harbor the identical nucleotide change. Finally, we note that all RS families with WNT5A mutations have a classical AD Robinow phenotype, whereas several less typically affected patients (with short-limbed dwarfism without a mesomelic pattern, with less typical facial dysmorphism and/or absence of genital hypoplasia) tested in our diagnostic laboratory do not show such mutations. Thus,
7
Roifman et al. both genotype and phenotype of WNT5A-associated RS are likely to be highly specific. Acknowledgements Whole exome sequencing was funded by the Government of Canada through Genome Canada, the Canadian Institutes of Health Research and the Ontario Genomics Institute (OGI-049). Additional funding was provided by Genome Quebec, Genome British Columbia, and the McLaughlin Centre. Sanger sequencing was performed by the Oxford Medical Genetics Laboratory at the Churchill Hospital in Oxford, UK in Family 1, and by the Genome Diagnostics laboratory of the Department of Human Genetics in Nijmegen, The Netherlands in Families 2 and 3.
References 1. Robinow M, Silverman FN, Smith HD. A newly recognized dwarfism syndrome. Am J Dis Child 1969: 117 (6): 645–651. 2. Person AD, Beiraghi S, Sieben CM et al. WNT5A mutations in patients with autosomal dominant Robinow syndrome. Dev Dyn 2010: 239 (1): 327–337. 3. Nusse R. Wnt signaling in disease and in development. Cell Res 2005: 15 (1): 28–32. 4. Yamaguchi TP, Bradley A, McMahon AP, Jones S. A Wnt5a pathway underlies outgrowth of multiple structures in the vertebrate embryo. Development 1999: 126 (6): 1211–1223.
8
5. Schleiffarth JR, Person AD, Martinsen BJ et al. Wnt5a is required for cardiac outflow tract septation in mice. Pedriatr Res 2007: 61 (6): 386–391. 6. Oishi I, Suzuki H, Onishi N et al. The receptor tyrosine kinase Ror2 is involved in non-canonical Wnt5a/JNK signalling pathway. Genes Cells 2003: 8 (7): 645–654. 7. van Bokhoven H, Celli J, Kayserili H et al. Mutation of the gene encoding the ROR2 tyrosine kinase causes autosomal recessive Robinow syndrome. Nat Genet 2000: 25 (4): 423–426. 8. Afzal AR, Rajab A, Fenske CD et al. Recessive Robinow syndrome, allelic to dominant brachydactyly type B, is caused by mutation of ROR2. Nat Genet 2000: 25 (4): 419–422. 9. Afzal AR, Jeffrey S. One gene, two phenotypes: ROR2 mutations in autosomal recessive Robinow syndrome and autosomal dominant brachydactyly type B. Hum Mutat 2003: 22 (1): 1–11. 10. Krieger E, Koraimann G, Vriend G. Increasing the precision of comparative models with YASARA NOVA–a self-parameterizing force field. Proteins 2002: 47 (3): 393–402. 11. Vriend G. WHAT IF: a molecular modeling and drug design program. J Mol Graph 1990: 8 (1): 52–56. 12. Krieger E, Joo K, Lee J et al. Improving physical realism, stereochemistry, and side-chain accuracy in homology modeling: four approaches that performed well in CASP8. Proteins 2009: 77 (9): 114–122. 13. Schwabe GC, Trepczik B, Süring K et al. Ror2 knockout mouse as a model for the developmental pathology of autosomal recessive Robinow syndrome. Dev Dyn 2004: 229 (2): 400–410. 14. Thorvaldsdottir H, Robinson JT, Mesirov JP. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform 2013: 14 (2): 178–192.
