Article Text
PDF
Abstract
Background Ellis-van Creveld syndrome (EvC) is a recessive disorder characterised by acromesomelic limb shortening, postaxial polydactyly, nail-teeth dysplasia and congenital cardiac defects, primarily caused by pathogenic variants in EVC or EVC2. Weyers acrofacial dysostosis (WAD) is an ultra-rare dominant condition allelic to EvC. The present work aimed to enhance current knowledge on the clinical manifestations of EvC and WAD and broaden their mutational spectrum.
Methods We conducted molecular studies in 46 individuals from 43 unrelated families with a preliminary clinical diagnosis of EvC and 3 affected individuals from a family with WAD and retrospectively analysed clinical data. The deleterious effect of selected variants of uncertain significance was evaluated by cellular assays.
Main results We identified pathogenic variants in EVC/EVC2 in affected individuals from 41 of the 43 families with EvC. Patients from each of the two remaining families were found with a homozygous splicing variant in WDR35 and a de novo heterozygous frameshift variant in GLI3, respectively. The phenotype of these patients showed a remarkable overlap with EvC. A novel EVC2 C-terminal truncating variant was identified in the family with WAD. Deep phenotyping of the cohort recapitulated ‘classical EvC findings’ in the literature and highlighted findings previously undescribed or rarely described as part of EvC.
Conclusions This study presents the largest cohort of living patients with EvC to date, contributing to better understanding of the full clinical spectrum of EvC. We also provide comprehensive information on the EVC/EVC2 mutational landscape and add GLI3 to the list of genes associated with EvC-like phenotypes.
- Congenital, Hereditary, and Neonatal Diseases and Abnormalities
- Human Genetics
- Molecular Medicine
- Pediatrics
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- Congenital, Hereditary, and Neonatal Diseases and Abnormalities
- Human Genetics
- Molecular Medicine
- Pediatrics
WHAT IS ALREADY KNOWN ON THIS TOPIC
Ellis-van Creveld syndrome (EvC) is an autosomal recessive skeletal dysplasia primarily caused by variants in EVC or EVC2, which encode ciliary proteins that are positive mediators of Hedgehog (Hh) signalling.
A minority of EvC cases lack mutations in EVC or EVC2 and are associated with variants in other components of the Hh pathway.
Weyers acrofacial dysostosis (WAD) is a dominant condition allelic to EvC. Until now, four convincing WAD mutations have been described in the literature, all located at the 3′-end of EVC2.
WHAT THIS STUDY ADDS
This study represents the largest cohort of living patients with molecularly confirmed EvC up to date, improving the understanding of the full clinical spectrum of EvC/EvC-like phenotypes.
In 46 new patients with EvC, we identified 47 different variants in EVC, EVC2, WDR35 and GLI3, and ascertained the pathogenicity of selected variants of unknown significance in EVC, EVC2 and WDR35 through functional studies.
We also described a novel WAD-causing EVC2 variant in a multiplex family with three patients showing vertical transmission.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
We present functional assays for validation of variants of uncertain significance in EVC/EVC2.
We provide a basis to place GLI3 among the genes causing EvC phenotypes and further support for the association of WDR35 splicing mutations with an EvC phenotype.
Our extensive molecular, clinical and functional data will help refine the spectrum of EvC/EvC-related phenotypes and aid diagnosis and surveillance in the clinical setting.
Introduction
Ellis-van Creveld syndrome (EvC; MIM# 225500), also known as chondroectodermal dysplasia, is an autosomal recessive ciliary chondrodysplasia recognisable by a distinctive pattern of clinical features. Affected individuals have a skeletal phenotype characterised by disproportionate short stature with acromesomelic shortening of the limbs, postaxial polydactyly and narrow thorax at birth.1 Dysplastic pelvis observed in infancy usually normalises during childhood, along with the amelioration of the narrow thorax. Inability to make a tight fist and genu valgum are additional common findings.2
Congenital heart defects (CHDs) are among the hallmarks of the syndrome, occurring in approximately 60–70% of the patients.1 3 Atrioventricular septal defects (AVSDs) and atrial septal defects (ASDs), particularly common atrium, account for more than half of the EvC cases with CHDs.4 5
Ectodermal dysplasia, in particular nail and teeth dystrophy, is an almost invariable presentation which allows to differentiate EvC from other chondrodysplasias.6 Numeric, structural and morphological teeth anomalies in EvC include natal teeth, congenitally missing teeth, supernumerary teeth, premature loss, enamel hypoplasia, conical teeth and microdontia.7 8
Intraoral findings in EvC have been well defined. Soft tissue findings such as accessory, wide or otherwise aberrant oral frenulae, labiogingival adherences, alveolar notching or serrated alveolar appearance of the mandibular incisor region and submucosal clefts were reported.8 9
EvC is mainly caused by biallelic loss-of-function variants in EVC or EVC2, two adjacent genes located on 4p16 encoding the two subunits of the EvC ciliary complex.10 11 EVC and EVC2 function as positive mediators of Hedgehog (Hh) signalling and form a mutually protective protein complex that localises to the ciliary base.12–14 The EvC complex interacts with Smoothened, frizzled class receptor (SMO), the main activator of the Hh pathway, and regulates the dissociation of SUFU-GLI3 inhibitory complexes as well as the traffic of GLI3 to the distal end of the cilium after pathway activation.12 15 16
Weyers acrofacial dysostosis (WAD; MIM# 193530) is an autosomal dominantly inherited disorder associated with truncating variants in the final exon of EVC2.2 17–19 Patients with WAD present with polydactyly and ectodermal defects but normal stature. Variants associated with WAD code for EVC2 proteins lacking a C-terminal motif, indispensable to interact with another ciliary base complex formed by EFCAB7 and IQCE.20 As a result, EvC complexes containing EVC2 WAD proteins localise along the length of the cilium instead of being concentrated at its base. Causative variants for both EvC and WAD result in impaired Hh signalling.12 15
Previous studies show that mutations in EVC or EVC2 account for 70% to 100% of the affected individuals, indicating genetic heterogeneity.17 19 21 22 In agreement with this, biallelic variants in WDR35, GLI1, DYNCL2L1, DYNC2H1 and SMO, and heterozygous germline or mosaic mutations in PRKACA and PRKACB have been reported in a limited number of individuals with EvC-related phenotypes.22–27 Notably, all these genes play a role in Hh signalling.
Here, we studied a cohort of 46 individuals from 43 unrelated families with a clinical diagnosis of EvC, and 1 family with WAD with 3 affected individuals, to provide detailed clinical characterisation with emphasis on less common phenotypical findings and expand the knowledge on the mutation spectrum accountable for these diseases.
Material and methods
Clinical assessment
Patients with a clinical diagnosis of EvC and WAD were recruited through an international collaboration between 2010 and 2020. Polydactyly and intraoral findings of two affected patients with WAD were previously reported.28 The clinical diagnosis of EvC was established when postaxial polydactyly was present in combination with at least two other well-described features6: short stature with short limbs and/or brachydactyly; ectodermal findings (nail and/or teeth dysplasia); characteristic facial or intraoral findings (midline notching of upper lip and/or frenular/alveolar abnormalities); characteristic CHDs (AVSD or ASD). Informed written consent was obtained from all subjects or their legal guardians in accordance with the Declaration of Helsinki principles.
Molecular techniques
A diagram illustrating the stepwise molecular analysis of the cohort is shown in figure 1. Briefly, EvC probands were analysed for EVC (NM_153717.3) and EVC2 (NM_147127.5) variants by Sanger sequencing, except an individual (208.1) who was directly tested with exome sequencing (ES). Cases negative for pathogenic variants in EVC/EVC2 were subjected to multiplex ligation-dependent probe amplification (MLPA) analysis, homozygosity mapping by SNP-array hybridisation or ES. Homozygosity mapping was performed when consanguinity was present, and Sanger sequencing of a candidate gene (WDR35, NM_001006657.2) was conducted in one proband (170.1) who was not homozygous for the 4p16 EvC region. MLPA and SNP arrays were used to investigate large deletions and duplications in genomic DNA. ES was performed in another subject born to unrelated parents (169.1), after exclusion of EVC/EVC2 mutations, and Sanger sequencing was used for confirmation of the candidate variant found in GLI3 (NM_000168.6). In the family with WAD, only EVC2 exon 22 was analysed by Sanger sequencing. RT-PCR, western blot, immunofluorescence or a minigene assay was used to demonstrate the pathogenicity of variants of unknown significance (VUS) with possible effects on mRNA splicing and to evaluate the protein effect of a large intragenic duplication. Additional methods are provided as online supplemental information.
Supplemental material
Overview of the molecular testing procedure used in this study. Workflow diagram illustrating the molecular analyses conducted in probands. Individuals that resulted negative for pathogenic mutations in EVC or EVC2 by Sanger sequencing were examined through whole-genome SNP-array hybridisation, MLPA or ES. The pathogenicity of specific splice site mutations of uncertain significance was determined by RT-PCR in patient samples or through a minigene assay. Proband 208 was directly screened by ES. ES, exome sequencing; MLPA, multiplex ligation-dependent probe amplification.
Results
Patients and genetic analyses
This study includes 46 individuals with EvC from 43 unrelated families and 3 affected individuals from a family with a previous clinical diagnosis of WAD.28 Two fetal EvC cases were examined after termination of the pregnancy at 30 and 19 weeks of gestation. The age of assessment of the remaining EvC cases varied from birth to 63 years. Parental consanguinity was reported in 34 out of 43 (79%) families with EvC, including 3 families with 2 affected siblings. Nineteen (41.3%) patients with EvC were males and 27 (58.7%) were females.
Genetic analysis in the 43 families with EvC identified biallelic pathogenic variants in EVC or EVC2 in 41 (95.34%), a homozygous WDR35 variant in one and a heterozygous GLI3 variant in another. Affected individuals of the family with WAD had a heterozygous EVC2 variant. A list of the identified variants is shown in table 1.
Genetic variants identified in 46 individuals with clinical diagnosis of EvC and 3 with WAD
EVC/EVC2 variants
We detected 45 different EVC/EVC2 pathogenic variants in patients with EvC: 24 in EVC and 21 in EVC2, distributed along the coding sequence of both genes (table 1; figure 2A). One EVC variant (c.1694delC) was found in compound heterozygous state in two unrelated individuals (110.1 and 127.1), and one homozygous EVC2 variant (c.2092C>T) was recurring in two families (199.1 and 203.1). EVC variants included one missense variant (4.17%), two large duplication/deletions (8.33%), two in-frame microdeletions (8.33%), five nonsense (20.83%), seven frameshift (29.17%) and seven splicing variants (29.17%) (figure 2B). EVC2 variants included one missense variant (4.76%), two large deletions (9.52%), five frameshift (23.81%), five splicing (23.81%) and eight nonsense variants (38.10%) (figure 2B).
Schematic representation and classification of the EVC and EVC2 variants identified in this study. (A) Schematic representation of the exon structure of EVC (21 exons) and EVC2 (22 exons) indicating the position of the different variants detected in this work. The WAD variant in exon 22 of EVC2 is labelled in red. (B) Percentages of the various types of the EVC and EVC2 variants found in this study. The large majority of mutations in both genes are loss-of-function variants. Recurrent variants are counted once. EvC, Ellis-van Creveld syndrome; WAD, Weyers acrofacial dysostosis.
Analysis of missense variants in EVC/EVC2
The two missense variants of this study, c.919T>C; p.(Ser307Pro) in EVC and c.737C>A; p.(Ala246Asp) in EVC2, had CADD1.6 scores of 15.3 and 23.8, respectively (UCSC Genome Browser; GRCh38/hg38).29 30 The p.(Ser307Pro) variant was previously observed in several other EvC cases17 21 22 and is classified as likely pathogenic (PS4, PM2 as supporting and PM3) according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) 2015 variant interpretation criteria.31 32 The p.(Ala246Asp) change was classified as a VUS (PM2, PM3 and PP3). To clarify the pathogenicity of the Ala246Asp substitution, we carried out phenotype rescue experiments using Evc2-/- mouse embryonic fibroblasts (MEFs) in which we introduced either the wild-type version of human EVC2 or the Ala246Asp variant through retroviral delivery. Evc2-/- cells were previously demonstrated to have neither EVC2 nor EVC in the cilia because the two proteins form an interdependent protein complex.12 Immunofluorescence analysis on the retrotransduced cells showed that, unlike the wild-type EVC2 protein, expression of the Ala246Asp variant in Evc2-/- MEFs was not able to restore the presence of the endogenous mouse EVC protein at the base of cilia. The Ala246Asp EVC2 variant was thus reclassified as likely pathogenic with the addition of ACMG/AMP PS3 evidence (online supplemental figure 1).
The three affected individuals with WAD were found with a novel heterozygous truncating variant in exon 22 of EVC2: c.3751G>T; p.(Glu1251*). This variant is expected to result in an EVC2 protein lacking the last 58 amino acids (table 1; figure 2A; online supplemental figure 2).
Analysis of splicing variants in EVC/EVC2
Three homozygous variants, EVC c.2304G>A; p.(Lys768Lys) in patient 166.1, EVC2 c.1470+3A>T in patient 176.1 and EVC2 c.2830–30G>A in patient A479, which were initially considered as VUS, were predicted to affect splicing by in silico analysis with the NNSPLICE 0.9 splice site predictor program.33 To test this hypothesis, we carried out studies in patient samples. RT-PCR amplification and sequencing of a cDNA fragment spanning from exon 12 to exon 19 of EVC in 166.1 primary fibroblasts demonstrated that the c.2304G>A synonymous variant, which occurs in the last nucleotide of EVC exon 15, leads to the skipping of this exon and thus to the in-frame deletion of 69 amino acids (figure 3A). Consistently, a protein of lower molecular weight than the wild type was detected by anti-EVC immunoblotting in patient 166.1 (figure 3B).34 Immunofluorescence analysis also showed that only 50% of 166.1 cells had EVC/EVC2 in the cilium, compatible with the mutual dependence of EVC and EVC2 on each other for ciliary localisation (figure 3C,D).12 Pathogenicity of the EVC2 intron 10 donor splice site variant c.1470+3A>T was evaluated by conducting RT-PCR between exons 8 and 12 of EVC2 in fibroblasts from patient 176.1. Sequencing of the amplified product revealed an 8-nucleotide deletion due to activation of a cryptic 5′ splice site within EVC2 exon 10 (figure 3E). This deletion causes a frameshift and the early truncation of the protein (p.Gly488Valfs*46). Accordingly, EVC and EVC2 were not detected in the cilia of 176.1 fibroblasts (figure 3C). EVC protein levels, which are known to be diminished in the absence of EVC2,12 were also found to be lower in 176.1 cells than in control cells (figure 3B). Lastly, we analysed the EVC2 c.2830–30G>A intron 16 variant by RT-PCR amplification of an EVC2 fragment spanning from exon 12 to exon 22 in peripheral blood of patient A479 and a control. We sequenced both PCR products and found the inclusion of 28 extra nucleotides of intron 16 between exons 16 and 17 in the patient transcript due to usage of a new splice site acceptor created by the mutation. The insertion of these nucleotides also results in a frameshift and a premature termination codon (p.Arg945Trpfs*12) (figure 3F).
Functional evaluation of variants. (A) RT-PCR amplification of an EVC cDNA fragment encompassing the c.2304G>A variant in fibroblasts from patient 166.1 and a control individual. A smaller size PCR product was obtained in 166.1 cells compared with control fibroblasts. Sanger sequencing chromatogram of this product is on the right, demonstrating exon 15 exclusion and abnormal junction between exons 14 and 16. (B) Representative anti-EVC immunoblot of protein extracts from 176.1, 136.1 and 166.1 skin fibroblasts. Extracts from normal control fibroblasts, as well as from fibroblasts lacking EVC (EVC-/-) or EVC2 (EVC2-/-) reported earlier,34 were also included. Similar to EVC2-/- cells, EVC protein levels in 176.1 fibroblasts are decreased with respect to normal control cells, indicating absence of EVC2 and loss of the EVC-EVC2 mutually protective effect. Normal size EVC protein is marked with a black arrow. In the case of 136.1, who had a homozygous in-frame duplication of exons 3 to 11 of EVC, a high molecular weight band was detected (top arrowhead). In 166.1 fibroblasts, an EVC protein (lower arrowhead) smaller than the wild type was observed due to in-frame skipping of exon 15. n=3. *Non-specific bands. Tubulin (TUB) was used as loading control. (C) Representative shifted-overlay immunofluorescence images of EVC and EVC2 in the primary cilia of normal control and patient primary fibroblasts showing that both proteins are absent in the cilia from 176.1 and 136.1 cells. In fibroblasts from 166.1, only a fraction of the cilia was positive for EVC and EVC2. Red: EVC (top panel); EVC2 (lower panel); green: acetylated-TUB (AcTUB, ciliary axoneme)+ɣ-TUB (basal body/centrioles); blue: nuclei. Scale bars: 10 µm; n=2. (D) Quantification of the ciliary presence of EVC (left diagram) and EVC2 (right diagram) in 166.1 and control fibroblasts. For each culture, n=100 cilia identified by AcTUB immunostaining from two independent experiments were analysed. Data are mean±SD. p<0.05; **p<0.01. Student’s t-test. (E) Sanger sequencing chromatograms of an RT-PCR product spanning from exon 8 to exon 12 of EVC2 amplified from fibroblasts of proband 176.1 and a control. The c.1470+3A>T variant in 176.1 activates an upstream cryptic donor splice site placed within exon 10 (underlined nucleotides in the control sequence) causing the deletion of the last eight nucleotides of this exon. (F) Sanger sequencing chromatogram of an EVC2 RT-PCR fragment from exon 12 to exon 22 amplified from whole-blood cDNA of patient A479 demonstrating inclusion of 28 nucleotides of intron 16 between exons 16 and 17 due to a new acceptor splice site created by the c.2830–30G>A mutation. EvC, Ellis-van Creveld syndrome.
Large insertions and deletions (indels) in EVC/EVC2
Large EVC/EVC2 indels detected in our cohort comprise a homozygous deletion of EVC exon 7 (c.802-?_939+?del) in patient 138.1; a homozygous deletion of EVC2 exons 4–6 (c.451-?_816+?del) in patient 205.1; a maternally inherited heterozygous deletion of EVC2 exons 6 to 10 (c.707-?_1470+?del) in compound heterozygosity with c.737C>A; p.(Ala246Asp) in patient 159.1; and a homozygous duplication of EVC from exon 3 to exon 11 (c.301-?_1563+?dup) in patient 136.1. The effect of the in-frame duplication of exons 3 to 11 of EVC was explored by anti-EVC immunoblotting and immunofluorescence in 136.1 fibroblasts. We observed an extra-large EVC protein in the cells of this patient, which did not localise to the primary cilia (figure 3B,C).
A novel splicing mutation in WDR35 in a patient with an EvC-like phenotype
SNP-array hybridisation showed that patient 170.1, born to consanguineous parents, was not homozygous at the EVC/EVC2 region but presented an 8.8 Mb block of homozygosity on chromosome 2 encompassing WDR35. Sanger sequencing of WDR35 coding exons and flanking intronic regions revealed a novel, homozygous five-nucleotide intronic deletion, located 10 nucleotides upstream of the acceptor splice site of exon 15 (c.1504–11_1504-7delATTTA), which was not listed in gnomAD V.2.1.1. We investigated the pathogenicity of this variant using a minigene splicing assay due to unavailability of patient material. This experiment showed inclusion of exon 15 into the minigene transcript in the cells transfected with the wild-type minigene but not in the cells transfected with the minigene construct carrying the c.1504–11_1504-7delATTTA variant, indicating pathogenicity (figure 4). Exclusion of exon 15 would result in the deletion of 18 evolutionary conserved amino acids of the WDR35-encoded protein (IFT121).
Minigene assay proving pathogenicity of the WDR35 intron variant identified in patient 170.1. (A) Schematic representation of pSPL3/WDR35 hybrid minigenes showing the two artificial exons (ExA and ExB) of pSPL3 along with WDR35 exon 15 and adjacent intronic sequences inserted into the pSPL3 multicloning site with EcoRI and XhoI. SDv and SAv indicate ExA and ExB donor and acceptor splice sites, respectively. The 170.1 intronic mutation is depicted in red and the dotted lines represent splicing events occurring in the wild-type minigene construct. (B) Representative agarose gel image showing the RT-PCR products amplified using SD6 and SA2 primers in hTERT RPE-1 cells transfected with the empty vector (pSPL3) or with the wild-type or mutant minigene (n=3). Non-transfected cells and a non-template negative control (NC) were also included. Schematic representation of the PCR products from each transfection illustrating exon composition and product sizes is on the right.
Identification of a de novo mutation in GLI3 in a patient with an EvC-like phenotype
ES analysis in patient 169.1 revealed a heterozygous truncating variant in the last exon of GLI3, c.3299dupT; p.(Leu1100Phefs*29), placed within the first half of the C-terminal transcriptional activator domain of the GLI3 protein.35 Sanger sequencing in proband 169.1 and his parents indicated de novo occurrence of the mutation.
Clinical characteristics of EvC individuals with EVC and EVC2 variants
Detailed clinical features observed in 44 patients with EvC with EVC/EVC2 pathogenic variants are described in online supplemental table 1. Facial, oral, radiological and skeletal findings of selected individuals are illustrated in figure 5.
Supplemental material
Facial, oral, radiological and skeletal findings of individuals with EvC from the present study. (A–K) Clinical photographs and radiological images illustrating EvC findings in this cohort: (A) Midline notching of the upper lip with long philtrum. (B) Widely spaced teeth, missing upper lateral incisors, shallow upper labiogingival sulcus with multiple frenulae. (C) Multiple frenulae, alveolar notching, delayed eruption of teeth with microdontia. (D) Midline diastema, tapering teeth, bifid crown of the left central incisor. (E) Postmortem babygram at the 20th gestational week showing narrow thorax, short and horizontally placed ribs, narrow sacrosciatic notch, disproportionately shortened tubular bones and type A postaxial polydactyly. (F) Disproportionate short stature with acromesomelic shortening of the limbs, markedly narrow thorax, protuberant abdomen, postaxial polydactyly of the hands and large capillary malformation on the facial midline. (G) Genu valgum and postaxial polydactyly. (H) Pelvic X-ray showing coxa valgae. (I) Sparse scalp hair, high forehead, laterally sparse eyebrows, strabismus, smooth philtrum and prominent ears. (J) Postaxial type A polydactyly, brachydactyly, hypoplastic and dysplastic nails. (K) brachydactyly, hypoplastic and dysplastic nails, and position anomalies of toes.
Natal and neonatal history
Excluding the two fetal cases, 95.2% (40/42) of the individuals were born between 38 and 42 weeks of gestation, whereas the remaining 4.8% (2/42) were born late preterm at 37 weeks of gestation. Neonatal respiratory problems necessitating specific management were evident in 24.4% (10/41) immediately after birth. While 70% (7/10) of patients with neonatal breathing problems had narrow thorax, 73.1% (19/26) of the patients having a small chest showed no breathing problems.
Growth
SD corresponding to mean birth weight of the cohort was −1.18 SD±0.45. Length at birth was below −2 SD in 47.3% (9/19) of the patients with length recorded at birth. Most individuals older than 3 months of age (26/36, 72.2%) demonstrated short stature (≤−2 SD), and 58.3% (21/36) had a height lower than −2.5 SD. Only one patient had a height above the mean. Adult height was evaluated in four patients (range 18.5 to 63 years), with a mean SD value of −4.03 SD (range −3.27 to −5.55 SD). Three patients demonstrated growth hormone deficiency.
Skeletal abnormalities
All patients had postaxial polydactyly, involving the hands in 97.7% (43/44) patients and the feet in 40.9% (18/44). Polydactyly was almost always bilateral, except two patients who had unilateral polydactyly of the feet. Of the 42 patients, 28 (66.7%) had narrow thorax. Failure to form a fist was observed in 85.7% (30/35). Subjects older than 2 years were assessed for genu valgum, which was present in 70% (21/30). Scoliosis was observed in 20.5% (8/39). No patients required orthoses or surgery for scoliosis.
Cardiac anomalies
Out of 44 patients, 43 underwent echocardiographic screening, including the two fetal cases evaluated by detailed antenatal ultrasound. Out of 43 patients, 30 (69.8%) had congenital heart anomalies. The most common cardiac malformation was AVSD (16/43, 37.2%) and ASD (11/43; 25.6%), followed by patent ductus arteriosus (6/43, 14%), mitral valve diseases including congenital cleft (6/43, 14%), common atrium (5/43, 11.6%), patent foramen ovale (PFO) (2/43, 4.7%), double superior vena cava (2/43, 4.7%) and abnormalities of the aortic arch including narrowing and hypoplasia (2/43, 4.7%). Tetralogy of Fallot, single ventricle with situs ambiguous, pulmonary hypertension and tricuspid regurgitation were observed in a single case each.
Facial and intraoral findings
We observed several dysmorphic features such as hypertelorism, upslanting palpebral fissures, smooth philtrum and low-set ears; however, no recognisable facial appearance was noted. Most patients (84%, 37/44) had at least one intraoral finding: midline notching of the upper lip, accessory or aberrant oral frenulae, and alveolar notching or serrated gingival appearance, which was observed in 55.8% (24/43), 67.4% (29/43) and 72.1% (31/43) of the patients, respectively. Bifid and long uvula were present in a single case each.
Ectodermal findings
Natal teeth were observed in 29.3% (12/41) of the patients. Delayed eruption of teeth, hypodontia/oligodontia and conical teeth/microdontia were present in 60.6% (20/33), 68.6% (24/35) and 62.9% (22/35), respectively. One patient was reported to have a bifid crown of the central maxillary incisor, and another presented with enamel hypoplasia.
Most patients (39/44, 88.6%) demonstrated nail dysplasia. Sparse hair with or without sparse eyebrows was also a frequent finding, affecting 54.8% (23/42).
Global development and structural central nervous system anomalies
Out of 41 patients, 5 (12.2%) had a history of global developmental delay (GDD) and/or intellectual disability (ID), mild to moderate in 3 and severe in 2. Patient 137.1 with severe GDD, hypotonia and swallowing difficulties had hypoplastic cerebellar hemispheres and severe congenital hydrocephalus. She was treated with a ventriculo-peritoneal shunt and followed up until age 5 years, showing only mild delay in acquisition of language and social skills. One patient had mild callosal dysgenesis and mild to moderate GDD/ID, and another presented with seizures and epileptic activity on electroencephalogram.
Other congenital anomalies
Out of 44 individuals, 5(11.4%) had congenital abnormalities of the urinary tract or kidneys, including renal fusion anomalies, bilateral grade I urinary stasis and unilateral renal enlargement. One patient had grade II vesicoureteral reflux with normal renal function, lipoma of the phylum terminale and tethered cord, necessitating spinal cord release at the age of four. Unilateral optic pit and hypoplastic labia minora were detected in a single patient each.
Clinical features of patients with WAD
The three patients with WAD in this study are a mother and her two daughters (online supplemental figure 2A). Clinical features observed in all three were four-limb polydactyly and hypoplastic nails, whereas short stature, narrow thorax, mild arm-span shortening, brachydactyly, aberrant oral frenulae were inconsistently present. Clinical features are further documented in online supplemental table 1.
Clinical presentation of the patient with a homozygous WDR35 variant
Patient 170.1 was born at term with a normal weight and length for gestational age. He had transient neonatal respiratory difficulty, but thoracic circumference was normal. At 3 months of age, his weight was 7060 g (1.32 SD) and length was 54 cm (−2.92 SD). Physical examination revealed smooth philtrum, natal teeth, postaxial type A polydactyly of the hands and feet, brachydactyly, smooth finger flexion creases, difficulty forming a fist and hypoplastic nails. Echocardiography showed PFO (online supplemental table 1).
Clinical presentation of the patient with a heterozygous GLI3 variant
Patient 169.1 was born at the 40th gestational week to non-consanguineous parents with a weight appropriate for gestational age and length of 40 cm (−5.23 SD). At 10 months of age, weight was 7900 g (−2.04 SD) and length was 62 cm (−4.48 SD). On physical examination, he had sparse hair, midline notching of the upper lip, conical teeth, cleft palate and postaxial type A polydactyly of the hands. Thoracic circumference was normal. Developmental milestones were achieved on time. Echocardiography showed PFO (online supplemental table 1).
Discussion
Large EvC cohorts with detailed clinical and molecular data are scarce due to the rarity of the disease. A study by Hills et al only focused on the cardiac phenotype in a cohort of 32 new, molecularly non-confirmed patients.5 In 2022, Aubert-Mucca et al reported 50 individuals from 45 families, but more than 60% of the cases were fetal cases analysed after termination of pregnancy or cases who died prematurely after birth.22 Our study represents the largest series of living individuals harbouring pathogenic EVC/EVC2 variants, providing valuable phenotypic information for full comprehension of EvC.
This cohort includes a consecutive case series of patients with EvC who underwent molecular testing between 2010 and 2020. During the same period, as part of collaborative work with other laboratories, we reported seven patients with EvC with variants in WDR35, GLI1, PRKACA and PRKACB.23–25 If these patients had been included in the current cohort, the proportion of EvC attributable to variants in EVC/EVC2 would have decreased from 95.34% to 82.0%. In any case, there would be no mutation-negative patients. Taken together with previous literature, our results suggest that little of the genetic heterogeneity of EvC remains unexplored.
Clinical features associated with EVC/EVC2 variants
The phenotypic analysis of our cohort underlines the following features as the most consistent clinical findings of the EVC/EVC2 spectrum: postaxial polydactyly of the hands and feet, short stature, short arm span, dysplastic nails, brachydactyly, alveolar notching or serrated gingival appearance, abnormal oral frenulae, ASD or AVSD, genu valgum, narrow thorax and dentition abnormalities. Although a recognisable dysmorphic pattern was not present among patients, smooth philtrum was relatively common (61.9%). Genu valgum was observed in 70% and scoliosis in 20.5% of patients, significantly more frequent than previously reported, likely due to our cohort’s high median age. Uncommon findings observed in multiple individuals include intrauterine growth restriction, growth hormone deficiency, microcephaly, macroglossia, partial syndactyly of toes and long uvula. When interpreting the association of these rare findings with EvC, it is necessary to consider the possible presence of additional genetic factors that would not be detected by the targeted molecular approach used in our cohort.
In line with previous reports, 69.8% of the EVC/EVC2 cohort had congenital heart anomalies, whereas the prevalence of EvC-associated cardiac septation defects showed differences compared with the literature. AVSD, common atrium and ASD, findings which have been reported in 48%, 42% and 15% of patients with EvC,5 were only present in 34.9%, 11.6% and 25.6% of our cases, respectively. Postnatal mortality in our cohort was also lower than previously described, with only one patient dying at 2 months of age due to cardiac complications of situs ambiguous.5 22 36 These discrepancies between our study and previous literature may be due to patient selection criteria and limited follow-up time. Nineteen per cent of our patients were below 12 months of age at last examination, and none of our patients were recruited through cardiac surgery or pathology department, which could partly explain the lower prevalence of severe cardiac defects and discordances with previous literature in postnatal mortality.
Patient 200.1 had single ventricle with situs ambiguous. Situs abnormalities, which are well described in various ciliopathies such as primary ciliary dyskinesia, were reported to co-occur with the EvC phenotype in three patients in the form of situs inversus.37 38 To the best of our knowledge, patient 200.1 is the only patient with molecularly diagnosed EvC with situs ambiguous, suggesting that left-right asymmetry defects can be part of the cardiac phenotypic spectrum of EvC.
Among the dental defects observed in this study, teeth number alterations were the most common in patients with EVC/EVC2 mutations (62%), in agreement with a recent review of oral manifestations of EvC.8 Natal teeth were observed in 29.3% of the present cohort, as compared with 9% in the literature.8 Bifid central maxillary incisor and lower alveolar ridge cleft were previously undescribed features, present in one patient each.
ID and central nervous system anomalies including occipital encephalocele, hydrocephalus, cortical heterotopias, callosal anomalies, Dandy-Walker malformation and cerebellar vermis abnormalities have been infrequently reported in patients with EvC.39–42 In our study, 12.2% of EVC/EVC2 patients displayed GDD/ID, and two of them had structural brain anomalies: hydrocephalus with underdeveloped cerebellar hemispheres and callosal dysgenesis, respectively. Multiple Mendelian disorders are diagnosed in up to 8.5% of patients after positive postnatal exome analysis.43 As combined phenotypes in our patients with neurological abnormalities were not excluded through ES, further studies are required to include developmental delay, ID or structural central nervous system abnormalities as part of the EvC phenotype.
Challenging the previous view that there are no phenotypic differences between patients with biallelic EVC and EVC2 variants or truncating and missense alterations in these genes, a recent systematic review of molecularly confirmed EvC cases suggested novel genotype-phenotype correlations.19 21 44 The main conclusions of this review are that (1) biallelic EVC2 variants cause a more severe phenotype with increased prevalence of phenotypes associated with short tubular bones, (2) EVC missense variants cause a milder phenotype with reduced frequency of common skeletal findings, (3) variants affecting EVC and CRMP1, whose 3′ region partially overlaps with EVC, are associated with a higher prevalence of musculoskeletal and ectodermal dysplasia traits compared with those with EVC-only variants. Consistent with the results of this review, the data from our cohort show that specific musculoskeletal and ectodermal features are more common in patients with EVC/CRMP1 variants than in those with EVC-only variants (online supplemental table 2). In contrast, the frequency of specific phenotypes associated with short bones was reduced in our patients with biallelic EVC2 variants compared with those with EVC variants (online supplemental information). It is important to note that the number of patients in our cohort is not large enough to draw definitive conclusions. We were unable to assess the frequency of specific phenotypes in patients with missense variants and large indels due to the paucity of these variant types in our cohort.
In this study, the variant identified in the family with WAD confirms the association of WAD with heterozygous truncating variants in EVC2 exon 22. Prior to this study, six families with WAD had been published, and three had the same EVC2 variant.17–19 The variant reported here represents the fifth WAD-causing variant in the literature.
Features of patients with variants in WDR35 and GLI3
WDR35 codes for the intraflagellar transport protein IFT121 and has classically been linked to short-rib thoracic dysplasia with/without polydactyly and cranioectodermal dysplasia (CED). Biallelic WDR35 splicing variants were previously described in patients with EvC showing clinical overlap with CED in the form of renal and hepatic involvement, dolichocephaly and characteristic facial appearance.24 We report a novel homozygous splicing variant in a patient with classical findings of EvC, including natal teeth, four-limb postaxial polydactyly with brachydactyly and hypoplastic nails but no cardiac septation defects. The patient did not demonstrate classical craniofacial findings of CED. There was no evidence of renal and hepatic involvement, although age-related penetrance cannot be excluded due to the patient’s young age (3 months).
GLI3 is a dual transcription factor that mediates Hh signalling, whose function depends on the primary cilia. Pathogenic variants in GLI3 are typically associated with Greig cephalopolysyndactyly and Pallister-Hall syndrome. GLI3 mutations have also been implicated in various ciliopathy phenotypes comprising acrocallosal syndrome, orofaciodigital syndromes, isolated and syndromic forms of pre/postaxial polydactyly.45–49 We report for the first time the occurrence of a GLI3 variant in a patient with classical EvC features including nail hypoplasia, conical teeth, midline notching of the upper lip and postaxial polydactyly. This patient additionally had narrow thorax, which was previously not described in association with GLI3 variants, and cleft palate, which is not a strongly EvC-associated finding. Multiple frenula, a common manifestation of GLI3-associated phenotypes and EvC, was not present.
This large cohort of 49 patients with EvC/WAD widens the spectrum of EvC phenotypes. Our molecular results using a variety of techniques confirm that the vast majority of patients clinically diagnosed with EvC have pathogenic variants involving one or several nucleotides or structural variants in EVC or EVC2 and provide a comprehensive view of the genetic landscape of variants in these genes.
Our study also emphasises the diagnostic importance of functional validation of EVC/EVC2 VUS. Several splicing variants and one missense variant we initially classified as VUS can now be reclassified as likely pathogenic by the 2015 ACMG/AMP criteria, after their damaging effects were demonstrated through cDNA analysis in patient cells, and protein work, including immunoblotting and immunofluorescence, when possible. For one splicing variant, a minigene assay was performed due to the absence of patient material. Since EVC and EVC2 depend on each other to maintain their normal expression levels and ciliary localisation,12 phenotype rescue experiments using EVC or EVC2-knockout cells are useful tests for functional validation of EVC/EVC2 variants. We used this approach to assess the clinical significance of the EVC2 Ala246Asp variant, and similar tests were previously applied for evaluation of EVC variants.34
This work uncovers a new mutation in WDR35 in a patient diagnosed with EvC, giving further support to the association of WDR35 splicing mutations with an EvC phenotype. In addition, we report a de novo heterozygous GLI3 frameshift variant in one individual demonstrating EvC characteristics with no clear-cut clinical distinction with EVC/EVC2 patients. This provides a basis to place GLI3 among the genes causing EvC phenotypes.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by Hospital La Paz-CSIC: PI-4624. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We thank all patients and their families/legal guardians for participation in this study. We acknowledge Mert Kaya, PhD student, for his assistance in the preparation of tables.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
X @ElisaFernndez1
UA and AP-C contributed equally.
Contributors UA, HK and VR-P designed the study and coordinated the international collaboration. UA and VR-P drafted the manuscript with the help of GTT and C-LF. UA, NG, GTT, GO, RE, AA, AK, RFE, CICF, YG, EÖ, BSE, MI-R, BRD, BB, HA, ZY, RY, EÜ, EA, MA, HK and BT provided clinical data and collected patient samples. AP-C, JN, PL, MV, AI, EF-N, C-LF, PA, JT and VR-P performed molecular analyses and interpretation of molecular data. UA and GTT analysed phenotypic data. VR-P is responsible for the overall content as guarantor. All authors read and approved the final manuscript.
Funding This work was funded by the Spanish Ministry of Economy and Competitiveness (SAF2010-17901, SAF2013-43365-R, SAF2016-75434-R, PID2019-105620RB-I00/AEI/10.13039/501100011033) and FEDER funds through ISCIII grant PI20/01053 and PMP21/00063/Instituto de Salud Carlos III.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.