Angelika A Noegel / Peter Nürnberg - C 11

Role of Wnt signaling in the etiology of Filippi syndrome and ectrodactyly ectodermal dysplasia


Filippi syndrome (FS) and ectrodactyly ectodermal dysplasia without cleft lip/palate (EECWOCLP) are developmental disorders characterized by craniodigital and limb deformities, respectively. These disorders have received insufficient attention in the past. We identified disease-causing DNA variants in CSNK2B and LEF1 implicated in Wnt signaling pathway for FS and EEC-WOCLP, respectively. We aim to study the functions of their corresponding proteins and to identify further genes in unsolved families. 


Filippi syndrome (MIM 272440) is a genetic disorder of craniodigital abnormalities associated with short stature, microcephaly, characteristic face, syndactyly and intellectual disability as the major phenotypes. In total, 32 patients from 24 families from different countries have been reported. In spite of the large number of reported patients and advancement in techniques to identify disease-associated genes, the molecular genetic cause of the Filippi syndrome was unknown until 2014. In that year, we published that mutations in CKAP2L may cause Filippi syndrome. Therewith, we defined the first disease gene for Filippi syndrome. 

Ectrodactyly ectodermal dysplasia without cleft lip/palate (EEC-WOCLP): (MIM 129810) is reported with clinical manifestation of split-hand/foot malformation (SHFM) accompanied by ectodermal anomalies like hypotrichosis and abnormal dentition. To date, no molecular genetic cause of this disorder is known. 

Our genomic analyses led to the discovery of the second causative gene of FS and first genetic cause implicated in the EEC-WOCLP. Investigating the functions of the corresponding proteins, we may get new insight into the molecular mechanisms underlying human embryonic development.

Mutation in CSNK2B, encoding casein kinase II subunit beta (CK2β), causes Filippi syndrome 

To date, we recruited 22 families of FS and were the first to describe a genetic cause of this rare disorder by revealing mutations in the CKAP2L gene in 6 of these families (Hussain et al., 2014). The remaining 16 families represent a precious pool to identify further molecular genetic causes. WES of a trio (both parents and a patient) representing an Italian family (FP2) led to the identification of a de novo mutation (c.94G>C; p.Asp32His) in CSNK2B, which we propose as the second candidate gene for Filippi syndrome. This mutation is not listed in any of the public databases and is predicted to be disease causing and probably damaging by Mutation Taster and PolyPhen-2, respectively. CSNK2B encodes CK2β which is the regulatory β subunit of protein kinase CK2, a holoenzyme with important roles in the Wnt signaling pathway. The protein kinase CK2 holoenzyme is a heterotetramer of the following compositions — αββα, αββα′ or α′ββα′. Our in silico analysis and depletion assay revealed that the mutation abolishes the interaction with the alpha subunit (Fig. 1). For the depletion assay, serial dilutions of CK2α were incubated with constant amounts of recombinant CK2β. In this assay significant amounts of unbound CK2β mutant were observed compared to wild type (Fig. 1B).   

Furthermore, patient derived LCLs revealed higher amounts of CK2β which mislocalized and abundant amounts were seen in the cytoplasm compared to wild type. The data were corroborated by overexpressed mutant CK2β in HeLa cells. Significant increase of β-catenin was seen in the cytoplasm of mutant LCLs compared to wild type. The mutation also impaired the interaction of CK2β with DVL3 (Dishevelled Segment Polarity Protein 3). As β-catenin and DVL3 are important members of the Wnt signaling pathway, we conclude that the mutation affects this pathway which is crucial for craniofacial development. 

A mutation in LEF1, encoding Lymphoid enhancer-binding factor 1 (LEF-1), causes EEC-WOCLP

We recruited three families of ectrodactyly from Pakistan and Yemen. The Yemeni family consists of two affected members (male and female) born to consanguineous healthy parents. The female patient died due to unknown medical reasons. The living boy is 10 years old. Both patients presented with ectrodactyly (Split-Hand/Foot) with syndactyly, hypotrichosis, oligodontia, mild intellectual disability and hypoplasia of mammary glands and nipples. WES identified a homozygous duplication of three base pairs in LEF1 (c.69_71dupGAT; p.M23_I24insM). This mutation is not reported in dbSNP138, the 1000 Genomes database, genomAD and EVS. Interestingly, Mutation Taster predicted this variant as disease causing. LEF1 encodes a protein of 399 amino acids and has a N-terminal β-catenin binding domain and C-terminal SOX-TCF_HMG-box domain. The mutation p.M23_I24insM is located in the β-catenin binding domain, which is highly conserved and essential for the Wnt signaling pathway. Met-23 in human LEF-1 In ß-catenin, the binding domain is situated in its armadillo (arm) repeat domain (Fig. 2A). The insertion of a further methionine in this region of LEF-1 may impair the specific arrangements of these consensus residues and thus the affinity for the β-catenin interaction may be abolished. Interestingly, the Lef1 knockout mouse showed phenotypes of absent teeth, mammary gland and hairs similar to those seen in our LEF1 patients but without limb deformities, whereas heterozygous mice did not show any phenotype. Ablation of Lef1 along with Tcf1 in mice mimics the limb defects seen in our LEF1 mutant.  

Immunofluorescence and confocal microscopy on patient derived primary fibroblasts and HeLa cells transiently expressing mutant and wild type protein revealed unchanged amount and localization of mutant LEF-1 compared to wild type. The binding efficiency of ß-catenin to the mutant LEF-1:p.M23_I24insM protein is impaired as compared to wild-type (Fig. 2B). RNA-seq. of mutant and control fibroblasts reveals differential expression of key targets of LEF-1/β-catenin during early limb development, like LGR5 (upregulation) and SALL4 (downregulation) compared to wild type. Intriguingly, differential expression of HOX genes (downregulation of HOXC10 and HOXB-AS3, upregulation of HOXD10, HOXD11 and HOXD13) was also monitored. This data was further corroborated by quantitative PCR. Interestingly, overexpression of Lgr5 in chicken embryos has already been reported with defective limbs morphology and mutations in HOXD13 cause digital abnormalities.


The planned analyses regarding CK2β and LEF-1 will add important facts to the emerging network of the involved pathways. To understand these novel pathways will be instrumental for the development of a better treatment of these disorders. Further, this study will improve the diagnostics of the patients and provide a solid basis for genetic counseling of the affected families. 

Selected publication

Hussain, M.S. et al. (2014). Mutations in CKAP2L, the Human Ortholog of the Mouse Radmis Gene, Cause Filippi Syndrome. Am. J. Hum. Genet. 95, 622-632.

Prof. Dr. Angelika A Noegel

Institute for Biochemistry I

Prof. Dr. Angelika A Noegel

Principal Investigator C 11

Work +49 221 478 6980

Fax (Work) +49 221 478 6979

Institute of Biochemistry I
Joseph-Stelzmann-Str. 52
50931 Cologne

Publications - Angelika A Noegel

Link to PubMed

Prof. Dr. Peter Nürnberg

Cologne Center for Genomics (CCG)

Prof. Dr. Peter Nürnberg

Co-Principal Investigator C 11

Work +49 221 478 96801

Fax (Work) +49 221 478 96803

Cologne Center for Genomics (CCG)
Weyertal 115 b
50931 Cologne

Publications - Peter Nürnberg

Link to PubMed

Group Members

Muhammad Sajid Hussain (Research Associate)
Salem Alawbathani (doctoral student)
Syeda Seema Waseem (doctoral student)
Arwa Khayyat (doctoral student)
Kathrin Schrage (doctoral student)
Maria Asif (doctoral student)
Maria Iqbal (doctoral student)
Martina Munck (technician)

Figure 1

CMMC Research Noegel
(A) Three-dimensional structures of CK2α and CK2β subunits with the salt bridge at the alpha/beta interface. (B) Coomassie Blue stained SDS-PAGE showing a depletion assay.

Figure 2

CMMC Research Noegel
(A) Three-dimensional structures of β-catenin armadillo domain (blue) along with LEF-1 consensus sequence shown as green stick, where Met-23 of LEF-1 in shown in red (PBD code 3oux). (B) A pull-down assay was performed to show the difference in the association of wild-type and mutant LEF-1 with β-catenin protein. LEF-1 was fused to GFP and β-Catenin to GST. Western blots were performed by probing with mouse monoclonal antibody K3-184-2 against GFP.1