Article Text

Download PDFPDF

Mutation analysis in the candidate Möbius syndrome genes PGT and GATA2 on chromosome 3 and EGR2 on chromosome 10
  1. B van der Zwaag1,
  2. H T F M Verzijl1,
  3. D Beltran-Valero de Bernabe2,
  4. V L Schuster3,
  5. H van Bokhoven2,
  6. H Kremer2,
  7. M van Reen2,
  8. G H Wichers2,
  9. H G Brunner2,
  10. G W Padberg1
  1. 1Department of Neurology, University Medical Centre Nijmegen, Nijmegen, The Netherlands
  2. 2Department of Human Genetics, University Medical Centre Nijmegen, Nijmegen, The Netherlands
  3. 3Department of Medicine, Albert Einstein College of Medicine, Bronx, New York, USA
  1. Correspondence to:
 Dr B van der Zwaag, University Medical Centre Nijmegen, 321 Reinier Postlaan 4, 6525 GC Nijmegen, The Netherlands;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Möbius syndrome (MBS, MIM 157900) is a rare congenital disorder characterised by paralysis of the facial nerve. This paralysis may be complete or partial and unilateral or bilateral. Other cranial nerves are often implicated, most frequently the abducens and hypoglossal nerve. Limb malformations and facial dysmorphism occur frequently. Features seen less often in MBS are structural anomalies of the ear, defective branchial musculature (Poland syndrome, MIM 173800), and mild mental retardation.1 Although Möbius syndrome usually occurs in isolated cases, familial recurrence has been reported. Patterns of inheritance observed in affected families suggest different modes of inheritance for the syndrome, ranging from autosomal recessive to autosomal dominant and X linked.

To date, four genetic loci for MBS have been described. In 1977, a reciprocal translocation of 13q12.2-13 was identified, cosegregating with the disease in a three generation MBS family.2 Slee et al3 reported a MBS patient with a deletion of chromosome 13q12.2 in 1991. Therefore, chromosome 13q12.2-q13 is thought to harbour a gene for MBS (MBS1). We identified two additional loci for MBS, MBS2 at 3q21-q224 and MBS3 at 10q21,1 in two large Dutch families with a mild Möbius phenotype, cosegregating in an autosomal dominant fashion with reduced penetrance. A fourth locus harbouring a gene for MBS on chromosome 1p22 was inferred from two reports. Donahue et al5 identified a t(1;11)(p22;p13) translocation in a patient with Möbius syndrome. The 1p22 locus was confirmed by a t(1;2)(p22.3;q21.1) translocation in a patient with Möbius-like syndrome, reported by Nishikawa et al.6 All these data combined prove genetic heterogeneity for MBS.

Although no conclusive evidence has been gathered so far, two modes of action have been postulated to explain the aetiology of MBS. Firstly, a primary metameric defect in the brain-stem nuclei in the region of the tegmentum could result in MBS. Secondly, an ischaemic process resulting from an interruption of the vascular supply of the brainstem and other structures during early fetal development could be causative.1 In line with these hypotheses, different genetic defects in a broad range of processes underlying neurogenesis, axonal outgrowth, or angiogenesis could result in the features observed in Möbius syndrome.

In order to identify candidate genes for Möbius syndrome, we directed our efforts at genes included in the MBS2 and MBS3 loci, which could play a role in the development of the hindbrain, the guidance of axons, and angiogenesis. Linkage analysis on additional members of the MBS2 linked family4 enabled us to reduce the MBS2 critical region to a 4.9 cM region between the markers D3S1589 and ACPP. Our mutation analysis included the human prostaglandin transporter (PGT) gene and the gene encoding the GATA binding protein 2 (GATA2) transcription factor at the MBS2 locus, and the early growth response 2 (EGR2) gene at the MBS3 locus. Mutation analysis comprised direct sequencing of subjects from the relevant families (either MBS2 or MBS3) on an ABI-PRISM 377 (EGR2) or 3700 DNA analyzer (GATA2, PGT), using BigDye terminator chemistry (Applied Biosystems).

PGT (or SLC21A2) was analysed because a correlation was suggested between the maternal use of a synthetic prostaglandin, Misoprostol, illegally used as an abortificant, and an increased incidence of Möbius syndrome in newborns following failed abortion attempts.7–9 Primers for the amplification and direct sequencing of PGT are listed in table 1. Mutation analysis of the PGT gene in patients from the MBS2 linked family showed only one nucleotide change, in exon 9. This A to G nucleotide change (base pair 1269) results in the substitution of threonine 396 for alanine (numbering according to sequence data from Genbank accession number NM_005630). The nucleotide change did not cosegregate with the disease, and several subjects related by marriage also carried the A to G nucleotide change. Therefore, this change in exon 9 can be classified as a polymorphism. No other changes were encountered in the coding sequence (CDs) and flanking intronic sequences of PGT. Therefore it is unlikely that the PGT gene is involved in MBS2.

The second gene analysed at the MBS2 locus at 3q21-q22 was GATA2, a member of the GATA binding protein family of transcription factors.10 This gene is an interesting candidate for two reasons. (1) GATA2 expression in the developing hindbrain is limited to rhombomere 4, an important structure for the development of the facial nerve,11 and (2) GATA2 expression is regulated by Hoxb-1, a homeobox domain protein that functions in patterning of the hindbrain.12 The Hoxb-1 knockout mouse shows a defect in the formation of the motor nucleus of the facial nerve, and is considered an appropriate animal model for Bell's palsy and Möbius syndrome.13 Primers were designed flanking the seven exons of GATA2 (two alternatively used exons of the 5`UTR and five coding exons), shown in table 1. No nucleotide changes were found in the CDs or splice sites in patients from the MBS2 linked family. This lack of mutations in the GATA2 gene suggests that this gene is not the causative gene for MBS2.

The early growth response 2 gene, coding for the EGR2 transcription factor, is located at 10q21.3 near the MBS3 critical region. The EGR2 gene was included in the mutation analysis based on the function of the mouse orthologue Krox-20. Krox-20 is preferentially expressed in rhombomeres 3 and 5 in the developing hindbrain,14 embracing rhombomere 4. During brain development, Krox-20 regulates the expression of several homeobox proteins in rhombomeres 3 and 5 that are important for patterning of the hindbrain. Genes regulated by Krox-20 include Hoxa-2, Hoxb-2, and Hoxb-3.15–17 As rhombomeres 4 and 5 are the site of origin of the majority of the cells that eventually make up the facial nerve,11 failure to express functional EGR2 could give rise to a Möbius phenotype. Although mutations in the EGR2 gene have been described in patients with congenital hypomyelinating neuropathy (CHN), and Charcot-Marie-Tooth disease type 1 (CMT1),18 allelism with MBS3 cannot be excluded. Primers were designed for amplification and direct sequencing of the EGR2 coding sequence and splice sites (table 1). No mutations were found in the coding sequence and splice sites of the EGR2 gene in patients from the MBS3 linked family. This suggests that the EGR2 gene is not involved in MBS3.

In summary, the lack of pathogenic mutations in the coding sequences and splice sites of the three genes investigated justifies the exclusion of these genes as candidate genes in MBS2 (PGT and GATA2) or MBS3 (EGR2).

In a future positional candidate gene approach, new candidate genes need to be considered for mutation analysis in MBS2 or MBS3 patients. As Möbius syndrome is a rare disorder, familial cases are hard to find. Additional familial Möbius syndrome patient material would be very useful, as this increases the chances of identifying causative mutations.

Table 1

Oligonucleotide sequences flanking each of the exons of the PGT, GATA2, and EGR2 genes. GATA2 has two alternatively used, non-coding, exons 1; we named these exon 1a and 1b. Exon 2 of the EGR2 gene was divided into eight amplicons (a–h) to facilitate PCR and subsequent sequencing. Amplification conditions are available upon request


We are very grateful to the members of the families who participated in this study. We thank Ms J W M Jeuken for critical reading of the manuscript. This work was partly funded by NWO grant 901-04-183 to HGB.