Article Text


The congenital myasthenic syndrome mutation RAPSN N88K derives from an ancient Indo-European founder
  1. J S Müller1,*,
  2. A Abicht1,*,
  3. G Burke2,
  4. J Cossins2,
  5. P Richard3,
  6. S K Baumeister1,
  7. R Stucka1,
  8. B Eymard3,
  9. D Hantaï3,
  10. D Beeson2,
  11. H Lochmüller1
  1. 1Friedrich-Baur-Institute, Department of Neurology, and Gene Center, Ludwig-Maximilians-University, Munich, Germany
  2. 2Neurosciences Group, Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, UK
  3. 3INSERM U.582, Institut de Myologie and Unité Fonctionnelle de Cardiogénétique et Myogénétique, Hôpital de la Salpêtrière, Paris, France
  1. Correspondence to:
 Dr Hanns Lochmüller
 Genzentrum München, Feodor-Lynen-Str. 25, 81377 München, Germany;

Statistics from

Only recently, mutations of the RAPSN gene have been recognised as causing acetylcholine receptor deficiency at the motor endplate resulting in early and late onset forms of congenital myasthenic syndromes (CMS).1–9 In most studies a single missense mutation of RAPSN (N88K) was detected either homozygously or compound heterozygously in numerous, unrelated patients of European and North American origin. Based on the analysis of a small number of intragenic single nucleotide polymorphisms (SNPs) and/or extragenic, polymorphic repeat markers, we hypothesised that RAPSN (N88K) may derive from a common founder.7 However, this hypothesis was disputed in a recent report in the Journal of Medical Genetics.10 In this study, 12 independent RAPSN (N88K) alleles from white North American CMS patients were compared to 37 wild type chromosomes. Five intragenic SNPs and two extragenic microsatellite markers appeared to be in linkage disequilibrium with the mutation, whereas four extragenic microsatellite markers failed to prove a significant association.10 To resolve this question we analysed 41 independent RAPSN (N88K) alleles using 21 SNPs flanking RAPSN on chromosome 11p11.11


We collected 21 CMS patients harbouring RAPSN (N88K) either homozygously (n = 20) or compound heterozygously (n = 1). Screening for the mutation N88K (264C→A) in exon 2 of the RAPSN gene was performed as described previously.7 Patients originated from Germany (n = 5), Austria (n = 1), Italy (n = 1), France (n = 2), the United Kingdom (n = 10), and the Indian subcontinent (n = 2). Consanguinity was not reported for any of the families. Most of the patients have been described, previously: patients G1–G4, Au1, UK10, and It1 (patients 1–7 in Müller et al7), patients UK1–7 (patients 3, 9, 11, 12, 13, 15, and 16 in Burke et al6), patient Ind2 (patient 1 in Burke et al6), patient Fr1 (patient 1 in Richard et al8), and patient Fr2 (in Yasaki et al9). In addition, patients G5, UK8 and 9, and Ind1, who are included in this study, are homozygous for RAPSN (N88K) and show clinical features typical of CMS (data not shown). For the second Indian subcontinent patient (Ind2) with a second, heteroallelic mutation (46insC6), analysis of parental DNA was used to construct a haplotype for the N88K allele.

By PCR and restriction digest or sequence analysis, we analysed a total of 21 SNPs on chromosome 11p11 flanking RAPSN (N88K). Primer sequences and restriction enzymes used to screen the selected SNPs are given in table 1. Optimal SNPs flanking RAPSN were selected with a threshold of the minor allele frequency >20% according to the NCBI SNP database ( They are located within 4.012 Mb centromeric (SNPs 13–21; fig 1) and 2.688 Mb telomeric of the mutation (SNPs 1–8; fig 1). Four SNPs are located within the RAPSN gene (SNPs 9–12). In addition to the 21 CMS patients, we analysed 20 European controls for all SNPs. We defined a core founder haplotype encompassing SNPs 7–16, and an extended founder haplotype encompassing SNPs 3–16. For controls, DNA from the parents was not available which prevented exact phase determination for control haplotypes. Arbitrarily, haplotypes were constructed for the controls to result in the maximum possible number of core founder haplotypes (10/40) and extended founder haplotypes (5/40). Therefore, it appears likely that the number of founder haplotypes is overestimated for the controls. The χ2 test was used to compare mutant alleles with normal control alleles.

Table 1

 Primer sequences and restriction enzymes used to screen the selected SNPs

Figure 1

 Haplotype analysis of 21 unrelated CMS patients (41 RAPSN N88K chromosomes). The 21 SNPs flanking the mutation have been used for analysis. Vertical columns represent individual haplotypes of the analysed patients, whereas each row represents the genotypes at one SNP. The SNPs are listed from telomeric to centromeric corresponding to table 1; SNPs 9–12 are intragenic. The distance of each SNP refers to the position of the mutation N88K. Genotypes dominant in mutant chromosomes have been coloured orange (representing a putative founder haplotype), while a differing genotype has been coloured white. Yellow has been assigned to a heterozygous constellation at one SNP. All patients share a common haplotype stretching from SNP 7–16. *Patient Ind2 carries only one K88 allele. Au, Austria; Fr, France; G, Germany; Ind, Indian subcontinent; It, Italy; UK, United Kingdom.

Key points

  • Mutations in various genes of the neuromuscular junction cause congenital myasthenic syndromes (CMS). The protein rapsyn is encoded by the RAPSN gene and clusters acetylcholine receptors (AChR) at the motor endplate. Recessive mutations of RAPSN result in AChR deficiency and impaired neuromuscular transmission.

  • A single missense mutation of RAPSN (N88K) detected frequently in patients of European ethnic origin results in early and late onset forms of CMS. Three studies suggested that RAPSN (N88K) may derive from a common founder, while a fourth study could not corroborate this hypothesis. Therefore, we investigated 21 patients of European and Indian ethnic origin studying a total of 41 mutant RAPSN (N88K) alleles.

  • Analysis of 21 single nucleotide polymorphisms (SNPs) flanking RAPSN on chromosome 11p11 revealed a common conserved haplotype encompassing a distance of about 360 kb. Our results support the hypothesis that RAPSN (N88K) derives from a single founder event in an ancient Indo-European population.


All 41 RAPSN (N88K) chromosomes showed a shared core haplotype encompassing SNPs 7–16 (fig 1). Statistical analysis revealed that this haplotype was linked to N88K (p = 0.0006). This provides a strong indication that all RAPSN (N88K) alleles are derived from a common founder. Moreover, five patients of different European origin were homozygous for all analysed markers. For SNPs 3–6, only one RAPSN (N88K) allele of an Indian subcontinent patient revealed a different haplotype as compared to the other mutant chromosomes (shared haplotype encompassing SNPs 3–16). By contrast, several mutant chromosomes revealed distinct genotypes for SNPs 1–2 and/or SNPs 17–21 (fig 1). Statistical analysis for the extended haplotype spanning SNPs 3–16 also revealed significant linkage to N88K (p = 0.0000174).

A possible founder effect for RAPSN (N88K) was discussed in detail in three previous studies.7,8,10 An unambiguous conclusion was hampered by two facts. Firstly, each study looked at a relatively small number of mutant alleles, and the statistical power to positively prove a founder effect was low. Secondly, because the chromosomal region showing a conserved haplotype is relatively small, remote repeat markers were not as appropriate to show linkage. Therefore, we analysed 21 SNPs of the suspected founder region on chromosome 11p11.2 in a total of 41 mutant alleles. Our data demonstrate a core founder haplotype for RAPSN (N88K) encompassing a genetic distance of about 0.36 Mb in European patients. Most of the data available for North American patients are also compatible with an old founder effect.10 All North American patients are of Caucasian descent and show identical intragenic SNPs. The microsatellite marker D11S1252, which showed the same value for 11 of the 12 North American chromosomes, lies between SNPs 3 and 4 and hence within the extended founder haplotype fragment spanning SNPs 3–16. D11S4117, linked to N88K in all three previous studies, is directly adjacent to SNP 16 and might also be part of the core founder haplotype.7,8,10 D11S986 was not linked to N88K in previous studies, and lies clearly outside of the founder haplotype defined by SNP analysis. However, microsatellite marker D11S4109 exhibits four different genotypes in North American patients,10 but is located within the core haplotype defined by SNP analysis (between SNP 14 and 15). This may indicate additional founder events in North American patients. Alternatively, the variation in D11S4109 may have occurred after the mutation event and may reflect the higher divergence rate of nucleotide repeat markers as compared to SNPs. This question could be resolved by SNP analysis of the North American patients.

The gene RAPSN is located pericentromeric (6 Mb from the centromere). Previous studies revealed that the rate of meiotic exchange is reduced across centromeres of human X chromosomes12 and of autosomes13 as compared to more telomeric regions. For the pericentromeric region of chromosome 11 a low recombination rate of 0–0.1 Mb/cM was observed ( Therefore, estimation of the age of the founder mutation is difficult. According to historical and linguistic evidence ancient tribes migrated and divided around 2000 BC giving rise to the related Indo-European populations and languages. Therefore, the RAPSN (N88K) mutation may have originated prior to these divisions. This is in good agreement with two findings. Firstly, RAPSN (N88K) was detected in patients of different European and Indian subcontinent origin. Secondly, RAPSN (N88K) has not been reported in CMS patients of sub-Saharan African or East Asian ethnic origin. Similarly, an age of approximately 4000 years has been assigned to the founder mutation of myotonic dystrophy type II detected in various European and Afghan patients.14


We thank the patients for participating in this study and Ursula Klutzny for expert assistance.


View Abstract


  • * These authors contributed equally to this paper.

  • This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG) and the Deutsche Gesellschaft für Muskelkranke (DGM) to HL and AA, by grants from the Medical Research Council (UK) and the Muscular Dystrophy Campaign/Myasthenia Gravis Association of Great Britain to DB, and by grants from the Institut National de la Santé et de la Recherche Médicale (INSERM), the Assistance Publique-Hôpitaux de Paris (APHP), and the Association Française contre les Myopathies (AFM) to PR, BE, and DH. JSM receives a scholarship from the Boehringer Ingelheim Fonds. SKB receives a scholarship through the program for molecular medicine of the Ludwig-Maximilians-University, Munich.

  • Conflict of interest: none declared.

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.