Editor—The study of achondroplasia, the most frequent skeletal dysplasia in man, has contributed several important insights into both developmental biology and human genetics, such as the recognition of the paternal age effect for dominant mutations,1 2 the first indication of the importance of FGFR molecules in growth and development,3 and the identification of the nucleotide with the highest mutation rate known so far in man, nucleotide 1138 of the FGFR3gene.4 Most cases of achondroplasia are associated with the g→a transition at nucleotide 1138 ofFGFR3.4

In spite of the frequency of achondroplasia, the birth of two or more children with achondroplasia to unaffected parents is surprisingly rare, with only a few examples published.5 6 One instance of half sibs with achondroplasia born to the same father has been reported.7 In contrast, observations of achondroplasia in people more distantly related are relatively more common.8-10 Thus, it is uncertain whether instances of achondroplasia in sibs born to unaffected parents are caused by somatic mosaicism (as suggested by the observation of three affected sibs from normal parents6 7) or by independent chance events.8 Undoubtedly, somatic or germinal mosaicism for achondroplasia must be orders of magnitude rarer than for osteogenesis imperfecta or other dominant conditions.11

Approximately, 90% of cases of achondroplasia are caused by de novo mutations, and all de novo achondroplasia mutations studied so far were found to have occurred on paternal chromosomes.12 We observed achondroplasia with characteristic clinical and radiographic signs in a brother and sister born to parents of normal stature, lacking any clinical sign of either hypochondroplasia or achondroplasia, aged 28 years (mother) and 25 years (father) at the time of birth of the first child (fig 1). The family agreed to have the molecular mechanism of recurrence investigated and consented to venepuncture and buccal smears. Both children were heterozygous (in leucocyte DNA) for the g1138a (G380R) FGFR3mutation, while that mutation was not found in parental leucocyte or buccal smear DNA by either SSCP analysis or direct sequencing of PCR products. This made parental somatic mosaicism unlikely.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Clinical appearance of the brother (right) and sister (left) with achondroplasia aged 12 and 10 years, respectively.

To investigate the origin of the mutation shared by the two affected sibs, inheritance of VNTR alleles on chromosome 4p was studied (fig 2). The affected children had two different maternal haplotypes but shared a paternal 4p haplotype encompassing theFGFR3 locus. As theFGFR3 g1138a mutation occurs exclusively on paternal chromosomes, and the affected children had two different maternal 4p haplotypes, the most likely explanations for these findings would be either two independent mutational events occurring by chance on the same paternal haplotype, or mosaicism at the spermatogonial level (before meiosis I) in the father. Paternal sperm was not available and the hypothesis of gonadal mosaicism could not be further substantiated.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Inheritance of chromosome 4 markers in the nuclear family. The markers are indicated on the left together with their genetic distance in centimorgans from the 4p telomere according to the GB4 map (http://www.ncbi.nlm.nih.gov/genemap98). F indicates the position of the FGFR3 locus and the asterisk indicates the presence of the achondroplasia mutation; its assignment to the paternal haplotype is not experimentally proven (but see text). The children have inherited two different chromosome 4 haplotypes from their mother, while they share a paternal haplotype on 4p encompassing the FGFR3 locus.

We conclude that recurrence of achondroplasia in this family was associated with de novo mutational event(s) occurring in the paternal germline, as is the case in sporadic cases,12 but could not distinguish between paternal gonadal mosaicism or the chance occurrence of two independent mutation events.8 The apparent rarity of somatic mosaicism forFGFR3 mutations in spite of the high mutation rate in achondroplasia remains unexplained.