Article Text

Germline and somatic mosaicism in achondroplasia

Statistics from

Editor—We describe a sib recurrence in achondroplasia with parents of normal stature. Both affected offspring carry the same causal mutation (G1138C) in the fibroblast growth factor receptor 3 (FGFR3) gene. Despite having no clinical features of achondroplasia, a proportion of the mother's peripheral blood leucocytes also contained the mutantFGFR3 allele. We conclude she is a germline and somatic mosaic for achondroplasia and that both children have inherited the condition from her. To our knowledge, this is the first confirmed case of germline mosaicism in achondroplasia.

Achondroplasia is the commonest form of short limbed dwarfism (birth incidence estimated at between 1:10 000 and 1:70 000)1and is transmitted as an autosomal dominant trait. As is often the case among dominant traits, a high proportion of cases are new mutations but achondroplasia is unusual in that the great majority are caused by one of two mutations at the same nucleotide in the transmembrane domain of the FGFR3 gene (G1138A transition and G1138C transversion).1 In common with otherFGFR3 mutations which cause skeletal dysplasia, the pathogenic effect of the achondroplasia mutations is thought to be altered mitogenesis and/or differentiation owing to constitutive activation of the receptor.2 There is a marked paternal age effect in achondroplasia and it has recently been shown that new mutations in achondroplasia are almost exclusively of paternal origin.3 We received peripheral blood DNA from a family with two children with achondroplasia; both parents were of normal stature. They had a total of four children of whom the second and fourth were affected. The mother was 27 years of age and the father 53 years at the birth of their second affected child. Our first thoughts in this case, taking into account the age of the father, were that the affected sibs were the result of two independent new mutations in the paternal germline, as would be expected to occur by chance as a very rare event.

To determine the FGFR3 mutation(s) in the affected offspring, blood was collected from the affected children and from both parents. DNA was extracted and exon 10 of theFGFR3 gene was amplified and products were digested with the restriction enzymes BfmI and MspI. The G1138A transition creates a restriction site for BfmI whereas the G1138C transversion creates a restriction site forMspI. Analysis showed that both children were heterozygous for the rare G1138C transversion. The father did not have the mutation in his blood but, surprisingly, the mother did. However, the ratio of the G1138C allele compared to the wild type allele was less than the 1:1, which would be expected for a straightforward heterozygote (fig 1). The relative proportion of the G1138C allele in the mother's blood leucocytes was determined using primer extension4 (fig 1) followed by densitometry. The proportion of the mutant allele in the mother was found to be 28%. She has a height of 169 cm, span of 171 cm, upper segment/lower segment 0.09, left hand 17.6 cm, and head circumference of 58 cm. Apart from her slightly larger head size and mild obesity her appearance is normal.

Figure 1

(Upper panel). Digestion of 320 bp FGFR3 genomic PCR product with MspI. Lanes 1, father; 2, mother; 3, first affected child; 4, second affected child. Primers used were 5′-GGAGATCTTGTGCACGGTGG-3′ and 5′-GCGCGTGCTGAGGTTCTGAG-3′. (Lower panel) Primer extension was carried out with primer 5′-GATGAACAGGAAGAAGCCCA-3′ (which binds 4 bp downstream of nucleotide 1138) using methods described in Loughlin et al.4 The primer was end labelled with γ32P dATP and the extension mix contained dA, dT, dC, and ddG. The 20mer primer was extended by 4 bp for the achondroplastic G1138C allele and by 5 bp for the wild type allele as shown below (added nucleotides are shown underlined, nucleotide 1138 is shown in bold). Wild type template: 5′-C G GGGTGGGCTTCTTCCTGTTCATC-3′ Extended primer (25mer): 3′-ddGCCCCACCCGAAGAAGGACAAGTAG-5′ G1138C template: 5′-CCGGGTGGGCTTCTTCCTGTTCATC-3′ Extended primer (24mer): 3′-ddGCCCACCCGAAGAAGGACAAGTAG-5′   The products were then separated by electrophoresis through a 15% denaturing polyacrylamide gel. The relative intensity of the 24mer and 25mer products was used to calculate the proportion of achondroplastic to wild type allele. Lane order and PCR primers as above.

We conclude that, despite her normal appearance, the mother is a germline and somatic mosaic for the G1138C mutation and both her affected children have inherited the mutant allele from her. Given the mother's relatively high proportion of mutant alleles, her lack of phenotypic expression is surprising; a hypochondroplasia-like phenotype, which is less severe than achondroplasia, might have been expected. The most likely explanation for this is the tissue specific distribution of the mosaicism, although the mutant allele is present in 28% of her peripheral blood leucocytes it may be at lower levels in her chondrocytes.

Germline and somatic mosaicism are both reasonably common features of genetic disorders. For example, in Duchenne muscular dystrophy and osteogenesis imperfecta, 15% and 6% of cases, respectively, inherit the condition from a detectably mosaic parent.5 Germline mosaicism results from a mutation in gamete precursors which then continue to divide, whereas combined germline and somatic mosaicism arises when the mutation occurs very early in development before the germline and somatic lineages have separated. As achondroplasia is a common condition which arises from a highly mutable nucleotide, high frequencies of mosaicism might have been expected. Surprisingly, the frequency of germline mosaicism as evidenced by sib recurrence is very low. A few cases of recurrence have been reported,5 6 but so infrequently that it has been calculated that they could be accounted for by independent mutations alone. Clinical reports of somatic mosaicism in achondroplasia are also extremely rare.7 For some reason, somatic and germline mosaics occur much more rarely in achondroplasia than in many other dominant traits. One possibility is that for reasons as yet unclear, FGFR3 nucleotide 1138 is only hypermutable in the male germline. Alternatively, there could be somatic selection against cells carrying the mutant allele. Interestingly, Apert syndrome, which is mainly caused by either of two point mutations in FGFR2, also seems to have low levels of germline mosaicism.5 This apparent low incidence of somatic mutation is at variance with recent findings that somatic activating mutations of FGFR3 are relatively common in multiple myeloma8 and carcinomas.9 However, all theFGFR3 mutations so far identified in these malignant neoplasms are identical to activating mutations that cause thanatophoric dysplasia. The greater severity of this phenotype in comparison to achondroplasia is thought to be a reflection of the more strongly activating nature of the thanatophoric dysplasia mutations.10 That only these highly activatingFGFR3 mutations have so far been found in neoplasms may suggest that the achondroplasia mutations, when they occur in somatic cells, do not activate the receptor to a level that it becomes oncogenic.

This is the first confirmed report of germline and somatic mosaicism for an achondroplasia mutation. FGFR3nucleotide 1138 appears to be highly mutable in the male germline, but somatic mutations resulting in mosaicism are rare. The reasons for this discrepancy are unknown but are clearly of importance to the understanding of mutagenesis. The observation that the mother has a normal appearance, despite a high proportion of the achondroplastic allele in her somatic tissues, exemplifies the fine balance that the fibroblast growth factors play in morphological determination.


View Abstract

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.