Editor—Fragile X syndrome is the leading cause of inherited mental retardation, with an incidence that is generally estimated to be about 1/1250 to 1/4000 in males and 1/2000 to 1/8000 in females.1-3 An extremely high estimate was reported by Rousseau et al,4who indicated that as many as 1/259 females from Quebec, Canada, are premutation carriers. In contrast, Tranebjaerg et al 5 reported a prevalence of 0.04/1000 males in Funen, Denmark. The syndrome is characterised clinically by the triad of (1) long, narrow face with protruding chin and large ears, (2) macro-orchidism, and (3) mental retardation.6 The molecular basis for the disease is usually an expanded triplet (CGG) repeat located in the 5′ region of the fragile X mental retardation (FMR1) gene.7 8 In normal subjects, there are fewer than 60 copies of this CGG repeat; carrier females and transmitting males have a premutation that usually ranges from 60-200 repeats and in affected subjects the number is expanded to >200 copies.7 The mutations are associated with absence or reduction in FMR1 gene expression. “Fully expanded” alleles are heavily methylated, in contrast to normal alleles, which are unmethylated.9

Contrary to what might be expected of a very common disorder that confers a selective disadvantage, normal alleles appear to have a low mutation rate. The conversion of a normal allele to a premutation, or to a full mutation, has not been reported.10 Indirect evidence of a low mutation rate is provided by the finding of founder chromosomes.11 Studies of populations from the United States,12 France and Spain,13 Belgium/The Netherlands,14 northern Europe and the United States,15 Italy16, United Kingdom,17 Sweden,18 and Finland19-21 have shown thatFMR1 mutations are in apparent linkage disequilibrium with specific alleles at microsatellite loci FRAXAC1 and DXS548, which are located 7 kb and 150 kb proximal to the CGG repeat region, respectively. About two thirds of full mutationFMR1 alleles are associated with a few specific haplotypes. In some reports, the 204 bp allele at the DXS548 locus was associated with approximately 25% of fragile X and 8% of normal chromosomes. The Swedish and Finnish studies found linkage disequilibrium between FMR1 and other DXS548 alleles.

Given the high prevalence of fragile X syndrome reported elsewhere,1-3 a population the size of Nova Scotia (one million) is expected to include approximately 200 to 550 cases, with two to six newly identified patients each year. However, despite testing of patients presenting with mental handicap by cytogenetic analysis since 1980 (n=423) and mutation analysis since 1991 (n>650), only a single case has been identified. This family had recently moved here from elsewhere in Canada. This disorder may be extremely rare in Nova Scotia; however, patients might be overlooked, misdiagnosed, or not referred for laboratory testing. It has been estimated that more than 50% of fragile X cases are undiagnosed in The Netherlands.22 Because the implications to extended family members who may be at high risk for having affected children are significant, we screened subjects with mental retardation and no known diagnosis from seven regions of the province of Nova Scotia for CGG expansion. We also investigated the FMR1allele size distribution in a random sample of Nova Scotians.

Molecular analyses of DNA samples from 177 males with significant mental handicap, using PCR methods previously described by Fuet al 7 and Milleret al,23 identified 30 different FMR1 alleles with repeat sizes ranging from 19 to 59 triplets (fig 1A). Although three subjects had repeat sizes in the upper normal range (54, 57, 59), no premutations or full mutations were found. The absence of mutation in the samples with these large normal alleles was confirmed by Southern blotting using probe pE 5.1 and restriction enzymes EcoRI and BssHII.8

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

FMR-1 (CGG)n allele size distribution in (A) 167 institutionalised subjects with mental retardation and (B) 1226 random alleles from Guthrie spots.

The FMR1 CGG repeat size distribution in 1226 random alleles (470 males and 378 females) from Guthrie newborn screening samples (dried blood spots taken at birth) from the general Nova Scotia population was determined according to the methods of Carduci et al 24 and Fuet al.7 They contained 36 different allele sizes ranging from seven to 55 repeats (fig 1B). No premutations or fully expanded alleles were identified using this PCR based method. Thirty eight females showed only one allele size. Some of these subjects could be premutation or mutation carriers; however, assuming Hardy-Weinberg equilibrium and allele frequencies observed in male samples, we expected 45 homozygotes, which is more than we observed. The FMR1 allele size distribution differed between the patient and general population sample (χ²=12.34, 42 df, p=0.015). However, this result is apparently because of differences in regional representation in the two samples; chi-square analysis that takes geographical origin of the subjects into account showed no significance differences (analysis not shown).

Table 1 shows the allele distribution at locus DXS548 (determined as described by Zhong et al 12) within our institutionalised population compared to that in published groups of fragile X patients and normal subjects.12-21 25 26 Because 37 of the 177 samples collected from the institutionalised subjects were either no longer available or did not amplify by PCR at this locus, the sample size for this analysis is 140. Our institutionalised population is markedly different from each of the published fragile X groups and more closely resembles the normal groups of other populations. The frequency of the 204 bp allele, which is associated with fragile X in linkage disequilibrium in many other regions, is low in our population relative to most of the published normal populations, but not as low as one of the Chinese populations reported by Zhong et al.25

A comparison of the DXS548 allele distribution between the previously reported fragile X populations12 21 showed that these differed from one another (p=<0.00001) to a greater extent than did normal groups (p=0.053) from the same geographical regions. The greater diversity among fragile X populations was still apparent, even when the relatively isolated Scandinavian populations18-21 were excluded from the calculation (p=0.00006).

Studies from elsewhere have shown that approximately 2-10% of developmentally delayed patients are positive for the fragile X mutation by cytogenetic or molecular analysis.3 27-29 For example, 16 of 219 male and 13 of 128 female mentally retarded children studied from Birmingham, UK, had mutations in the FMR1gene.27 Similarly, a study of children with special educational needs from Salisbury, UK, found an expanded allele in four out of 180 male and 0 of 74 female cases.28 We have analysed a similar number of males with mental handicaps and found none. This patient population and our screen of 1226 random alleles showed bimodal size distributions similar to those reported elsewhere,19 28 but identified no allele expansions into the premutation or full mutation ranges. Fragile X syndrome, as suspected from our previous lack of detection through cytogenetic and molecular testing, is therefore rare in Nova Scotia.

The conclusion that fragile X syndrome is rare in Nova Scotia is also supported by molecular analysis of the microsatellite locus DXS548 in the same “high risk” patient population with mental retardation. Alleles at this locus have been shown to be in linkage disequilibrium with FMR1 expansions in many populations.11-20 We have shown that the allele distribution in our high risk group resembles that of other normal populations, with a dearth of those alleles commonly associated with fragile X.

It is noteworthy that the DXS548 allele distribution in previously studied fragile X populations12-21 is more heterogeneous that those in the corresponding normal populations.30 This observation suggests the possibility of multiple origins of the fragile X mutations from a limited number of pre-premutation alleles,30 31 with additional diversity generated by recombination and mutation of tightly linked microsatellite marker loci. Nova Scotia has long been noted for having a high prevalence of specific rare genetic disorders in various regions, for example, Niemann-Pick type D disease,32 Huntington disease,33 Charcot-Marie-Tooth disease,34acute intermittent porphyria,35 Fabry disease,36 and nephrogenic diabetes insipidus.37 Common ancestry of affected subjects in each of these cases has been documented, and molecular analysis supports the conclusion of a founder effect. Given the low rate of mutation, the absence of fragile X syndrome in Nova Scotia can be seen as an example of an “absence of founder effect”. Considering that our population (approximately one million) has tens of thousands of founders, from multiple founding groups (chiefly English, Scottish, Irish, French, and German immigrants in the 18th and 19th centuries), this phenomenon appears remarkable.