Population screening at the FRAXA and FRAXE loci: molecular analyses of boys with learning difficulties and their mothers

Transcription

1 1996 Oxford University Press Human Molecular Genetics, 1996, Vol. 5, No ARTICLE Population screening at the FRAXA and FRAXE loci: molecular analyses of boys with learning difficulties and their mothers Anna Murray 1, Sheila Youings 1, Nick Dennis 2, Lorinda Latsky 1, Paul Linehan 3, Nicky McKechnie 2, James Macpherson 1, Michelle Pound 1 and Patricia Jacobs 1 1 Wessex Regional Genetics Laboratory, Salisbury District Hospital, Salisbury, Wiltshire SP2 8BJ, UK, 2 Wessex Clinical Genetics Service, Princess Anne Hospital, Southampton, Hampshire SO16 5YA, UK and 3 CRC Epidemiology Research Group, Princess Anne Hospital, Southampton, Hampshire SO16 5YA, UK Received February 12, 1996; Revised and Accepted March 13, 1996 Preliminary results on a large population-based molecular survey of FRAXA and FRAXE are reported. All boys with unexplained learning difficulties are eligible for inclusion in the study and data are presented on the first 1013 tested. Individuals were tested for the number of trinucleotide repeats at FRAXA and FRAXE and typed for four flanking microsatellite markers. Mothers of 760 boys were tested to determine the stability of the FRAXA and FRAXE repeats during transmission and to provide a population of control chromosomes. The frequency of FRAXA full mutations was 0.5%, which gives a population frequency of 1 in 4994, considerably less than previous reports suggest. No FRAXE full mutations were detected, confirming the rarity of this mutation. In the boys X chromosomes, we detected one FRAXA premutation with 152 repeats and one putative FRAXE premutation of 87 repeats. No full or premutations were seen in the control chromosomes. A significant excess of intermediate alleles at both FRAXA and FRAXE was detected in the boys X chromosomes by comparison with the maternal control chromosomes. This suggests that relatively large unmethylated repeats of sizes for FRAXA and for FRAXE may play some role in mental impairment. No instability was found in transmissions of minimal or common alleles in either FRAXA or FRAXE, but we saw two possible instabilities in transmission of FRAXA and two definite instabilities in transmission of FRAXE among 43 meioses involving intermediate or premutation sized alleles. We found no linkage disequilibrium between FRAXA and FRAXE but did find significant linkage disequilibrium between large alleles at FRAXE and allele 3 at the polymorphic locus DXS1691 situated 5 kb distal to FRAXE. INTRODUCTION Mutations at two loci on distal Xq, FRAXA and FRAXE, may cause mental impairment. Almost all deleterious mutations at the FRAXA locus are due to expansion of a CGG trinucleotide repeat at the 5 end of exon 1 in the FMR1 gene (1 3). This expansion can exist in two forms, an unstable premutation which has not been thought to be associated with mental impairment and which appears to have no effect on the level of the FMR1 protein (4,5), and the full mutation which is also unstable but in which the FMR1 gene is silenced by hypermethylation of the CpG island (6). The absence of the FMR1 protein is the cause of the mental impairment and the subtle but characteristic phenotype of the fragile X syndrome (7). While the degree of mental impairment associated with the full mutation in males is variable, the great majority have a moderate level of retardation with an IQ between 30 and 50 (8). The FRAXE locus was first identified in patients with folate-sensitive fragile sites at Xq27 28 but no evidence of CGG expansion at the FRAXA locus (9,10). Subsequent studies revealed FRAXE to be a polymorphic GCC repeat associated with a nearby CpG island, 600 kb distal to FRAXA (11). In patients expressing the fragile site, the number of repeats is significantly greater than normal and there is concomitant hypermethylation at the CpG island (11). As yet, the FRAXE gene has not been identified, although candidate genes have been suggested (12,13) and it is assumed that such a gene would operate in a manner similar to FRAXA (14). The FRAXE mutation is associated with a mild level of mental impairment but, unlike FRAXA, there appear to be no characteristic phenotypic features (15 17). However, relatively few families with a FRAXE mutation have been described and thus information about its phenotype or the behaviour of the GCC repeat is sparse. *To whom correspondence should be addressed

2 728 Human Molecular Genetics, 1996, Vol. 5, No. 6 In common with other triplet repeat diseases, FRAXE demonstrates anticipation with progressive increases in repeat number over successive generations in affected families (11,14,17). Reductions in repeat number at FRAXE are also frequently observed, including reversions from a full to a premutation, and in this respect FRAXE differs from FRAXA (11,16,17). Males with the full FRAXE mutation have been known to reproduce and may have affected daughters (16), in contrast to FRAXA where full mutation males rarely reproduce and when they do so they transmit a premutation allele to their daughters (18 20). There does not, therefore, appear to be the same sex-specific restriction on transmission of FRAXE alleles as is seen in FRAXA. The incidence of the fragile X syndrome has generally been thought to be between 1 in 1000 and 1 in 2600 males, based on cytogenetic surveys of defined populations of individuals with mental impairment (21,22). However, it is now evident that the fragile sites observed included the rare fragile site FRAXE as well as the common fragile site at Xq28. Thus the incidence of FRAXA full or premutations in the population is not reliably known. The incidence of FRAXE full mutations is unknown. The majority of FRAXE full mutations reported have been ascertained as a result of testing for FRAXA. Screening surveys for FRAXE are rare; one recently reported survey found no cases of FRAXE full mutation in 300 males tested (23). We are currently undertaking a large population survey of boys aged 5 18 in the state school system who have learning difficulties for which there is no well established cause. We are also obtaining DNA from their mothers. The X chromosome passed from mother to son is considered the experimental chromosome, while that not transmitted is the control chromosome. The purposes of the survey include the determination of (i) the frequency of unusual alleles, including those in the pre- and full mutation size range, at the FRAXA and FRAXE loci in the experimental and control chromosomes; (ii) the haplotype in relation to four flanking polymorphisms (DXS548, FRAXAC1, FRAXAC2, DXS1691) in the region of the X chromosome containing the FRAXA and FRAXE loci and the relationship between the haplotype and size of the FRAXA and FRAXE alleles; and (iii) the stability of alleles in transmission from mother to son. RESULTS The total population of 5 18 year old boys in the region studied was , of whom were in schools that agreed to participate. In the participating schools, there were 1502 (4%) boys who met our criteria for inclusion in the study and of these 1068 (71%) agreed to provide a DNA sample. The 434 boys from Table 1. Distribution of FRAXA and FRAXE alleles by category whom we did not obtain a sample consisted of 55 who declined to participate and 379 from whom we had no reply to our letters. We obtained a DNA sample from 760 mothers representing 71% of the participating boys. DNA was obtained from mouth brushes. We found no difficulty with this approach. Samples from 98% of boys and 95% of mothers yielded satisfactory DNA. Successful FRAXA results were achieved for 1013 boys and 726 mothers and successful FRAXE results for 992 boys and 725 mothers. The FRAXA and FRAXE alleles were partitioned into five categories on the basis of their size (Table 1). FRAXA locus We identified five males with a full mutation among the 1013 boys with learning disabilities whom we tested (Table 1). Four of these boys were mosaics with both a pre- and full mutation sized allele and two were already known to us. The five fragile X boys were drawn from a total population of boys. Of the 1502 eligible boys in this population, 1013 or 67% have been tested for FRAXA. Thus, in effect, the population size is (67% of ), giving an estimated population frequency of 1 in The distribution of allele sizes for FRAXA is shown in Figure 1 and the distribution by category in Table 1. When the minimal and common categories combined are compared with the intermediate and premutation categories combined, there is a significant excess of the latter class among the experimental chromosomes (χ 2 1 = 4.26, P = 0.039). The distribution of intermediate and premutation sized alleles in the experimental and control chromosomes is shown in Table 2. As can be seen, the excess of large alleles among the experimental chromosomes appears to involve alleles across the whole size range, not merely those with >50 repeats. FRAXE locus We found no cases of FRAXE full mutation in 992 boys successfully tested, giving a 95% confidence interval of (Table 1). Of the 1502 eligible boys in the population, 992 or 66% have been tested for FRAXE. Thus, the population size is effectively (66% of ). If we assume that all FRAXE full mutation males would have been included among the eligible boys, the 95% confidence interval in the total population is Twenty-one boys and one mother failed FRAXE PCR testing although successful for FRAXA. All were from the first school visited in the survey and failed for at least one other marker. This suggests DNA of poor quality rather than an unrecognized cluster of FRAXE full or premutations. These individuals were excluded from the analysis. Category FRAXA FRAXE No. of Experimental Control No. of Experimental Control repeats no. (%) no. (%) repeats no. (%) no. (%) Minimal (M) <11 1 (0.1) 0 <11 6 (0.6) 6 (0.8) Common (C) (95.9) 712 (98.1) (98.3) 717 (98.9) Intermediate (I) (3.6) 14 (1.9) (1) 2 (0.3) Premutation (P) (0.1) (0.1) 0 Full mutation (F) >200 5 (0.5) 0 > Total a 725 a a Twenty-one boys and one mother failed FRAXE testing although they were successful for FRAXA.

3 729 Nucleic Human Acids Molecular Research, Genetics, 1994, 1996, Vol. 22, Vol. No. 5, No Figure 1. Comparison of FRAXA allele distributions in experimental and control chromosomes. The distribution of FRAXE repeat alleles in the boys was compared with that in the control X chromosomes (Fig. 2). The distribution in each group had two modes, one at 15 repeats and a second at 24 or 25 repeats; the two distributions appeared superimposable, apart from the right hand tail, which was longer in the boys, as described below. Alleles of >30 or <11 repeats were rare. The smallest allele observed in this series was of four repeats seen in a patient with a FRAXA full mutation. No alleles were sequenced, so deletions in the region flanking the triplet repeat cannot be excluded in this or any other patient. When the minimal and common categories combined are compared with the intermediate and premutation categories combined there is a

4 730 Human Molecular Genetics, 1996, Vol. 5, No. 6 significant excess of the latter class among the experimental chromosomes (χ 2 1 = 4.39, P = 0.036). The distribution of the intermediate and premutation alleles is shown in Table 2. Linkage disequilibrium We confirmed the presence of linkage disequilibrium between DXS548, FRAXAC1, FRAXA and FRAXAC2, as has already been shown (24 26). The haplotype was much more frequent among alleles with repeat numbers >40 than those with <40 repeats (Table 3). Among alleles with >40 repeats, there was a higher proportion of the haplotype in the experimental than the control chromosomes, but the difference was not significant (χ 2 1 = 1.44, P>0.2). We also looked for linkage disequilibrium across the 600 kb FRAXA/FRAXE interval; analysis was performed on alleles grouped into categories for FRAXA and FRAXE and on individual pairs of alleles for the other loci. No significant disequilibrium was found between any loci across this 600 kb interval. The dinucleotide repeat polymorphism physically closest to the FRAXE triplet repeat in our study is DXS1691, which lies 5 kb distal to the repeat. This polymorphism has three common alleles (1 = 106 bp, 2 = 104 bp, 3 = 102 bp) and we found a highly significant association between allele 3 and intermediate alleles at FRAXE (P<0.001). Ninety two percent of FRAXE repeats in the intermediate and premutation range are in association with DXS1691 allele 3 compared with only 28% in the common and minimal range (Table 4). Table 2. Distribution of intermediate and premutation FRAXA and FRAXE alleles Category FRAXA FRAXE No. of CGG Experimental Control No. of GCC Experimental Control repeats (N = 1013) (N = 726) repeats (N = 992) (N = 725) Intermediate Premutation Total Table 3. Distribution of haplotype for loci DXS548, FRAXAC1 and FRAXAC2 in FRAXA FRAXA Experimental chromosome Control chromosome No. of repeats All Haplotype All Haplotype (Category) Haplotypes N (%) Haplotypes N (%) <41 (M+C) (2) (3) (I+P) (47) 14 4 (29) >200 (F) 5 1 (20) 0 Table 4. Distribution of alleles at locus DXS1691 in FRAXE FRAXE No. of Experimental chromosome Control chromosome repeats (category) Total Allele 3 haplotype Total Allele 3 haplotype N (%) N (%) <31 (M+C) (27) (31) (I+P) (91) 2 2 (100)

5 731 Nucleic Human Acids Molecular Research, Genetics, 1994, 1996, Vol. 22, Vol. No. 5, No Figure 2. Comparison of FRAXE allele distributions in experimental and control chromosomes. Stability of individual alleles Stability during transmission was examined by comparing allele sizes in mothers and boys; in this way 726 female meioses were studied. At FRAXA, two possible changes in repeat number were detected, excluding full and premutations, and both were in the intermediate range; a reduction of one repeat from 43 to 42 and an increase of one repeat from 53 to 54. The large size of these alleles means that the PCR products generated have numerous stutter bands and appear rather smeared. This makes interpretation difficult, hence we have described these as possible differences which await further investigation. At FRAXE, two cases of instability were detected. The first was an intermediate allele in a mother who had 37 and 17 repeats

6 732 Human Molecular Genetics, 1996, Vol. 5, No. 6 Figure 3. Gels showing FRAXE PCR products labelled with 32 P from the two cases in which an instability in repeat number during transmission was detected. whose son had a mosaic pattern of 37 and 27 repeats in both buccal and blood samples (Fig. 3a). A Southern blot showed a normal male pattern with a single band in the unmethylated range, excluding the possibility of further mosaicism into the affected range and also confirming the presence of only a single X chromosome in this individual. As expected, the two bands detected by PCR could not be resolved on the Southern blot. The second instance of instability at FRAXE was a putative premutation allele in a boy which had increased in size by 21 repeats, from 66 to 87 in transmission from mother to son (Fig. 3b). Southern blot analysis showed the allele to be unmethylated in both individuals, with an expansion size consistent with the PCR data. DISCUSSION The frequency of 0.5% FRAXA full mutations in a population of males with learning difficulties is considerably lower than those reported from other surveys and is in part due to the fact that we tested many boys who had only borderline intellectual disability. In the majority of large surveys of boys with learning difficulties, some 2 4% were found to be fragile X positive by cytogenetic diagnosis (21,22). It is now generally agreed that the frequency of full fragile X mutations is considerably less than previously reported on the basis of cytogenetic surveys (27 29). Molecular retesting of fragile X individuals originally identified in cytogenetic surveys gives an estimate for the frequency of full mutations of the FRAXA gene of 1 in 4000 males (28). This is a minimum frequency because molecular retesting of cytogenetically positive individuals identifies all false positives but does not identify any false negatives. Common sense and experience dictate that such individuals exist (30). While the frequency of 1 in 4994 males obtained in our survey is less than the recalculated estimates based on cytogenetic surveys, and also less than the frequency of 1 in 3625 males found in our pilot survey (27), the differences among the three estimates are not significant (χ 2 2 = 0.23, P >0.8). Thus the frequency of FRAXA may well be only 1 in in 5000 males in Caucasian outbred populations. The frequency of premutations is critically dependent on their precise definition, and on the accuracy with which the number of repeats is measured. We found a single premutation sized allele for FRAXA among the experimental chromosomes. No premutations were seen among 726 control chromosomes, although we did observe one allele with 52 and one with 53 repeats in the controls (Table 2). There are a number of reports of the frequency of premutation sized alleles in FRAXA (31 33). Unfortunately, the techniques used to determine the alleles and the definitions of premutation alleles differ widely among the surveys so it is almost impossible to compare them. In the largest (33), X chromosomes were examined and 41 (0.19%) found with >55 CGG repeats. These data suggest that 1 in 500 X chromosomes carries a premutation sized allele as defined by Rousseau et al. (33) on Southern blots. The frequency of intermediate and premutation sized alleles and of full fragile X mutations might well differ among different populations, depending on the distribution of haplotypes and their pattern of AGG interspersed repeats. Our data suggest that full mutations of the FRAXE gene are rare. There has been controversy over the FRAXE full mutation phenotype; although the majority of individuals identified with FRAXE full mutations have mild mental retardation (11,15,16), others are said to be not significantly different from family members who do not carry the mutation (9). Such findings may be due to assortative mating among the mildly retarded population, a situation not usually found in the more severely retarded, who tend not to reproduce. One boy studied had an unstably inherited FRAXE allele of 87 repeats, and to our knowledge this is the first example of a FRAXE premutation ascertained in the absence of a full mutation proband. Convincing non-mosaic premutations at the FRAXE locus have been rarely described, although premutation/full mutation mosaics are common (11,14,16,17). The apparent rarity of the FRAXE premutation in families in which the full FRAXE mutation is segregating may indicate that premutations are less stable in FRAXE than FRAXA. Our finding of an increased frequency of intermediate FRAXA alleles among the probands is unexpected because it is thought that intermediate alleles and premutations have no effect on intellectual performance (34 36). However, there are no studies in which intellectual performance in fragile X premutation carriers has been rigorously compared with intellectual performance in an appropriate control population. Recently it has been shown that premutations are associated with an increased risk of premature ovarian failure (37) and possibly also of twinning (38), so it may be that they also have a subtle effect on learning ability. We found a similar significant excess of FRAXE intermediate alleles among the experimental chromosomes. If this unexpected excess of intermediate and premutation alleles for both FRAXA and FRAXE loci is independently confirmed, it will suggest that such alleles are not as benign as previously supposed. Interestingly, inspection of data from at least one other survey, in which sizes of repeats in chromosomes of retarded individuals are compared with those of appropriate controls, suggests a similar phenomenon (39). Possible explanations for this apparent excess of intermediate alleles include: (i) somatic expansion in tissues not normally sampled such as the brain; (ii) unusual secondary structure at either the DNA or mrna level, with a concomitant

7 733 Nucleic Human Acids Molecular Research, Genetics, 1994, 1996, Vol. 22, Vol. No. 5, No adverse effect on gene function (40,41); and (iii) the presence of genes for mental retardation in linkage disequilibrium with the haplotypes associated with large alleles, i.e for FRAXA and DXS1691 allele 3 for FRAXE. Linkage disequilibrium analysis on our data shows that the six markers tested can be viewed as two separate groups, group 1 (DXS548, FRAXAC1, FRAXA and FRAXAC2) and group 2 (FRAXE and DXS1691), with linkage disequilibrium within groups but not between groups. This is not surprising, given the physical distance of 600 kb between the groups. It is well established that large normal alleles and FRAXA pre- and full mutations are found more often than expected on a haplotype for loci DXS548, FRAXAC1 and FRAXAC2 respectively (26). Similarly at FRAXE, all intermediate alleles but one were associated with allele 3 of DXS1691 (Table 4). Both full FRAXE mutations previously detected by us (11,14) also have allele 3 at locus DXS1691 as does the premutation case in the current survey. The excess of allele 3 in association with both large normal FRAXE repeats and full mutation alleles suggests that FRAXE full mutations can arise from a pool of large normal or intermediate alleles which are non-randomly associated with DXS1691 allele 3, and are at increased risk of expansion, presumably due to their size. It has been suggested for FRAXA that one mechanism of triplet repeat expansion may involve the loss of an AGG triplet from within the CGG repeat as an initial mutational event (42,43). It is postulated that such AGG punctuation of the otherwise pure CGG repeat confers stability upon the repeat tract during replication, and hence loss of an AGG would increase the risk of instability (44,45). Evidence to date shows that the FRAXE repeat is pure (46,47) and this suggests that length may be the main, if not the only, important indicator of the stability of a FRAXE allele. On the basis of present evidence, we suggest that alleles >30 repeats (i.e. the classes we term intermediate and premutation) be regarded as potentially unstable. Such alleles are rare, accounting for only 0.76% of our sample. Two of these, with 37 and 66 repeats respectively, were unstably inherited. This on-going survey is accumulating a large sample of data which indicates a new, provisional estimate of 1 in 5000 for the prevalence of FRAXA. Continuation of the survey to include a further 3000 tested boys will refine this figure and also provide a more accurate estimate of FRAXE prevalence. The recognized associations between haplotype, trinucleotide repeat substructure and pattern of instability should become more clearly defined, enabling more confident predictions of the mechanism(s) of instability. In addition, more substantive data on the stability of intermediate alleles and their possible role in mild mental impairment should help to resolve important dilemmas in genetic counselling. MATERIALS AND METHODS The survey population and tests done In order to identify boys in the state school system who were considered to have learning difficulties for which there was no established cause, we contacted the head teachers in all 253 state schools in the study districts and 240 agreed to participate in the project. The school medical officers obtained a list of the relevant boys in each participating school. The survey was not exclusive, and any boy considered to have learning difficulties was included. The reason for this was our desire to obtain frequencies for the FRAXE mutation which is thought to be associated with only a mild degree of learning disability. When the population eligible for testing was identified, we contacted the parents by letter asking firstly whether they would be willing for us to obtain a mouth brush sample from their son and secondly whether the mother herself would be willing to provide us with a mouth brush sample. If within 3 weeks there was no reply to our initial letter, we sent a second letter. However, if we had no reply to the second letter no further action was taken. DNA extraction and PCR Each participating school was visited, and two mouth brushes, one from the inner aspect of each cheek, obtained from each boy for whom we had parental consent. A padded, addressed envelope containing two mouth brushes was sent to each consenting mother together with instructions for self-sampling. Buccal cells from both mouth brushes (Medical Wire and Equipment Co. Ltd, Corsham, Wiltshire, UK) were suspended in 500 µl of resuspension buffer (75 mm NaCl, 24.5 mm EDTA, ph 8.0), plus proteinase K (0.3 mg/ml) and SDS (0.2%) and incubated at 37 C overnight. DNA was extracted by the salt precipitation technique (48). The resulting DNA pellet was resuspended in 50 µl of TE; 1 µl of DNA solution was used in each PCR reaction. Initial PCR reations were carried out using fluorescently labelled oligonucleotide primers supplied by Oswel DNA service. Two multiplex PCR reactions were carried out; for FRAXA and FRAXE a final reaction volume of 15 µl was used with 1 Amplitaq buffer, 2.2 mm magnesium chloride, 0.2 mm datp/dctp/dttp (Sigma), 10% dimethylsulphoxide (DMSO; Sigma), 0.15 mm 7-deaza dgtp (Boehringer), 0.05 mm dgtp (Sigma), 0.2 µm of primers c (labelled 5 with 6-FAM) and f (31), 0.7 µm of primers 598 (labelled 5 with HEX) and 603 (11) and U/µl of Amplitaq (Perkin Elmer Cetus). PCR was carried out, following an initial hot start, for 30 cycles of denaturation at 95 C for 1.5 min, annealing at 65 C for 1 min and extension at 72 C for 2 min, followed by a final extension at 72 C for 7 min. The dinucleotide repeat multiplex was also in a final reaction volume of 15 µl plus 1 Amplitaq buffer, 1.5 mm MgCl 2, 0.2 mm dntps, 0.05% glycerol, 0.3 µm of primers AC1A (labelled 5 with TAMRA) and AC1B (24), µm of primers AC2A (labelled 5 with 6-FAM) and AC2B (24), 0.13 µm of primers F322 (labelled 5 with HEX) and F010 (16), 0.13 µm of primers DXS548A (labelled 5 with HEX) and DXS548B (1) and 0.05 U/µl of Amplitaq (Perkin Elmer Cetus). PCR was carried out, following an initial hot start, for 10 cycles of denaturation at 94 C for 1 min, annealing at 60 C for 1.5 min and extension at 72 C for 1.5 min. This was followed by a second round of amplification for 25 cycles, of denaturation at 94 C for 1 min, annealing at 55 C for 1.5 min and extension at 72 C for 1.5 min, followed by a final extension at 72 C for 7 min. From both multiplex reactions, 0.5 µl were mixed with 0.5 µl of ROX 500 molecular weight marker (ABI/Perkin Elmer) and 1.5 µl of formamide dye buffer, denatured and separated on a sequencing grade, 6%, polyacrylamide 7 M urea gel (Gibco BRL). The gels were run on an ABI 373A Stretch machine using 24 cm well to read plates for 8 h at 40 W, 2500 V and 35 ma. Gel data were analysed on 672 GENESCAN software (ABI/Perkin Elmer) and then imported into GENOTYPER software (ABI/Perkin Elmer) to assign alleles. Due to inefficient amplification of larger FRAXA

8 734 Human Molecular Genetics, 1996, Vol. 5, No. 6 and FRAXE alleles with fluorescent primers, a conventional 32 P-labelled PCR reaction was occasionally required (11,27). As the PCR technique we used does not distinguish between homozygotes and hemizygotes for any allele and, as it does not amplify many expansions over 100 repeats, it is not possible to distinguish between females who are truly homozygous at the FRAXA or FRAXE loci and those with one normal sized allele and a second pre- or full mutation sized allele. Therefore, we obtained a blood sample from all women who appeared homozygous for FRAXA or FRAXE alleles for which the expected frequency of homozygosity was <1% and tested them by Southern blot analysis. This meant that we did not retest females who were apparently homozygous for the common FRAXA alleles with 29, 30 and 31 repeats and the FRAXE alleles with 14, 15, 16, 17 and 18 repeats. We also obtained a blood sample from any individual who appeared from initial PCR analysis to have a FRAXA allele with >50 repeats or a FRAXE allele with >40 repeats, and used Southern blot analysis to check for possible mosaicism into the full mutation ranges. Southern blot DNA (0.5 µg) was double digested with either EcoRI and BstZI (methylation-sensitive) for the FRAXA protocol or HindIII and NotI (methylation-sensitive) for FRAXE. Gels were Southern blotted and hybridized with 32 P-labelled probe, StB12.3 for FRAXA (34) or OxE20 for FRAXE (11). ACKNOWLEDGEMENTS We would like to thank Hampshire County Council Education Department and the head teachers of all schools that took part in the survey. We are extremely grateful to the community paediatricians for Southampton and Winchester, Dr C. A. Smalley, Dr V. Shrubb and Dr A. G. Antoniou and their teams of clinical medical officers without whose help and advice the study would not have been possible. We thank Professor N. E. Morton for very helpful discussion. This work was supported by a programme grant from the Wellcome Trust. REFERENCES 1. Verkerk, A.J.M.H., Pieretti, M., Sutcliffe, J.S., Fu, Y.H., Kuhl, D.P.A., Pizzuti, A., Reiner, O., Richards, S., Victoria, M.F., Zhang, F.P., Eussen, B.E., van Ommen, G.J.B., Blonden, L.A.J., Riggins, G.J., Chastain, J.L., Kunst, C.B., Caskey, C.T., Nelson, D.L., Oostra, B.A. and Warren, S.T. (1991) Identification of a gene (FMR-1) containing a CGG repeat coincident with a Fragile X breakpoint cluster region which exhibits length variation in fragile-x syndrome. Cell, 65, Oberlé, I., Rousseau, F., Heitz, D., Kretz, C., Devys, D., Hanauer, A., Boue, J., Bertheas, M.F. and Mandel, J.L. (1991) Instability of a 550-base pair DNA segment and abnormal methylation in Fragile X syndrome. Science, 252, Yu, S., Pritchard, M., Kremer, E., Lynch, M., Nancarrow, J., Baker, E., Holman, K., Mulley, J.C., Warren, S.T., Schlessinger, D., Sutherland, G.R. and Richard, R.I. (1991) Fragile X genotype characterized by an unstable region of DNA. Science, 252, Devys, D., Lutz, Y., Rouyer, N., Bellocq, J.P. and Mandel, J.L. (1993) The FMR-1 protein is cytoplasmic, most abundant in neurons and appears normal in carriers of a Fragile X premutation. Nature Genet., 4, Feng, Y., Lakkis, L., Devys, D. and Warren, S.T. (1995) Quantitative comparison of FMRI gene expression in normal and premutation alleles. Am. J. Hum. Genet., 56, Bell, M.V., Hirst, M.C., Nakahori, Y., MacKinnon, R.N., Roche, A., Flint, T.J., Jacobs, P.A., Tommerup, N., Tranebjaerg, L., Froster-Iskenius, U., Kerr, B., Turner, G., Lindenbaum, R.H., Winter, R., Pembrey, M., Thibodeau, S. and Davies, K.E. (1991) Physical mapping across the fragile X: hypermethylation and clinical expression of the Fragile X syndrome. Cell, 64, Pieretti, M., Zhang, F.P., Fu, Y.H., Warren, S.T., Oostra, B.A., Caskey, C.T. and Nelson, D.L. (1991) Absence of expression of the FMR-1 gene in Fragile-X syndrome. Cell, 66, Fisch, G.S., Arinami, T., Froster-Iskenius, U., Fryns, J.-P., Curfs, L.M., Borghgraef, M., Howard-Peebles, P.N., Schwartz, C.E., Simensen, R.J. and Shapiro, L.R. (1991) Relationship between age and IQ among fragile X males: a multicenter study. Am. J. Med. Genet., 38, Sutherland, G.R. and Baker, E. (1992) Characterisation of a new rare fragile site easily confused with the fragile X. Hum. Mol. Genet., 1, Flynn, G.A., Hirst, M.C., Knight, S.J.L., Macpherson, J.N., Barber, J.C.K., Flannery, A.V., Davies, K.E. and Buckle, V.J. (1993) Identification of the FRAXE fragile site in two families ascertained for X linked mental retardation. J. Med. Genet., 30, Knight, S.J.L., Flannery, A.V., Hirst, M.C., Campbell, L., Christodoulou, Z., Phelps, S.R., Pointon, J., Middletonprice, H.R., Barnicoat, A., Pembrey, M.E., Holland, J., Oostra, B.A., Bobrow, M. and Davies, K.E. (1993) Trinucleotide repeat amplification and hypermethylation of a CpG island in FRAXE mental retardation. Cell, 74, Gedeon, A.K., Keinänen, M., Adès, L.C., Kääriäinen, H., Gécz, J., Baker, E., Sutherland, G.R. and Mulley, J.C. (1995) Overlapping submicroscopic deletions in Xq28 in two unrelated boys with developmental disorders: identification of a gene near FRAXE. Am. J. Hum. Genet., 56, Chakrabarti, L., Knight, S.J.L., Flannery, A.V. and Davies, K.E. (1996) A candidate gene for mild mental handicap at the FRAXE fragile site. Hum. Mol. Genet., 5, Knight, S.J.L., Voelckel, M.A., Hirst, M.C., Flannery, A.V., Moncla, A. and Davies, K.E. (1994) Triplet repeat expansion at the FRAXE locus and X-linked mild mental handicap. Am. J. Hum. Genet., 55, Dennis, N.R., Curtis, G., Macpherson, J.N. and Jacobs, P.A. (1992) Two families with Xq27.3 fragility, no detectable insert in the FMR-1 gene, mild mental impairment and absence of the Martin Bell phenotype. Am. J. Med. Genet., 43, Hamel, B.C..J., Smits, A.P.T., Degraaff, E., Smeets, D.F.C.M., Schoute, F., Eussen, B.H.J., Knight, S.J.L., Davies, K.E., Assmanhulsmans, C.F.C.H. and Oostra, B.A. (1994) Segregation of FRAXE in a large family: clinical, psychometric, cytogenetic, and molecular data. Am. J. Hum. Genet., 55, Mulley, J.C., Yu, S., Loesch, D.Z., Hay, D.A., Donnelly, A., Gedeon, A.K., Carbonell, P., Lopez, I., Glover, G., Gabarron, I., Yu, P.W.L., Baker, E., Haan, E.A., Hockey, A., Knight, S.J.L., Davies, K.E., Richards, R.I. and Sutherland, G.R.(1995) FRAXE and mental retardation. J. Med. Genet., 32, Yu, W.D., Wenger, S.L. and Steele, M.W. (1990) X-chromosome imprinting in Fragile-X syndrome. Hum. Genet., 85, Willems, P.J., Van Roy, B., De Boulle, K., Vits, L., Reyniers, E., Beck, O., Dumon, J.E., Verkerk, A. and Oostra, B. (1992) Segregation of the fragile X mutation from an affected male to his normal daughter. Hum. Mol. Genet., 1, Reyniers, E., Vits, L., Deboulle, K., Vanroy, B., Vanvelzen, D., Degraaff, E., Verkerk, A.J.M.H., Jorens, H.Z.J., Darby, J.K., Oostra, B. and Willems, P.J. (1993) The full mutation in the FMR-1 gene of male fragile-x patients is absent in their sperm. Nature Genet., 4, Webb, T.P., Bundey, S.E., Thank, A.I. and Todd, J. (1986) Population incidence and segregation ratios in the Martin Bell syndrome. Am. J. Med. Genet., 23, Turner, G., Robinson, H., Laing, S. and Purvis-Smith, S. (1986) Preventive screening for the Fragile X syndrome. N. Engl. J. Med., 315, Allingham-Hawkins, D.J. and Ray, P.N. (1995) FRAXE expansion is not a common etiological factor among developmentally delayed males. Am. J. Hum. Genet., 57, Richards, R.I., Shen, Y., Holman, K., Kozman, H., Hyland, V.J., Mulley, J.C. and Sutherland, G.R. (1991) Fragile-X syndrome diagnosis using highly polymorphic microsatellite markers. Am. J. Hum. Genet., 48, Oudet, C., Mornet, E., Serre, J.L., Thomas, F., Lenteszengerling, S., Kretz, C., Deluchat, C., Tejada, I., Boué, J., Boué, A. and Mandel, J.L. (1993) Linkage disequilibrium between the fragile-x mutation and two closely linked CA repeats suggests that fragile-x chromosomes are derived from a small number of founder chromosomes. Am. J. Hum. Genet., 52,

9 735 Nucleic Human Acids Molecular Research, Genetics, 1994, 1996, Vol. 22, Vol. No. 5, No Macpherson, J.N., Bullman, H., Youings, S.A. and Jacobs, P.A. (1994) Insert size and flanking haplotype in fragile X and normal populations possible multiple origins for the fragile X mutation. Hum. Mol. Genet., 3, Jacobs, P.A., Bullman, H., Macpherson, J., Youings, S., Rooney, V., Watson, A. and Dennis, N.R. (1993) Population studies of the fragile X: a molecular approach. J. Med. Genet., 30, Turner, G., Webb, T., Wake, S. and Robinson, H. (1996) The prevalence of the fragile X syndrome. Am. J. Med. Genet., in press. 29. Hagerman, R.J., Wilson, P., Staley, L.W., Lang, K.A., Fan, T., Uhlhorn, C., Jewellsmart, S., Hull, C., Drisko, J., Flom, K. and Taylor, A.K. (1994) Evaluation of school children at high risk for Fragile X syndrome utilizing buccal cell FMR-1 testing. Am. J. Med. Genet., 51, Van den Ouweland, A.M.W., de Vries, B.B.A., Bakker, L.G., Deelen, W.H., de Graaff, E., van Hemel, N.O., Oostra, B.A., Niermeijer, M.F. and Halley, D.J.J. (1994) DNA diagnosis of the Fragile X syndrome in a series of 236 mentally retarded subjects and evidence for a reversal of mutation in the FMR-1 gene. Am. J. Med. Genet., 51, Fu, Y.H., Kuhl, D.P.A., Pizzuti, A., Pieretti, M., Sutcliffe, J.S., Richards, S., Verkerk, A.J.M.H., Holden, J.J.A., Fenwick, R.G., Warren, S.T., Oostra, B.A., Nelson, D.L. and Caskey, C.T. (1991) Variation of the CGG repeat at the fragile-x site results in genetic instability resolution of the Sherman paradox. Cell, 67, Reiss, A.L., Kazazian, H.H., Krebs, C.M., Mcaughan, A., Boehm, C.D., Abrams, M.T. and Nelson, D.L. (1994) Frequency and stability of the fragile X premutation. Hum. Mol. Genet., 3, Rousseau, F., Rouillard, P., Morel, M.-L., Khandjian, E.W. and Morgan, K. (1995) Prevalence of carriers of premutation-size alleles of the FMR1 gene and implications for the population genetics of the Fragile X syndrome. Am. J. Hum. Genet., 57, Rousseau, F., Heitz, D., Biancalana, V., Blumenfeld, S., Kretz, C., Boué, J., Tommerup, N., Vanderhagen, C., Delozierblanchet, C., Croquette, M.F., Gilgenkrantz, S., Jalbert, P., Voelckel, M.A., Oberlé, I. and Mandel, J.L. (1991) Direct diagnosis by DNA analysis of the Fragile X syndrome of mental retardation. N. Engl. J. Med., 325, Reiss, A.L., Freund, L., Abrams, M.T., Boehm, C. and Kazazian, H. (1993) Neurobehavioural effects of the fragile X premutation in adult women: a controlled study. Am. J. Hum. Genet., 52, Rousseau, F., Heitz, D., Tarleton, J., Macpherson, J., Malmgren, H., Dahl, N., Barnicoat, A., Mathew, C., Mornet, E., Tejada, I., Maddalena, A., Spiegel, R., Schinzel, A., Marcos, J.A.G., Schorderet, D.F., Schaap, T., Maccioni, L., Russo, S., Jacobs, P.A., Schwartz, C. and Mandel, J.L. (1994) A multicenter study on genotype phenotype correlations in the Fragile X syndrome, using direct diagnosis with probe StB 12.3: the first 2,253 cases. Am. J. Hum. Genet., 55, Conway, G.S., Hettiarachchi, S., Murray, A. and Jacobs, P.A. (1995) Fragile X premutations in familial premature ovarian failure. Lancet, 346, Turner, G., Robinson, H., Wake, S. and Martin, N. (1994) Dizygous twinning and premature menopause in Fragile X syndrome. Lancet, 344, Hofstee, Y., Arinami, T. and Hamaguchi, H. (1994) Comparison between the cytogenetic test for Fragile X and the molecular analysis of the FMR-1 gene in Japanese mentally retarded individuals. Am. J. Med. Genet., 51, Mitchell, J.E., Newbury, S.F. and McClellan, J.A. (1995) Compact structures of d(cng) n oligonucleotides in solution and their possible relevance to fragile X and related human genetic diseases. Nucleic Acids Res., 23, Gacy, A.M., Goellner, G., Juranic, N., Macura, S. and McMurray, C.T. (1995) Trinucleotide repeats that expand in human disease form hairpin structures in vitro. Cell, 81, Eichler, E.E., Hammond, H.A., Macpherson, J.N., Ward, P.A. and Nelson, D.L. (1995) Population survey of the human FMR1 CGG repeat substructure suggests biased polarity for the loss of AGG interruptions. Hum. Mol. Genet., 4, Snow, K., Tester, D.J., Kruckeberg, K.E., Schaid, D.J. and Thibodeau, S.N. (1994) Sequence analysis of the fragile X trinucleotide repeat: implications for the origin of the fragile X mutation. Hum. Mol. Genet., 3, Eichler, E.E., Holden, J.J.A., Popovich, B.W., Reiss, A.L., Snow, K., Thibodeau, S.N., Richards, C.S., Ward, P.A. and Nelson, D.L. (1994) Length of uninterrupted CGG repeats determines instability in the FMR1 gene. Nature Genet., 8, Kunst, C.B. and Warren, S.T. (1994) Cryptic and polar variation of the fragile X repeat could result in predisposing normal alleles. Cell, 77, Ritchie, R.J., Knight, S.J.L., Hirst, M.C., Grewal, P.K., Bobrow, M., Cross, G.S. and Davies, K.E. (1994) The cloning of FRAXF: trinucleotide repeat expansion and methylation at a third fragile site in distal Xqter. Hum. Mol. Genet., 3, Zhong, N.(1995) Presentation at VII International Workshop on the Fragile X and X-Linked Mental Retardation. 48. Miller, S.A., Dykes, D.D. and Polesky, H.F. (1988) A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res., 16,