Instability of a 550-base pair DNA segment and abnormal methylation in fragile X syndrome.

The fragile X syndrome, a common cause of inherited mental retardation, is characterized by an unusual mode of inheritance. Phenotypic expression has been linked to abnormal cytosine methylation of a single CpG island, at or very near the fragile site. Probes adjacent to this island detected very localized DNA rearrangements that constituted the fragile X mutations, and whose target was a 550-base pair GC-rich fragment. Normal transmitting males had a 150- to 400-base pair insertion that was inherited by their daughters either unchanged, or with small differences in size. Fragile X-positive individuals in the next generation had much larger fragments that differed among siblings and showed a generally heterogeneous pattern indicating somatic mutation. The mutated allele appeared unmethylated in normal transmitting males, methylated only on the inactive X chromosome in their daughters, and totally methylated in most fragile X males. However, some males had a mosaic pattern. Expression of the fragile X syndrome thus appears to result from a two-step mutation as well as a highly localized methylation. Carriers of the fragile X mutation can easily be detected regardless of sex or phenotypic expression, and rare apparent false negatives may result from genetic heterogeneity or misdiagnosis.

On some technical aspects of direct DNA diagnosis of the fragile X syndrome.

Direct DNA analysis of fragile X [Fra(X)] mutations has already shown its clear superiority for postnatal and prenatal diagnosis of the disorder and for carrier detection. However, it is of great importance to have conditions which guarantee optimal reliability and sensitivity. Some mutations may be more difficult to detect, especially in female carriers: this is the case for small amplifications of the CGG repeat (premutations) or for smears which can be generated by the instability of the full mutation in somatic tissues. We present the various alternatives (probe/enzymes combinations) for Southern blot based diagnosis, the possible artefacts and our detailed experimental protocol, which has given excellent results on a large number of families. While detection of amplification, using for instance EcoRI, appears sufficient for initial testing of mentally retarded patients, once the fra(X) diagnosis has been established, we favor the use of an EcoRI+EagI digest, which detects both amplification and abnormal methylation, for analysis of the family, including carrier detection and prenatal diagnosis. We discuss the place of proposed PCR based techniques for detection of mutations, or for indirect tracking using polymorphic microsatellites in the immediate vicinity of the fra(X) locus.

Direct DNA analysis of fragile X syndrome in Spanish pedigrees.

Eleven complete Spanish pedigrees with fragile X syndrome were analysed by Southern blotting with the DNA probe StB12.3 previously isolated and described by Oberlé et al. [1991]. This probe allowed the direct detection of affected males and carrier females and was able to distinguish between normal males and normal transmitting males (NTMs). One hundred and twenty three individuals were analyzed, 115 from the pedigrees and 8 from the general population. Five mosaic cases were found (4 males and one female) showing both the premutation and the full mutation. One half of the females with the full mutation were mentally retarded but no female with mental retardation carried the premutated pattern, suggesting that the absence of the full mutation in females is a very good criterion for pre-or postnatal diagnosis of normal mental status.

Improved DNA markers for efficient analysis of fragile X families.

We report the characteristics of two new probes that detect BclI RFLPs useful for analysis of fragile X families. With these two probes and a single blot, 34% of women are heterozygous both for the proximal marker DXS105 (closer to the fragile X locus than the factor IX gene) and for the distal markers DXS52 or the factor VIII gene. Combined with the analysis of previously described polymorphic markers, it is possible to have a majority of families fully informative for flanking markers using a limited number of probes and restriction digests.

Analysis of full fragile X mutations in fetal tissues and monozygotic twins indicate that abnormal methylation and somatic heterogeneity are established early in development.

The fragile X syndrome, the most common cause of inherited mental retardation, is characterized by unique genetic mechanisms, which include amplification of a CGG repeat and abnormal DNA methylation. We have proposed that 2 main types of mutations exist. Premutations do not cause mental retardation, and are characterized by an elongation of 70 to 500 bp, with little or no somatic heterogeneity and without abnormal methylation. Full mutations are associated with high risk of mental retardation, and consist of an amplification of 600 bp or more, with often extensive somatic heterogeneity, and with abnormal DNA methylation. To analyze whether the latter pattern is already established during fetal life, we have studied chorionic villi from 10 fetuses with a full mutation. In some cases we have compared them to corresponding fetal tissues. Our results indicate that somatic heterogeneity of the full mutation is established during (and possibly limited to) the very early stages of embryogenesis. This is supported by the extraordinary concordance in mutation patterns found in 2 sets of monozygotic twins (9 and 30 years old). While the methylation pattern specific of the inactive X chromosome appears rarely present on chorionic villi of normal females, the abnormal methylation characteristic of the full mutation was present in 8 of 9 male or female chorionic villi analyzed. This suggests that the methylation mechanisms responsible for establishing the inactive X chromosome pattern and the full mutation pattern are, at least in part, distinct. Our results validate the analysis of chorionic villi for direct prenatal diagnosis of the fragile X syndrome.

Three families with high expression of a fragile site at Xq27.3, lack of anomalies at the FMR-1 CpG island, and no clear phenotypic association.

We report on 3 families where the presence and segregation at high frequency of a fragile Xq27.3 site is not associated with the mutations and methylation anomalies typically seen in the fragile X [Fra(X)] syndrome. In one family, a folate insensitive fragile site was associated with Robin sequence in the propositus. In a second family a fra(X) negative mother has two fra(X) positive sons (one mentally retarded and the other newborn). The third family presents very high expression of a folate sensitive site, unlinked to mental retardation, and was described previously by Voelckel et al. [1989]. The fragile sites in these or similar families recently described must be different from the one associated with the fra(X) syndrome. Their association with a clinical phenotype or with mental retardation is certainly not consistent, and may represent an ascertainment bias. However, the relatively high frequency with which they have been found among previously diagnosed fra(X) families suggests that, at least in some cases, the association with mental impairment may be significant. In two families reported up to now, a male with high expression of such variant fra(X) site failed to transmit it to his daughter, which may reflect an imprinting effect. Previously diagnosed families should be reinvestigated before direct DNA analysis is used for prenatal or carrier diagnosis of the fra(X) syndrome.

New polymorphism and a new chromosome breakpoint establish the physical and genetic mapping of DXS369 in the DXS98-FRAXA interval.

Recently some of us cloned a new probe RN1 (DXS369), which appears a close marker for the fragile X locus (FRAXA) [Oostra et al.: Genomics 1990]. We present here new evidence for its physical and genetic mapping in the DXS98--FRAXA interval. We used 2 different somatic cell hybrid lines with breakpoints in the Xq27-q28 region: L10B Rea and PeCHN, and we established the order: (DXS105, DXS98)-L10B Rea-DXS369-PeCHN- (DXS304, DXS52). We detected an additional TaqI RFLP at the DXS369 locus which increases its informativeness up to 57%. Two point linkage analysis in a large set of families gave high lod scores for the FRAXA-DXS369 linkage (z(theta) = 10.1 at theta = 0.044) and for DXS369-DXS304, a marker distal to FRAXA (z = 19.2 at theta = 0.070). By multipoint analyses we established the localization of DXS369 in the DXS98-FRAXA interval. DXS369 is a much closer proximal marker for FRAXA than DXS105 or DXS98 and any new probe mapping between the breakpoints in L10B Rea and PeCHN will be of potential interest as a marker for FRAXA.

Physical and genetic mapping of the CDR gene with particular reference to its position with respect to the FRAXA site.

This study narrows down the localization of the gene coding for the cerebellar degeneration-related protein (CDR 34) to the upper boundary of the FRAXA and reports the finding of two common RFLPs respectively identified at an RsaI site flanking the 3' end of the gene and at a Hincll site flanking its 5' end. Segregation analysis carried out in the CEPH-pedigrees for the new CDR/RsaI-RFLP versus other polymorphic loci of the region has established a tight linkage with the markers DXS105/DX98 and absence of measurable linkage with two clusters of markers respectively located proximally to the FRAXA (F9, DXS102, DXS51, and DXS369) or distally to it (DXS52, DXS304). In addition, two recombinants were found among 23 scorable sibs identified in the Sardinian pedigrees segregating for the Martin-Bell Syndrome (MBS) and the CDR/RsaI variants. The overall evaluation of the in situ and genetic data reported suggest that the CDR locus 1) is located at the upper boundary of the FRAXA site; 2) is distal to DXS51 and proximal to DXS 389; and 3) segregates in a close linkage association with the loci DXS98 and DXS105 and, to a lesser extent, with the locus for MBS.

Methylation and mutation patterns in the fragile X syndrome.

Chromosomes carrying the mutation causing the fragile X [fra(X)] syndrome have been shown to have an unstable DNA sequence close to or within the fragile site. The length variation is located within a DNA fragment containing a CGG trinucleotide repeat which is unstable in both mitosis and meiosis. We have used the probe StB12.3 from the region to analyze the mutations and the methylation patterns in 21 families segregating for the fra(X) syndrome. Among 40 fra(X) males all showed an abnormal pattern. The normal 2.8 kb band was absent in 36 individuals and replaced by a heterogeneous smear of larger size. The remaining four were shown to be "mosaics" with the presence of both mutated, unmethylated and mutated, methylated fragments. We found four normal transmitting males, one which was a great-grandson of another normal transmitting male indicating that the pre-mutation can remain stable through two meioses in the female. In nine fra(X) positive females the abnormal pattern consisted of a smear, usually seen in affected males, in addition to the normal bands. Five of these females were mentally normal. Of clinical importance is the prediction of mental impairment in females. We suggest that this is not made by the detection of the full mutation alone, but rather by the degree of methylation of the normal X chromosome. Our results suggest that difference of clinical expression in monozygotic twins may be correlated with difference in methylation pattern. Six out of 33 fra(X) negative females at risk were diagnosed as carriers. Our observations indicate that molecular heterogeneity is responsible for variable expression of the fra(X) syndrome in both males and females.

A new polymorphic marker very closely linked to DXS52 in the q28 region of the human X chromosome.

We have isolated an X chromosome probe, St35.691 (DXS305), which detects two RFLPs with TaqI and PstI, whose combined heterozygosity is about 60%. This probe has been assigned to Xq28 by physical and genetic mapping and is very closely linked to DXS52, DXS15, and the coagulation factor VIII gene (F8C). The best estimate of the recombination fraction for the DXS52-DXS305 interval is 0.014, with a lod score of 50.1. Multipoint analysis places DXS305 on the same side of F8C as DXS52, but complete ordering of the three loci was not possible with our present data. This highly informative marker should be useful in the precise mapping of the many disease genes that have been assigned to the Xq28 band.

Genetic mapping of nine DNA markers in the q11----q22 region of the human X chromosome.

We have ordered nine polymorphic DNA markers within detailed map of the proximal part of the human X chromosome long arm, extending from band q11 to q22, by use of both physical mapping with a panel of rodent-human somatic hybrids and multipoint linkage analysis. Analysis of 44 families (including 17 families from the Centre d'Etude du Polymorphisme Humain) provided highly significant linkage data for both order and estimation of map distances between loci. We have obtained the following order: DXS1-DXS159-DXYS1-DXYS12-DXS3-(DXS94 , DXS178)-DXYS17. The most probable location of DXYS2 is between DXS159 and DXS3, close to DXYS1 and DXYS12. The high density of markers (nine loci within 30 recombination units) and the improvement in the estimation of recombination frequencies should be very useful for multipoint mapping of disease loci in this region and for diagnostic applications.

A new MspI restriction fragment length polymorphism in the hemophilia B locus.

Using a partial cDNA probe for human coagulation factor IX, we have detected a new restriction fragment length polymorphism in human DNA digested with MspI. The frequency of the minor allele is 0.20 +/- 0.05 and average heterozygosity is about 0.32. The MspI RFLP is in strong linkage disequilibrium with the TaqI RFLP previously described, but should nevertheless be useful in segregation analysis in case of homozygosity for the TaqI minor allele.

New informative polymorphism at the DXS304 locus, a close distal marker for the fragile X locus.

The polymorphic DNA marker DXS304 detected by probe U6.2 has recently been shown to be closer to the fragile X locus than previously available markers. Its usefulness has however been limited by its relatively low heterozygosity. We have isolated, by cosmid cloning, a 67 kilobase region around probe U6.2 and have characterized a new probe (U6.2-20E) that detects BanI and BstEII restriction fragment length polymorphisms (RFLPs). The BanI RFLP has a heterozygosity of 0.49 and is in partial linkage disequilibrium with the previously described polymorphism, with a combined heterozygosity of 0.63. Furthermore, we have found that the U6.2 original probe, which probably detects an insertion-deletion polymorphism, is also informative in BanI digests. Thus, the two informative RFLPs at the DXS304 locus can be conveniently tested in a single hybridization with a single digest. An updated linkage analysis confirms that DXS304 is distal to the fragile X locus. This informative locus can now be used effectively for genetic mapping of the Xq27-q28 region, and for diagnostic applications in fragile X or Hunter syndrome families.

Selection in blood cells from female carriers of the fragile X syndrome: inverse correlation between age and proportion of active X chromosomes carrying the full mutation.

We have studied the patterns of mutation and X inactivation in female carriers of a fragile X mutation, to try to correlate them with various phenotypic features. We used a simple assay, which shows simultaneously the size of the mutation, its methylation status, and DNA fragments that represent the normal active and inactive X chromosomes. We have observed an age dependent process, whereby the 'full' fragile X mutation is found preferentially on the inactive X in leucocytes in adult females, but not in younger ones. This phenomenon was not observed in female carriers of a 'premutation', who have little phenotypic expression. Preliminary data suggest that young females who show preferential presence of a full mutation on the active X in leucocytes may be at increased risk for mental retardation. We have also obtained preliminary evidence for an age dependent decrease in the somatic heterogeneity of full mutations, possibly owing to selection for smaller mutated fragments. If confirmed, the latter phenomenon might account for the known decrease with age of the expression of the fragile site. Our observations suggest that a gene whose expression is affected by the presence of a full mutation (possibly the FMR-1 gene) has a cell autonomous function in leucocytes, leading to a slowly progressive selection for cells where the mutation is on the inactive X chromosome.

Abnormal pattern detected in fragile-X patients by pulsed-field gel electrophoresis.

The fragile-X syndrome is the most frequent inherited form of mental retardation, with an incidence of 1 in 1,500 males. It is characterized by the presence of a fragile site at Xq27.3 induced in vitro by folate deprivation or by inhibitors of deoxynucleotide synthesis. Its mode of inheritance is unusual for an X-linked trait, with incomplete penetrance in both males and females. Some phenotypically normal males transmit the mutation to all their daughters who rarely express any symptoms, but penetrance is high in sons and daughters of these carrier women. Genetic and physical mapping of the Xq27-q28 region has confirmed that the disease locus is located at or very near the fragile site. Hypotheses proposed to account for the abnormalities in the inheritance of the disease include sequence rearrangements by meiotic recombination or a mutation that affects reactivation of an inactive X chromosome during differentiation of female germ cells. To detect such rearrangements, or methylation changes that may reflect a locally inactive X chromosome, we used pulsed-field gel analysis of DNA from fragile-X patients with probes close to the fragile-X locus. The probe Do33 (DXS465) detected abnormal patterns in fragile-X patients, but not in normal controls or in non-expressing male transmitters.

A case of female hemophilia with a 46,XXr karyotype studied with X-chromosome DNA probes.

A case of female hemophilia with a 46,XXr/45,X karyotype and signs of Turner syndrome, has been followed for the past 10 years. One of her brothers also has hemophilia A. A study with polymorphic DNA probes located in the Xq27-qter region has enabled us to demonstrate that the ring chromosome is of paternal origin and that the factor VIII gene region is deleted. The hemizygous state allowed expression of the hemophilia A mutation, present on the morphologically normal X chromosome, inherited from her carrier mother.

More precise localization of the gene for Hunter syndrome.

A linkage analysis between the Hunter syndrome locus (IDS) and four polymorphic loci of the Xq27-Xq28 region, DXS105, DXS98, DXS304, and DXS52, was performed in large families. A significant lod score was obtained between DXS304 and the Hunter gene (Zmax = 6.57 at theta max = 0.0). The Hunter gene can be localized within 7 cM of this marker. In addition, the translocation breakpoint of the Hunter female case described by J. Mossman et al. (1986, Arch. Dis. Child. 58: 911-915) was localized between DXS98 and DXS304 using somatic cell hybrids. These two results are in agreement and give the following order: DXS105-DXS98-IDS-DXS304-DXS52. Probes for these marker loci can thus be used for carrier detection.

The telomeric region of the human X chromosome long arm: presence of a highly polymorphic DNA marker and analysis of recombination frequency.

A DNA fragment (named St14) derived from the human X chromosome reveals a small family of related sequences that have been mapped to the Xq26-Xq28 region by using a panel of rodent-human somatic cell hybrids. The probe detects in human DNA digested by Taq I a polymorphic system defined by a series of at least eight allelic fragments with a calculated heterozygosity in females of 80%. With Msp I, we found three additional restriction fragment length polymorphisms, each of them being defined by two alleles. These polymorphisms are also common in Caucasian populations. The genetic locus defined by probe St14 has been localized more precisely to the distal end of the X chromosome (in band q28) by linkage analysis to other polymorphic DNA markers. The results obtained suggest that the frequency of recombination is distributed very unevenly in the q27-qter region of the X chromosome, with a cluster of seven tightly linked loci in q28 showing about 30% recombination with the gene for coagulation factor IX located in the neighboring q27 band. Probe St14 reveals one of the most polymorphic loci known to date in the human genome, and 17 different genotypes have already been observed. It constitutes the best marker on the X chromosome and should be of great use for the genetic study of three important diseases: hemophilia A, mental retardation with a fragile X chromosome, and adrenoleukodystrophy.

Physical and genetic mapping of polymorphic loci in Xq28 (DXS15, DXS52, and DXS134): analysis of a cosmid clone and a yeast artificial chromosome.

Sequences corresponding to the Xq28 loci DXS15, DXS52, DXS134, and DXS130 were shown to be present in a 140-kb yeast artificial chromosome (YAC XY58, isolated by Little et al.). This YAC clone appears to contain a faithful copy of this genomic region, as shown by comparison with human DNA and with a cosmid clone that contains probes St14c (part of the DXS52 sequences) and cpX67 (DXS134). cpX67 and St14c are contained in 11 kb and detect the same MspI RFLP polymorphism. A comparison of the YAC restriction map and pulsed-field gel electrophoresis data leads us to propose the following order of loci: DXS52(VNTR)-DXS33-DXF22S3-DXS130-DXS134 -DXS52-DXS15-DXS52, this whole cluster being comprised within 575 kb. The physical proximity of the DXS15, DXS52, and DXS134 loci led us to reinvestigate recombination events that had been reported between these loci in families from the Centre d'Etude du Polymorphisme Humain. Our results do not support the assumption that this region shows increased recombination.

The polymorphic marker DXS304 is within 5 centimorgans of the fragile X locus.

The fragile X syndrome, which is the most common cause of inherited mental retardation, poses important diagnostic problems for genetic counseling. The development of diagnostic strategies based on DNA analysis has been impaired by the lack of polymorphic markers very close to the disease locus. Here we report that the polymorphic probe U6.2 (locus DXS304) is much closer to the fragile X locus than all the previously reported markers. A recombination fraction of 0.02 between DXS304 and the fragile X locus was estimated by multipoint linkage analysis (confidence interval 0.002 to 0.05). Our data suggest that DXS304 is distal to the fragile X locus. This marker thus represents a major improvement for carrier detection and prenatal diagnosis in fragile X families.