Fragile XA syndrome appeared on the cytogenetic scene in the late 1970s, although by the mid-1990s it occupied a predominantly molecular genetic stage. But it is entirely appropriate that it retains a place, and indeed a whole chapter, in this book, with the evolution of cytogenetics into molecular cytogenetics. Fragile XA syndrome is the second most important genetic cause of mental deficiency after Down syndrome, and is the most common familial cause. The premutational state is associated with a risk for premature menopause in the female and a degenerative neurological syndrome in late middle-age in males. Fragile XE syndrome is rare, but may be the most common form of nondysmorphic mild intellectual disability (Knight et al., 1996; Gecz, 2000).
BIOLOGY
Fragile XA syndrome is named for the folate-sensitive fragile site, FRAXA, at Xq27.3 (Fig. 14-1). There are two other rare fragile sites distal to FRAXA, named FRAXE and FRAXF. FRAXE is a rare folate-sensitive fragile site that is the chromosomal manifestation of an expanded CCG repeat that silences the FMR2 gene, which is the molecular basis of its associated form of nonspecific mental retardation (Gécz et al., 1996). FRAXF appears to be a harmless rare variant. FRAXD is a common fragile site, proximal to FRAXA, and is also harmless and a part of normal chromosome structure.
The FRAXA fragile site exists in the 5 region of the FMR1 (fragile X mental retardation-1) locus that encodes a protein named FMRP. FMRP binds to mRNA and is necessary for normal brain development and function, having a role in either synaptic function or dendrite growth (Darnell et al., 2001; Irwin et al., 2001). A specific observation in brain tissue from patients with fragile X syndrome is an excess of immature forms of the “spines” that project from the branches (dendrites) of the neurons and comprise the physical basis of synapses (connections between neurons).
The reader wishing to learn more about fragile XA syndrome in general is referred to Hagerman and Hagerman (2002), and to Bardoni et al. (2000) for a treatment of the psychopathology.
Incidence
Current best estimates of the prevalence of fragile XA syndrome, i.e., the full mutation with intellectual disability, are 1 in 4000 males and 1 in 6000 females (Turner et al., 1996).
Earlier estimates of much higher frequencies have not stood the test of time. There are limited data on the prevalence of females of normal intelligence who carry premutations and thus are at risk of having children with either a full mutation or a larger premutation. About 1 in 260 women in Quebec carry a premutation of 55–101 copies of the CCG repeat (Rousseau et al., 1995) and 1 in 250 women in Finland have more than 60 copies of the repeat unit (Ryynänen et al., 1999). In Israel, however, 1 in 110 women have 55 or more repeats: this high frequency was suggested as being due to a founder effect (Toledano-Alhadef et al., 2001). The prevalence of premutation carrier males who are at risk of having fragile X syndrome grandchildren is estimated at 1 in 5000 (Turner et al., 1996). Castellví-Bel et al. (2000) showed that the incidence of “intermediate” FRAXA alleles (41–60 repeats) was not different in retarded and in appropriate control populations.
|
|
|
Figure 14-1. Plain-stained sex chromosomes from a normal female (a) and male (b) compared with those of a fragile X female (c) and a fragile X male (d). The fragile site is arrowed. The segment distal to the fragile site appears as a satellite, as though it were about to break off. |
Intellectual disability associated with FRAXE is rare and probably affects less than about 1 in 100,000 individuals. It is, however, one of the more common forms of nonspecific X-linked mental retardation, with more than 50 families having been recorded (Gecz, 2000). As fragile X diagnosis has moved from cytogenetics to molecular genetics, FRAXE may come increasingly to be underdiagnosed
MOLECULAR GENETICS OF THE FRAGILE X LOCI
FRAXA
Fragile XA syndrome is a “dynamic mutation” disorder (Sutherland and Richards, 1993). At the fragile site there is a section of DNA comprising the triplet cytosine-cytosine-guanosine (CCG) repeated many times.1 The number of repeats determines the genotype of normality, gray zone, premutation, or full mutation (Table 14-1). Normal X chromosomes have from 6 to about 54 sequentially repeated copies of the CCG triplet, with copies above 40 being regarded as being in a “gray zone.” Normal carrier males, sometimes called “transmitting males,” have approximately 55–230 copies of the triplet-repeat (premutation). Males with the full mutation of more than 230 triplet-repeats (up to about 1000 copies) have fragile XA syndrome (Kremer et al., 1991; Oberlé et al., 1991; Yu et al., 1992). The situation is a little more complex for females, probably because they have two X chromosomes and one of these is inactivated. Female carriers with the premutation do not have fragile XA syndrome, but about half of those with the full mutation have some degree of intellectual disability (Thompson et al., 1994).
|
Table 14.1. Trinucleotide Lengths and Associated Phenotypes |
||||||||||||||||||||||||
|
||||||||||||||||||||||||
The CCG triplet-repeat sequence, when increased beyond a critical size (in the vicinity of 55 copies), is unstable. It can change in copy number when transmitted from parent to child. Thus, a mother with a premutation can transmit a full mutation to her child. A mother with a full mutation can transmit a larger or smaller full mutation to her child. A female with a gray zone allele can transmit an expansion of this allele to her child, into the premutation range, but not a full mutation. In contrast, fathers only transmit premutations to their children (daughters), regardless of whether they themselves carry a premutation or, very rarely, a full mutation. This process, whereby the initial change to the DNA sequence alters the chance of it undergoing further change, is termed “dynamic mutation” (Richards and Sutherland, 1992). The change from normal copy number to full mutation is a multistep process, proceeding through premutation steps presumably over several generations, rather than a single event characteristic of classical (or static) mutation.
The varying sizes of the fragile site DNA can be visualized on a Southern blot. Figure 14-2 shows the patterns seen in various types of individual and demonstrates the instability in a family; further examples are shown in Figure 14-3. If DNA is digested with the enzyme PstI and probed with pfxa3, the normal X chromosome gives a fragment of approximately 1.0 kb. This <1.0 kb band represents about 900 bp of DNA flanking the repeat plus the 18–165 bp of the actual CCG repeat sequence itself (which amounts to 6–55 triplet-repeats). The DNA fragment from a fragile XA chromosome, having within it the additional copies of the CCG repeat sequence, is larger by this amount. For example, in a person with 230 copies (230 × 3 = 690 bp) of the triplet-repeat, which is about where the premutation merges into the full mutation, the fragment is 1.6 kb (~900 bp + 690 bp) in size. This is an increase (“amplification”) of ~600 bp over the normal size of 1 kb.
Methylation of the FMR1 Gene
The mechanism by which the CCG repeat causes the molecular pathology that in turn leads to the fragile XA phenotype is not yet fully clarified, but hypermethylation appears to play a part (Sutcliffe et al., 1992). The repeat sequence is located within a noncoding portion (the 5-untranslated region) of the FMR1 gene (Verkerk et al., 1991). Once there are more than about 230 copies of the repeat the DNA surrounding the repeat and the repeat itself become hypermethylated. (Methylation is the addition of methyl groups to the C [cytosine] bases in CpG sequences in the DNA molecule.) Hypermethylation is associated with inactivation of the FMR1 gene containing the triplet-repeat sequence, and it ceases to be transcribed (Pieretti et al., 1991). The consequential lack of FMR1 gene product (FMRP) is presumed to be the cause of the abnormal phenotype of the fragile XA syndrome (Ashley et al., 1993). Methylated DNA is not cut by methylation sensitive restriction enzymes, and this property can be exploited in the laboratory. Determination of repeat copy number and, if necessary, assessment of methylation as an indicator of the activity state of the gene can be used to help predict phenotype.
|
|
|
Figure 14-2. Inheritance of the fragile XA unstable element in a four-generation lineage from a large affected pedigree. Chromosomal DNA was digested with PstI and probed with pfxa3. The control probe pS8 was included in the hybridization. Pedigree symbols: normal carrier female expressing the fragile XA on cytogenetic study ([circle with right half black]), affected fragile XA syndrome male expressing the fragile XA (▪); normal female (○). All carriers (⊙, ⊙) are obligate carriers. |
|
|
|
Figure 14-3. PstI-digested DNA probed with pfxa3 and pS8. (1) Carrier female, (2) transmitting male, (3) noncarrier female, (4) affected male, (5) carrier female, (6) affected male, (7) noncarrier female, (8) normal male, and (9) carrier female. |
Methylation status of the triplet-repeat and adjacent DNA in affected males cannot be resolved from the routine FRAXA PstI digestion DNA test. High-functioning affected males may have amplification values of between 0.6 and 1.0 kb above the normal size. If a more accurate genotype–phenotype correlation is required, analysis of methylation status is done by probing a double digest of EcoRI and EagI. EcoRI/EagI double digests of DNA from males within this range of amplification values and probed with pfxa3 are shown in Figure 14-4. The probes StB12.3 and Ox1.9 would give an identical result. Interpretation is simpler in males than in females. The normal chromosome of carrier females is also methylated in half the cells, on average, as part of the normal random X-inactivation process.
The rationale for methylation studies is as follows: EcoRI gives a normal band size of 5.2 kb. DNA of normal males is unmethylated and gives a 2.8 kb band from EcoRI/EagI double digestion. In carrier males the size of these bands is increased by the length of the (CCG)n repeat. However, amplification beyond 230 copies of the CCG repeat is usually associated with methylation. EagI does not cut methylated DNA, and EcoRI/EagI then gives only a band identical to that which would be obtained by EcoRI digestion alone (5.2 kb plus the size of the amplification). Males who are mosaic for premutation and full mutation copy numbers of the triplet-repeat give the combination of unmethylated and methylated patterns depending upon the fragment sizes. In females, chromosomes carrying the CCG amplification behave the same as in males, but there are the additional bands contributed by the normal X chromosome. This normal X is randomly inactivated and gives both a 5.2 kb (inactive X) and 2.8 kb (active X) fragment. These patterns have been fully described in Rousseau et al. (1991). More technical aspects of molecular diagnosis are discussed in Mulley and Sutherland (1994).
|
|
|
Figure 14-4. Restriction patterns for fragile XA males who show methylation differences. EcoRI digests are shown in odd-numbered lanes, EcoRI/EagI double digests are shown in even-numbered lanes. Normal control male, lanes 1–2 and 11–12. The normal EcoRI fragment is 5.2 kb and the normal unmethylated EcoRI/EagI fragment is 2.8 kb. Lanes 3 through 10 for affected individuals showfragments of higher than normal molecular weight because of amplification. (Females would exhibit additional complexity because additional methylated and unmethylated bands occur for the normal X chromosome.) The affected individual in lanes 3–4 is unmethylated. Lanes 5–6 show a methylation mosaic; most of his cells have the EagI restriction site unmethylated, but in some cells it is methylated. The individual in lanes 7–8 is fully methylated, and the one in lanes 9–10 unmethylated. The tissue tested was blood, which may not necessarily reflect methylation patterns in other tissues. (From Loesch et al., 1993, with permission of the American Society of Human Genetics.) |
While methylation may have a role, it is not the only process influencing the function of FMR1. In the absence of methylation, the gene is transcribed, but the mRNA is inefficiently translated, and the degree of inefficiency increases in proportion to repeat copy number. Feng et al. (1995) analyzed clones of cells having 207, 266, and 285 repeats: these produced, respectively, 24%, 12%, and essentially 0% of the normal amount of the FMR1 protein. Possibly, this reduction in protein is cell type specific and some tissues (for example, chorionic villi) with large unmethylated repeats can still make FMR1 protein. Tassone et al. (2000b) have challenged the widely accepted concept that hypermethylation of the repeat silences transcription of the FMR1 gene, and their work awaits confirmation.
Effect of FRAXA Premutations on Carrier Males
The FRAXA premutation was originally considered to be harmless, without any effect on the intellectual phenotype. But doubts began to creep in over the years, and it is now appreciated that in some individuals there can be subtle and in some, with age, obvious neurological compromise (Daly et al., 2001; Loesch et al., 2002; Leehey et al., 2003). The most severe involvement is expressed as a progressive neurological syndrome with cerebellar and parkinsonian features, accompanied by decline in executive function and cognition, with overt onset usually in late middle age. A consistent neuropathological pattern, with widespread neuronal intranuclear inclusions, was shown in four cases studied at postmortem, and equally the cerebellar neuroradiology is characteristic (Greco et al., 2002; Brunberg et al., 2002). In a retrospective study of families in New South Wales, based mostly on reports by their daughters of transmitting grandfathers, and using their fathers-in-law (the paternal grandfathers) as a comparison group, a significant excess of these “normal transmitting males” were shown to have developed such a neurological syndrome, mostly in their fifties and sixties (Rogers et al., 2003). The fractions were 15/81 (18.5%) in the premutation-carrying maternal grandfathers,2 compared with 4/73 (5.5%) in the paternal grandfathers, with an average age of 68 years in the former and 73 in the latter. These long-term effects on brain function may reflect overexpression of FMR1 mRNA from premutations, and findings in a mouse model are consistent with this view (Tassone et al., 2000b; Willemsen et al., 2003).
Effect of FRAXA Premutations on Carrier Females
Typically, females who carry FRAXA premutations are not cognitively impaired, but some recent lines of evidence point to subtle effects in some carriers. Imaging shows a slight diminution in brain volume, with variation of metabolic rates in certain regions (Murphy et al., 1999). Those with larger premutations, more than 100 copies of the CCG repeat, are more likely to show some of the psychological traits of full mutation carriers, including lack of skill in social interaction and depressed mood (Johnston et al., 2001). On measurement of the level of FRMP protein in peripheral blood, a correlation is seen with subtle deficits in certain specific areas of psychological functioning (Loesch et al., 2002). These several reports provide support for the proposition that cerebral functioning in the female premutation carrier is not necessarily entirely unscathed, although the effect may be subtle, and psychological test results would mostly remain within the range otherwise displayed in the family.
It had long been suspected that female pre-mutation (but not full mutation) carriers might be prone to early menopause, due to premature ovarian failure (POF). An international collaborative study concluded that such an effect did indeed exist. Among 395 premutation carriers, 16% had early (before age 40) menopause, compared with only 0.4% in a control group comprising noncarrier relatives (Allingham-Hawkins et al., 1999). Hundscheid et al. (2000a) give likelihoods for POF according to age as follows: 4% by age 30 years, and 25% by age 40 years.3
Two Types of FRAXA Mosaicism
Individuals with full mutations can show somatic instability of the amplified repeat sequence. Different cells in a single tissue can be genetically different, in terms of triplet-repeat length; and genetic differences can exist between tissues. This is manifest as a smear of DNA fragments on Southern blot (Fig. 14-3, lanes 4–6).
This instability can lead to two types of mosaicism. While, strictly speaking, individuals with smears of DNA in the full mutation size range are mosaics (there are different lengths of triplet-repeats), this term is reserved for two specific situations.
In the first type, mutational mosaicism, there are some cells with full mutations, which are fully methylated; and some cells with premutations, which are unmethylated and functional. The mental phenotype can vary, presumably depending on the type of mutation predominating in different parts of the brain, and thus the regional activity within cerebral tissues of the FMR1 gene. Up to 20% of fragile XA syndrome males are mutational mosaics detectable on Southern analysis and some may have an IQ within but at the lower end of the normal range. A rare type of mutational mosaic is the individual with some cells containing a normal number of repeat copies (6 to about 55) and others containing premutation or full mutation copy numbers. Polymerase chain reaction (PCR) is a more subtle tool and may indicate the presence of a premutation in a very small fraction of cells in many fragile XA males. De Graaff et al. (1995)were able to show on brain tissue from an affected male that about 1% of neurons expressed FMRP, and thus presumed that these individual neurons had a transcriptionally active premutation.
In the second type, methylation mosaicism, the number of repeat copies is characteristic of a full mutation, but the DNA is not methylated in all cells (Fig. 14-4, lane 6). This is less common than mutational mosaicism and occurs mostly at the lower end of the range of full mutation copy number. High-functioning fragile XA syndrome males have been described who are methylation mosaics with full mutations that are partly or completely unmethylated (Hagerman et al., 1994). This diminished DNA methylation correlates with a low level of cytogenetic expression of the fragile site (less than 5% of metaphases) and possibly a milder phenotype. These males have some FMR1 protein in cultured cells, and this may reflect the in vivo (specifically, cerebral) situation and explain their higher-functioning status.
Other Mutational Bases of FRAXA
Almost all cases of fragile XA syndrome are due to triplet-repeat expansion. A very few affected persons have a different mutational basis, and deletions (ranging in size from 1 bp to about 100 bp and up to megabase size) and point mutations in the FMR1 gene have been reported (Hammond et al., 1997). Grønskov et al. (1998) studied 118 intellectually disabled males who had returned a normal result on triplet-repeat testing, including 6 who had a typical FRAXA phenotype. On single-stranded conformational (SSCA) analysis, none had a pathogenic FMR1 mutation. Castellví-Bel et al. (1999) similarly screened 31 patients with a high clinical score for fragile XA syndrome and found no mutations. Mutations other than CCG expansions as a cause of fragile X syndrome are too rare to try and diagnose in a clinical setting.
Fragile XE
This fragile site has been much less studied than FRAXA. The mechanism is again a CCG repeat that expands through premutations to full mutations (Knight et al., 1993b, 1994; Hamel et al., 1994; Mulley et al., 1995). The CCG repeat is in the 5 untranslated region of a gene called FMR2, whose 9.5 kb transcript encodes a putative 1302 amnio acid protein. FMR2 transcription is silenced in males with CCG expansion and methylation of the adjacent CpG island (Gécz et al., 1996). Because of its milder phenotype, the diagnosis is less likely to be clinically suspected, although it is one of the more common forms of nonsyndromal mild X-linked mental retardation. In some FRAXE families there will be males who have full mutations who are not intellectually disabled, as the cognitive impairment associated with FRAXE overlaps the normal range (Gecz, 2000).
Gedeon et al. (1995) describe two intellectually disabled boys with deletions of part of the FRAXE gene, one of whom expresses a truncated FMR2 protein and has FRAXE syndrome, while the deletion in the other is confined to intronic sequences which may or may not affect FMR2 production (Gécz et al., 1996). Small deletions of the FMR2 gene have been found more frequently in women with POF than in association with mental retardation, and these may be responsible for up to 1.5% of POF (Murray et al., 1999).
|
|
|
Figure 14-5. Morphological characteristics of the fragile XA. Shown are expression of the fragile site on chromosomes at different stages of compaction (a–c); despiralization of the chromatin distal to the fragile site (d); double-satellite appearance at the site (e, f); expression of the fragile site in skin fibroblast metaphase (g); appearance of the fragile X on G-banding (h–j); early replication of the fragile X (thick arrow) by BrdU labeling, with late replication of normal X (thin arrow) in a single metaphase (k); and late replication of the fragile X (thick arrow) by BrdU labeling, with early replication of normal X (thin arrow) from a single metaphase (l). The appearances of fragile XE and fragile XF are identical to those of fragile XA. (Reproduced from Sutherland, 1983, with the permission of VCH.) |
Fragile XF
FRAXF is also due to dynamic mutation of a CCG repeat (Parrish et al., 1994), but in this case an apparently harmless one, in that it is not associated with a familial form of intellectual disability.
CYTOGENETICS
The fragile X chromosomes have a characteristic appearance (Fig. 14-5). The fragile sites on the X chromosome are not spontaneously expressed when most standard cytogenetics methods are used. They must be induced by one of various methods (Sutherland, 1991). These induction methods all lead to a relative deficiency of either thymidine or deoxycytidine at the time of DNA synthesis. Either of these conditions appears to be a requirement for the fragile site expression of fragile X chromosomes.
The fragile Xs are expressed cytogenetically in only a relatively small proportion of cells (10%–40% in most fragile X syndrome males) after the cells have been appropriately cultured. The proportion of cells expressing FRAXE and FRAXF may be higher than for FRAXA, especially in females with full mutations. Males with premutations do not express the fragile X cytogenetically; most of those with full mutations do. Some females with premutations express the fragile site in up to 10% of metaphases. This suggests that another requirement for fragile X expression is that the repeat sequence be hypermethylated.
There are four fragile sites recognized cytogenetically on the end of the long arm of the X chromosome (Fig. 14-6). Only one, FRAXA in band Xq27.3, is associated with fragile XA syndrome. FRAXD in band Xq27.2 is a common fragile site that can probably be induced on all X chromosomes. It is thus a part of normal chromosome structure and of no pathological significance. It is the only fragile site that can be distinguished from the others by the cytogeneticist, being clearly in band Xq27.2 (Sutherland and Baker, 1992).
GENETICS
FRAXA
Originally it was presumed that the fragile XA syndrome followed standard X-linked recessive inheritance, but atypical properties were soon recognized. The proportion of females with the fragile XA chromosome or who were obligate carriers of it and who exhibited features of the syndrome was high (on the order of 35%); perplexingly, there were normal male carriers. Segregation patterns of the fragile XA syndrome were examined by Sherman et al. (1984, 1985) when only cytogenetic testing was available. Sherman delineated the paradox that bears her name, stating that the incidence of fragile XA syndrome is higher in the offspring of daughters of normal carrier males than in the offspring of mothers of these males. Since the discovery of the molecular basis of FRAXA, the transmission of the triplet-repeat sequence has been studied in many families (Rousseau et al., 1991, 1994; Verkerk et al., 1992; Yu et al., 1992) and the inheritance of the syndrome is now well understood. A number of points have emerged:
1. No new mutation has been observed— that is, no individual with a full mutation has been observed as the offspring of parents with normal numbers of copies of the CCG repeat on their X chromosomes. The mothers of all fragile XA syndrome individuals are carriers of at least a pre-mutation, and where study has been possible, so is a grandparent. The rate of mutation from normal CCG copy number to premutation thus appears to be very low. Morris et al. (1995) provide a mathematical model for a progression from normal allele → “intermediate” allele → premutation → full mutation at a population level.
2. When the unstable sequence is transmitted by males it characteristically does not increase in size, and may decrease. In males with full mutations, only premutations are seen in sperm (Reyniers et al., 1993).
3. When the unstable triplet-repeat sequence is transmitted by females it usually increases in size (although rare decreases have been reported, and even more rarely, gene conversion events have reduced mutation-range fragile X chromosomes back to normal [Losekoot et al., 1997]). Women with small numbers of CCG repeats usually show less increase in size than women with larger numbers. Thus, women with less than 70 triplet-re-peats (“low-end” premutations) mostly have children who also have premutations (and these offspring are thus normal carrier sons and daughters), although the premutations in these sons and daughters are characteristically larger than those of their mothers. Women with “high-end” premutations (90 or more triplet-repeats) and carriers who themselves have a full mutation almost always transmit a full mutation (Table 14-2).
4. The mothers of “normal transmitting males” have a smaller number of triplet-repeats than do their granddaughters through these transmitting males. Thus, as per Sherman's observation, these mothers have fewer intellectually handicapped sons than do the daughters of the transmitting males; and the Sherman paradox is no longer paradoxical (Fu et al., 1991; Rousseau et al., 1991; Yu et al., 1992). This property of FRAXA, displaying increased penetrance of mental defect in the present generation compared with a previous generation, is a form of genetic anticipation.
|
|
|
Figure 14-6. The fragile sites in distal Xq. FRAXA, FRAXE, and FRAXF are folate-sensitive and can only be distinguished cytogenetically using FISH with probes known to map between them. FRAXD is a common fragile site. FRAXA, FRAXE, and FRAXF can be distinguished from each other by molecular methods. |
FRAXE
Not enough fragile XE families have been reported in detail for segregation analyses to be performed. Notably, and distinct from FRAXA, instances of a FRAXE full mutation being passed from father to daughter are on record (Hamel et al., 1994; Carbonell et al., 1996). This stands in contrast with the observation in one case of a full-mutation patient having small unmethylated expansions in sperm (Carbonell et al., 1996). A possible explanation (unconfirmed at the present time) is that the FRAXE premutation triplet sequence is particularly unstable in early postzygotic mitoses and can expand to a full mutation.
Rare Complexities
There are cases on record of women who are FRAXA compound heterozygotes having a full mutation on their maternal X and a premutation on the X from their father (Linden et al., 1999; Hegde et al., 2001). We have seen a family in which both FRAXA and FRAXE are segregating (Mulley et al., 1995). A FRAXE man married a FRAXA heterozygote and they had two daughters, one a FRAXE heterozygote and the other a FRAXE/FRAXA compound heterozygote. The latter, in turn, has had a son with fragile X syndrome, and FRAXE and FRAXA carrier daughters.
The fragile X phenotype can coexist with other abnormalities as part of a contiguous gene syndrome. Quan et al. (1995) report a child in whom a deletion in Xq26.3–q27.3 removed FMR1 and adjacent loci. His phenotypically normal mother showed selective inactivation of the deleted X.
The fragile X mutation can coexist with an abnormality on another chromosome, quite by chance. Such an example is offered by Missirian et al. (2000), who identified a mother heterozygous for (1) a FMR1 premutation and (2) a 22q11.2 deletion. Two of her children had DiGeorge syndrome (one also with an FMR1 premutation), and one had fragile X syndrome with a full mutation.
Diagnosis
The first line of laboratory investigation is molecular analysis, with PCR amplification of the CCG repeat and sizing by capillary electrophoresis. This can be used for both FRAXA and FRAXE. Males without a clear product in the normal size range and females who are not clearly heterozygous for two normal alleles should then be studied by Southern blot analysis for amplification of the repeat. For prenatal diagnosis, Southern blot is the gold standard and is all that is necessary. For FRAXA in this situation DNA should be cut with EcoR1 to compact diffuse smears that can be missed in a female fetus. In many jurisdictions, fragile XA testing is a routine and standard investigation of any child with developmental delay, male or female (although in our experience, very few families come to attention through a female index case). A more focused selection can be made according to certain phenotypic traits, some of which can be rather subtle and not always present (Lachiewicz et al., 2000).
Of course, routine chromosome analysis should still be performed, as this is part of the investigation of any child who is a candidate for fragile X testing. An option is to include fragile site analysis in the cytogenetic study. Immunohistochemical demonstration of FMRP is a methodology that has not been widely used as a diagnostic tool; it does not give definitive results in females or males with some FMRP expression, and it will not detect FRAXE syndrome. It may become a useful screening test for males but has less diagnostic power for females (Willemsen and Oostra, 2000).
FRAXF
FRAXF can be indirectly assumed in an individual with a cytogenetic fragile Xq28 chromosome by exclusion of FRAXA and FRAXE. This fragile X chromosome can then be recognized as having no clinical relevance. In a research laboratory, the FRAXF can be directly demonstrated by molecular study (Parrish et al., 1994; Ritchie et al., 1994).
Carrier Detection in FRAXA
Recognition of the FRAXA carrier is, in principle, unequivocal, since the CCG amplification can be directly demonstrated. The probes frequently used, all of which yield essentially the same results, are pfxa3, StB12.3, and Ox1.9. The long stretches of CCG repeats in full mutations and many of the pre-mutations are refractory to routine PCR amplification, and so routine diagnosis at the present time is primarily based on Southern analysis
Different Probes
The simplest detection system uses the probe pfxa3. Digestion with a single enzyme, PstI, gives a normal 1.0 kb fragment containing the CCG repeat and detects all premutations and full mutations. Alternative enzymes, such as EcoRI alone, which gives a 5.2 kb fragment, do not resolve the smaller premutations although they do permit easier visualization of dispersed smears. Note our caveat above on the use of Pst1; we have missed a very diffuse smear in a full-muta-tion fetus using this enzyme alone and now recommend using EcoR1 for prenatal diagnosis. The pfxa3 probe has been validated on a collection of families previously diagnosed by linkage and cytogenetics (Mulley et al., 1992). Neither StB12.3 nor Ox1.9 are able to detect the above-mentioned PstI fragment; their sequences do not overlap with the sequences within the PstI fragment containing the actual CCG repeat.
Results of hybridizing pfxa3 to PstI digests are shown in Figures 14-2 and 14-3. The restriction fragment containing FRAXA has a higher molecular weight because of the additional copies of the CCG repeat. The pS8 probe, an anonymous X-linked marker, is used as a control in double hybridization. It detects a 0.8 kb fragment in all individuals and confirms the presence of digested DNA in the track. Probe ratio is adjusted to obtain signal intensity approximately equal for each of the probes for DNA from normal individuals. This enables the number of non–fragile XA chromosomes in females to be estimated by inspection of relative intensity of pS8 and pfxa3 fragments. Although this dosage test is inadequate as a primary diagnostic procedure, it does represent a simple and valuable check to avoid missing carriers with an amplification consisting of a faint smear of fragments. (Affected males have no 1.0 kb band, and so there is no difficulty in diagnosis, however dispersed a smear).
Screening for Carrier Status in Women of Reproductive Age
The prevention of fragile X syndrome in further pregnancies in a family where an affected child has been identified is “secondary prevention.” Primary prevention would require that carrier women be detected before having a child with fragile X—in other words, in families without a proband. Toledano-Alhadef et al. (2001) describe their experience of offering screening to women in Israel. They tested 14,334 women without any family history of mental retardation and found 3 full mutation carriers and 127 premutations of 55 repeat copies or more. Among the premutation carriers who were, or who became pregnant, prenatal diagnoses were performed in 97, and 5 full-mutation fetuses were detected. This small risk for a full mutation (only 5%) may reflect different likelihoods for premutations to expand to a full mutation in cases ascertained other than through a fragile X syndrome proband (compare the two groups listed in Table 14-2). Alternatively, most of the premutations in this group may have been right at the low end of the CCG premutation range.
Decisions to conduct such screening programs will depend on funding of health care locally, and may be influenced by the frequency of carriers in the population (which is apparently higher in Israel than in other comparable populations). In a detailed economic analysis, based on the particular circumstances obtaining in The Netherlands, Wildhagen et al. (1998) determined that the cost of detecting one carrier is about $45,000, whether testing is conducted through prenatal clinics, by pre-conceptual screening, or through schools. This figure contrasts with lifetime costs of care for FRAXA males and females of about $960,000 and $530,000, respectively.
GENETIC COUNSELING
As always, good genetic counseling depends on accurate diagnosis. In the fragile X context, this means the molecular confirmation of the diagnosis of FRAXA and the molecular identification of carriers. As McConkie-Rosell et al. (1999) point out, women who are carriers often have strong views about the desirability of avoiding having affected children and want their relatives to be informed. People have different coping mechanisms to deal with the challenge of learning of their carrier status, and McConkie-Rosell et al. (2001) reproduce excerpts from a number of carriers they had interviewed, illustrating this fact. These authors have also addressed the issue of what might be a suitable age to discuss the question of the genetic risk and to actually offer testing: each individual needs to be judged on their merits, and “there is no universal ‘right’ age” (McConkie-Rosell et al., 2002).
Risks of Having an Abnormal Child
FRAXA
Male with Premutation
For carrier males (“normal transmitting males”) there is, according to current knowledge, no risk of having a mentally impaired child due to the FRAXA gene per se. All daughters receive the FRAXA gene in its premutation form, whatever its size in terms of triplet-repeats. None receives a full mutation and it is presumed that none shows somatic expansion. Thus, none would have fragile XA syndrome. All 46,XY sons receive their father's Y chromosome, and obviously his carrier state implies no genetic risk to them.
|
Table 14.2. Risks for Fragile X Premutation Carriers to Have a Child with a Full Mutation, if the Fragile X Chromosome is Transmitted. The Risk Becomes Higher with a Larger Expansion Size in the Mother |
|||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||
Male with Full Mutation
Procreation in this group is extremely rarely documented. Retrogression to a premutation in sperm dictates that daughters would be (other things being equal) of normal intelligence. Sons receive the Y chromosome.
Female with Premutation
The risk of transmitting the FRAXA X chromosome is 50%. If it is transmitted, the risk of having a child with fragile XA syndrome (in other words, a child with a full mutation) depends on the size of the mother's premutation (Table 14-2). Women with low-end premutations of less than 60 copies of the triplet-repeat have only a small risk of having a child with fragile XA syndrome. At the other end, for those with more than 90 copies (high-end pre-mutation), the risk is in the range of 90% to 100%. Triplet-repeats in the 40–60 (or perhaps 40–54) range are termed “gray zone” (Mogk et al., 1998; Patsalis et al., 1999).4 The practical and prudent conclusion is that for any woman in a fragile XA syndrome family who carries 55 or more copies of the repeat, the risk of having an intellectually disabled child is significant (while acknowledging that the risk for those in the 55–60 bracket may be very small). Indeed, there is no report of an allele of less than 55 copies expanding to a full mutation in a single generation. There is the additional concern of a neurodegenerative syndrome coming on in late middle age in some men who inherit a premutation (see p. 223 and p. 20), for which the level of risk remains to be determined.
The possibility of premature menopause should be brought to the attention of women who are premutation carriers (Sherman, 2000), and it might be advisable to have children earlier rather than later. Currently it is not possible to offer young women, as a routine, ovarian biopsy for cryopreservation of oocytes for possible future use. Of course an ovum donated by a noncarrier sister would avoid the genetic risk.
Female with Full Mutation
Offspring who inherit the fragile XA locus will all have full mutations. Hence, of their children with the fragile XA mutation, all the boys will have fragile XA syndrome, as will about 60% of the girls. The intellectual disability in these affected females is usually less severe than in the male.
FRAXE
The fragile XE story is still evolving and referral to a specialist in this area is necessary. For the time being, counseling should include the following caveats: (1) the penetrance of intellectual disability in males and females is unclear, but may be lower than for fragile XA; (2) the severity of intellectual disability is less in both sexes than for fragile XA; (3) variation between and within families appears to be considerable; and (4) although the inheritance pattern of fragile XE may be similar to that of fragile XA, there are apparently some differences, perhaps the most important being that daughters of fragile XE men can have full mutations.
The phenotypic consequences of inheriting a full mutation are unclear. There is some risk of females being mildly handicapped, which is as yet unquantified but probably less than that for fragile XA. Some males with full mutations appear to be relatively normal, although not all have had detailed assessments. Nevertheless, we presume the risk of significant intellectual impairment in the male to be substantial, albeit less than 100%. Until the situation becomes clearer, it may be best to offer prenatal diagnosis and present what is currently known to the couples involved. We await a more detailed interpretation of findings at prenatal diagnosis, but the picture is likely to be similar to that of FRAXA.
FRAXF
Fragile XF appears to be without phenotypic effect and may be an entirely harmless fragile site, although very few FRAXF families have been documented. The counselor will need to consult current literature and seek expert advice.
Prenatal Diagnosis of Fragile X Syndrome
Given the high genetic risks that may apply, preimplantation diagnosis (PGD) would seem, in principle, an attractive proposition. In practice, however, two issues arise. First, it may be difficult to obtain good numbers of ova from FRAXA carriers (note the comments above about premature ovarian failure). Second, single-cell PCR of even the normal-length repeat, let alone large expansions, is difficult, and thus distinction of hemizygous/heterozygous and unaffected status is difficult to make. The use of linked markers is an alternative approach. Thus, for most services, chorionic villus sampling (CVS) continues to be the mainstay of prenatal diagnosis. Pioneers working on PGD may develop more robust methodologies that in due course could be taken on board by the routine laboratory (Sermon et al., 1999, 2001; Platteau et al., 2002).
FRAXA
Prenatal diagnosis is offered to the female carrier. The essential prerequisite is the molecular confirmation of fragile X carrier status of the mother prior to CVS. Southern analysis requires a larger DNA sample and takes more time to complete than the many PCR-based molecular diagnoses carried out for a range of inherited disorders. Times conservatively quoted for a result from Southern analysis could be 2–3 weeks (compared with 2–3 days for PCR-based results). Sufficient DNA must be extracted from the CVS sample for at least one digest (approximately 10 µg) with sufficient additional tissue to initiate a cell culture as a source of backup DNA. Diagnosis is based on repeat length; methylation status of CVS can be misleading, and nonmethylation of CVS has been associated with methylation in fetal tissues (Castellví-Bel et al., 1995). Amniocentesis is not recommended: it is done at a later gestational stage, and then takes further time for cell culture to provide enough DNA. The possible outcomes of prenatal diagnoses for fragile XA are as follows:
1. A normal male fetus (1.0 kb band on pfxa3 probe; triplet-repeat copy number 6–54).
2. A male fetus with a premutation (up to 1.6 kb band; approximately 55–230 copiesof triplet-repeat) and thus a male carrier is predicted. There is a risk, the level of which is yet to be defined, for the pre-mutation ataxia syndrome.
3. An affected male fetus with a full mutation (band[s] <1.6 kb size and/or smear; greater than 230 triplet-repeats). The fragile XA mental retardation syndrome is predicted. The phenotype of an affected male fetus with copy number mosaicism, i.e., a mixture of full and premutations, cannot be accurately predicted from CVS; most cases would be affected to some extent.
4. A normal female fetus (1.0 kb band on pfxa3 probe and 5.2 kb band on EcoR1 digest to exclude diverse smears; triplet-repeat copy number 6–54).
5. A female fetus with a premutation (up to 1.6 kb band; approximately 55–230 triplet-repeat copies) and thus a female carrier is predicted. She would have a risk of premature ovarian failure.
6. A female fetus with a full mutation (band[s] > 1.6 kb size and/or smear; greater than 230 triplet-repeat copies). Mental impairment of variable degree is predicted in at least half of full-mutation females. There is some evidence that the size of the full mutation may be related to the level of intellectual functioning, but the variability is too great for this to have any predictive value.
For male carriers, who will obligatorily transmit the premutation to a daughter, prenatal diagnosis is not usually an issue, according to our present understanding that these daughters have no major mental defect due to their FRAXA carrier status.
FRAXE
There is little experience in the prenatal diagnosis of FRAXE, and expert advice should be sought.5
Notes
1. There is some confusion in the way trinucleotide repeats are expressed (Sutherland and Richards, 1993). Taking the convention that the bases be written in alphabetical order from 5 to 3, the FRAXA repeat is CCG, and the “anchoring triplet” (note 4) is CCT. Some authors “read” the opposite DNA strand, and write “CGG” and “AGG” instead.
2. Two of these premutation carriers also had a smear on Southern blot, indicating mosaicism for a premutation and a full mutation.
3. There may be an imprinting effect in premature ovarian failure. In a study of Dutch families, Hundscheid et al. (2000a) proposed that paternally inherited FRAXA pre-mutations were those responsible for early menopause. Their findings were not supported by others (Murray et al., 2000; Vianna-Morgante and Costa, 2000), whose subjects showed no difference in age of cessation of menses according to the parent of origin of their premutation. These workers and Hundschied et al. (2000b) have debated possible reasons for the discrepancies.
4. “Anchoring” CCT triplets in place of some CCG triplets may characterize stable transmission (Gacy et al., 1995; Zhong et al., 1995). Mogk et al. (1998)report a family in which two brothers had “gray zone” alleles with 47 repeats, and no interspersed anchoring CCT sequences. One showed stable transmission of the 47-repeat allele to his daughter, whereas the other showed a stepwise 1-repeat expansion from himself, to daughter, to granddaughter: these three had 47, 48, and 49 triplet repeats, respectively. (Elsewhere in the family, there were five males with fragile X syndrome.)
5. Carbonell et al. (1996) report diagnosing a male and a female fetus, both with full mutations. A striking finding was the presence of the fragile site in 25% of metaphases from amniotic fluid cell cultures; the conditions of culture were not stated. Both mothers continued their pregnancies because of uncertainties in interpreting the results.
