Fragile X Syndrome

A number sign (#) is used with this entry because fragile X (FXS) is caused by mutation in the FMR1 gene (309550). The vast majority of cases are caused by a trinucleotide (CGG)n repeat expansion (309550.0004) of greater than 200 repeats.

See also fragile X tremor/ataxia syndrome (FXTAS; 300623), which is caused by expanded FMR1 (CGG)n repeats that range in size from 55 to 200 repeats and are referred to as 'premutations.'

Description

Fragile X syndrome is characterized by moderate to severe mental retardation, macroorchidism, and distinct facial features, including long face, large ears, and prominent jaw. In most cases, the disorder is caused by the unstable expansion of a CGG repeat in the FMR1 gene and abnormal methylation, which results in suppression of FMR1 transcription and decreased protein levels in the brain (Devys et al., 1993).

Reviews

Fragile X syndrome accounts for about one-half of cases of X-linked mental retardation and is the second most common cause of mental impairment after trisomy 21 (190685) (Rousseau et al., 1995).

McCabe et al. (1999) summarized the proceedings of a workshop on the fragile X syndrome held in December 1998.

Jacquemont et al. (2007) provided a review of fragile X syndrome, which they characterized as a neurodevelopmental disorder, and FXTAS, which they characterized as a neurodegenerative disorder.

Clinical Features

Lubs (1969) reported a family in which 4 males spanning 3 generations had mental retardation. Cytogenetic studies showed an unusual constriction of the long arm of the X chromosome in 10 to 33% of cells. In a follow-up report of the same family, Lubs et al. (1984) noted that affected individuals had large testes, low-set large ears, and asymmetric facial features with prominent angle of the jaw.

Cantu et al. (1976) reported 4 male sibs with congenital bilateral macroorchidism and severe mental retardation. Detailed endocrinologic evaluation, including sperm analysis, indicated normal testicular function.

Mattei et al. (1981) reported 20 patients from 15 unrelated families with fragile X syndrome. In all 19 affected male and 1 affected female proband, the fragile X site was detected in 10-61% of lymphocyte or fibroblast cells; there seemed to be no correlation between the frequency of the fragile site and clinical severity. Three sisters of probands were mildly affected, but carrier females were unaffected. Affected male individuals showed characteristic facies, including long face, high forehead, midface hypoplasia, large mouth with long upper middle incisors, thick lips, high-arched palate, large jaw with prominent chin, and large, poorly formed ears. None of 14 prepubertal males showed macroorchidism. Mental retardation was very variable, but language development was usually very delayed. Motor development was often delayed. Most showed unusual behavior, with alternating anxiety and hilarity, disordered hyperactivity, and aggressiveness.

Lubs et al. (1984) reported a large African American kindred in which 10 males had mental retardation and macroorchidism associated with the abnormal X chromosome marker. Other variable clinical features included asymmetric facies and large hands. Six females were similarly affected. Meryash et al. (1984) studied 18 affected males, aged 18 to 69 years. Of 15 subjects, 13 had macroorchidism. Average height was less than published standards. Of the 18 subjects, 17 had absolute or relative macrocephaly and 12 were dolichocephalic.

Jacobs (1982) encountered a man and Daker et al. (1981) reported 2 brothers with marXq28 and average intelligence. Similarly, Fryns and Van Den Berghe (1982) presented a kindred in which the fragile X chromosome was transmitted by at least 3 normal males. These men died at ages 68, 72, and 76 years and had a normal phenotype with normal intelligence; one was an administrator and 2 were officers. Voelckel et al. (1988) reported 3 brothers with the fragile X; only 2 were mentally retarded. Johnson et al. (1991) described a large kindred with 10 mentally retarded, fragile X-positive males, and 2 normal transmitting males. Pellissier et al. (1991) also described a kindred with 2 normal transmitter brothers.

Loesch and Hay (1988) presented the clinical findings on 113 fragile X female heterozygotes from 44 families. In 85% of a subsample of 92 adult females, the nonverbal IQ score was 85 or less. Verbal ability deficits were much less common. Typical facial characteristics, irregular teeth, and hypermobility of finger joints occurred in approximately 40% of adult females, but facial abnormalities were less common in children. Although the frequency of miscarriages was increased, a moderate increase in the number of children was found in female carriers with borderline intellectual impairment. The question of whether ovarian size is increased in females with the fragile X was addressed by Goodman et al. (1987).

A Prader-Willi-like subphenotype of the fragile X syndrome was described by de Vries et al. (1993). Clinical features included extreme obesity with a full, round face, small, broad hands and feet, and regional skin hyperpigmentation. Unlike the Prader-Willi syndrome (176270), the patients lacked the neonatal hypotonia and feeding problems during infancy followed by hyperphagia during toddlerhood. In a group of 26 patients with suspected Prader-Willi syndrome but without detectable molecular abnormalities of chromosome 15, one fragile X patient was found. General overgrowth was described in 4 fragile X patients, all of whom came from families with other affected relatives who showed the classic Martin-Bell phenotype (de Vries et al., 1995). Schrander-Stumpel et al. (1994) found the FMR1 mutation in a 3-year-old boy with unexplained extreme obesity and delayed motor and speech development. They compared the clinical features with those in 9 reported patients with the fragile X syndrome and extreme obesity. They suggested that behavioral characteristics such as hyperkinesis, autistic-like behavior, and apparent speech and language deficits may help point toward the diagnosis of the fragile X syndrome.

Limprasert et al. (2000) described unilateral macroorchidism in a boy with fragile X syndrome and discussed the possible explanations.

Backes et al. (2000) evaluated a group of boys with fragile X syndrome, ascertained by molecular genetic methods, to determine a cognitive and behavioral profile. The cognitive phenotype revealed a general intelligence corresponding to mild to moderately severe mental retardation. Psychiatric comorbidity was high, and attention deficit hyperactivity disorder (ADHD), oppositional defiant disorder, enuresis, and encopresis predominated. No significant correlation between the specific features of the phenotype and genotype were found.

Other Features

Stigmata of connective tissue abnormalities in fragile X syndrome have been reported, including finger joint hypermobility, instability of other joints (Opitz et al., 1984; Hagerman et al., 1984), and mitral valve prolapse (Pyeritz et al., 1982). Hagerman and Synhorst (1984) not only confirmed mitral valve prolapse but also demonstrated mild dilatation of the ascending aorta.

Davids et al. (1990) found that of 150 male patients with the fragile X syndrome, 75 had flat feet, 85 had excessive laxity of joints, and 10 had scoliosis. In 29 of the patients, flat feet had been evaluated or treated by an orthopedic surgeon before the diagnosis of fragile X syndrome had been made.

Langenbeck et al. (1984) found that mean corpuscular hemoglobin was increased in this disorder and asked whether this was a reflection of a defect in folate metabolism.

Rodewald et al. (1987) described ganglioglioma of the cauda equina in a 17-year-old male with familial Martin-Bell syndrome. Because of the association of neoplasms with autosomal chromosome abnormalities, Rodewald et al. (1987) suggested that this may be more than coincidence. However, they found no published reports of tumors associated with Martin-Bell syndrome.

Fryns (1993) noted periorbital hyperpigmentation and scrotal hyperpigmentation about the time of puberty.

Reiss et al. (1994) showed that the volume of the hippocampus was enlarged in fragile X patients compared to controls. Furthermore, there was an age-related increase in the volume of the hippocampus and an age-related decrease in the volume of the superior temporal gyrus. In another study, Reiss et al. (1991) showed that fragile X males had a significantly decreased size of the posterior cerebellar vermis and increased size of the fourth ventricle, when compared with age- and sex-matched groups of fragile X-negative, developmentally disabled subjects and individuals with normal IQ. Reiss et al. (1991) showed that young fragile X females had decreased size of the posterior cerebellar vermis and increased size of the fourth ventricle, when compared with normal age-, sex-, and IQ-matched females. The findings were intermediate between those of the fragile X males and the non-fragile X control groups.

Jakala et al. (1997) found that males with the full fragile X (fM) mutation showed worse cognitive performance than did males with the premutation (pM); deficits in females with the fM were qualitatively similar but less severe than in males with the fM. In a visual memory test, both fM groups were impaired. Hippocampal volumes normalized for intracranial or brain area did not significantly differ between fM and pM groups. Minor abnormalities in temporal lobe structures were found by MRI in fM subjects.

Lachiewicz et al. (2000) compared physical characteristics of young boys with fragile X with those of a control group. After adjustment for multiple comparisons, only 4 of 42 characteristics studied differed significantly in their distributions between the 2 groups. These included adverse response to touch on the skin, difficulty touching the tongue to the lips, soft skin over the dorsum of the hand, and hallucal crease. Ten other characteristics were identified that may also have predictive value for fragile X syndrome.

Gould et al. (2000) compared sleep patterns and endogenous melatonin profiles in 13 boys with fragile X to 8 age-matched normal controls. Results showed greater variability in total sleep time, difficulty in sleep maintenance, and significantly greater nocturnal melatonin production in the boys with fragile X.

Koekkoek et al. (2005) observed a severe defect in eyeblink conditioning in 6 patients with fragile X syndrome, indicating a deficit in cerebellar motor learning. Fmr1-null mice also showed deficits in classic delay eyeblink conditioning, and Fmr1-null mouse cerebellar Purkinje cells showed elongated irregular dendritic spines and enhanced long-term depression induction at the parallel fiber synapses that innervate these spines. The findings indicated that a lack of FMRP leads to cerebellar dysfunction.

Moro et al. (2006) reported 2 unrelated boys with fragile X syndrome who had periventricular heterotopia on brain MRI. One had 3 heterotopic nodules, and the other had a single nodule. The findings suggested that abnormal neuronal migration contributes to the neurologic phenotype.

Gothelf et al. (2008) compared the neuroanatomy of 84 children and adolescents with fragile X syndrome to 72 control individuals using various MR imaging methods. Although there was no difference between the groups for total brain volume, separate analysis for different brain regions showed that patients had significantly smaller cortical lobes, significantly increased size of the caudate nucleus, and decreased size of the posterior cerebellar vermis, amygdala, and superior temporal gyrus compared to controls. The combination of a large caudate with small posterior cerebellar vermis, amygdala, and superior temporal gyrus could distinguish children with fragile X syndrome from control subjects with high sensitivity and specificity. Large caudate and small posterior cerebellar vermis were associated with lower FMRP levels and more pronounced cognitive deficits and aberrant behaviors, including autistic features. Gothelf et al. (2008) suggested that abnormal development of the prefrontal-striatal pathway and the orbitofrontal-amygdala circuit characterizes a neuroanatomic phenotype associated with fragile X syndrome.

In a systematic chart review of over 500 patients with fragile X syndrome, Berry-Kravis et al. (2015) found that male patients had significantly reduced levels of total cholesterol, low density lipoprotein (LDL), and high density lipoprotein (HDL) compared to matched controls. These findings were not related to body mass index (BMI).

Male Premutation Carriers

Some boys with expanded FMR1 (CGG)n repeats that range in size from 55 to 200 repeats, referred to as 'premutations,' may exhibit similar, but possibly milder, clinical features to those with full expansions. Aziz et al. (2003) reported the clinical features of 10 boys with FMR1 CGG expansions between 45 and 198 repeats. Most had increased testicular volume and enlarged outer canthal distance, and most exhibited variable defects in social impairment, speech and language deficits, autistic features, hyperactivity, and/or developmental abnormalities.

Chonchaiya et al. (2012) examined 50 boys with FMR1 premutations for seizures and autistic features. Twenty-five boys were found to carry a premutation after direct referral for developmental issues ('probands'), and 25 additional boys were found to carry a FMR1 premutation after testing following identification of a family member with either the full mutation or a premutation ('non-probands'). The mean age of both groups was 9 years. All individuals with the premutation had increased FMR1 mRNA compared to sibs without a premutation. The probands with a premutation were significantly more likely to have features of autism spectrum disorder (68%) and seizures (28%) compared to controls (1.7% and 1% for autism spectrum disorder and seizures, respectively). Although none of the non-probands with premutations had seizures, 28% had features of autism spectrum disorder. Chonchaiya et al. (2012) concluded that boys with the FMR1 premutation should be assessed for autistic features and seizures.

Female Premutation Carriers

For a discussion of premature ovarian failure (POF) associated with premutation in the FMR1 gene, see POF1 (311360).

Rousseau et al. (1991) observed an age-dependent phenomenon: the full fragile X mutation was found preferentially on the inactive X in leukocytes 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 suggested that young females who show preferential presence of a full mutation on the active X in leukocytes may be at increased risk for mental retardation. There is a known decrease with age of the expression of the fragile site.

Steyaert et al. (2003) used the Sonneville Visual Attentions Task (SVAT) method to assess reaction time on different tasks in 3 groups of female subjects: premutation carriers, full mutation carriers, and control subjects. Their findings supported earlier findings that the fragile X premutation may affect neurocognitive function, in particular aspects of attention.

Hundscheid et al. (2003) investigated whether premutation carriers have an increased risk for diseases other than POF. Among 264 women from fragile X families, they found no statistically significant differences in the occurrence of diseases known to be associated with menopause, such as cardiovascular diseases and osteoporosis; however, lower bone mineral density was observed only in premutation carriers. Once a premutation carrier experiences premature ovarian failure, she is at risk for early estrogen deprivation which, if not treated, may lead to premature decrease in bone density.

Hunter et al. (2008) found no significant differences in neuropsychologic testing scores between 63 males under the age of 50 who were carriers of intermediate (20 to 55 repeats) or premutation (55 to 199 repeats) FMR1 alleles compared to 75 male controls. A comparison of 389 female intermediate or premutation allele carriers showed an association with increasing repeat length and self-reported attention difficulties compared to 117 female controls, but there were no differences in the other neuropsychologic testing scores.

Pathogenesis

Colak et al. (2014) demonstrated that FMR1 (309550) silencing is mediated by the FMR1 mRNA. The FMR1 mRNA contains the transcribed CGG-repeat tract as part of the 5-prime untranslated region, which hybridizes to the complementary CGG-repeat portion of the FMR1 gene to form an RNA/DNA duplex. Disrupting the interaction of the mRNA with the CGG-repeat portion of the FMR1 gene prevents promoter silencing. Colak et al. (2014) concluded that their data linked trinucleotide repeat expansion to a form of RNA-directed gene silencing mediated by direct interactions of the trinucleotide repeat RNA and DNA.

Cytogenetics

Lubs (1969) first described an abnormality of the distal long arm of the X chromosome, Xq, in 4 mentally retarded males from a single family. A secondary constriction of the chromosome gave the appearance of large satellites. Lubs (1969) suggested that either the anomalous region itself or a closely linked recessive gene might account for X-linked mental retardation. This observation went unconfirmed for years until cytogeneticists reverted to a folate-deficient medium for tissue culture such as Lubs (1969) employed. Appearance of this secondary constriction, referred to as a 'fragile site,' was localized to Xq27-q28 and was shown to be dependent on folate deficiency in the culture medium, which leads to deficiency of thymidine monophosphate (Giraud et al., 1976; Harvey et al., 1977; Sutherland, 1977). Sutherland was in Melbourne when he made his initial observations on the fragile X. When he went to Adelaide, he upgraded his laboratory, changing from 199 to F10 culture medium to give better chromosomes for banding. The failure to find the fragile X with the new medium led to his discovery of the critical role of folate (Gerald, 1983).

At least 12 other heritable secondary constrictions ('fragile sites') on other chromosomes were proved by the early 1980s (Sutherland, 1981; Hecht et al., 1982), but none had an association with a particular phenotype. In all pedigrees of marXq28 studied, no crossing-over between the marker and mental retardation had occurred. This suggested that the marker, rather than being closely linked to a gene causing mental retardation, is a direct cytologic indicator of the genetic mutation causing this phenotype (Kaiser-McCaw et al., 1980).

Turner et al. (1978) suggested labeling the marker 'secondary constriction Xq27'; however, convention requires that 'a break suspected at an interface between two bands is identified arbitrarily by the higher of the two band numbers' (ISCN, 1978; section 2.4.4.2). Brookwell and Turner (1983) again concluded that the fragile site is in band Xq27, close to the q27-q28 interface.

Lubs (1969) and Martin et al. (1980) found that the fragile X marker was not preferentially inactivated in female heterozygotes.

In a survey of retarded females who had no obvious physical abnormalities, Turner et al. (1980) found that 7% expressed marXq28 in lymphocytes. Among obligate heterozygotes, the likelihood of detecting marXq28 correlated with severity of retardation (Howard-Peebles and Stoddard, 1980; Jacobs et al., 1980). In 2 heterozygous sisters who were slow learners, Uchida and Joyce (1982) found that the fragile X was active in approximately 70% of cells, whereas 2 heterozygous relatives of normal intelligence had the fragile X active in approximately 30 to 50% of cells.

An earlier suggestion that the proportion of cells exhibiting marXq28 decreases with increasing heterozygote age (Sutherland, 1979; Jacobs et al., 1980; Turner et al., 1980) may have been an artifact due to ascertaining fewer retarded women in older age groups (Jacobs et al., 1982).

Inheritance

See review by Nussbaum and Ledbetter (1986).

In 4 of 27 large fragile X pedigrees, Fryns (1984) found strong evidence of transmission by normal males.

By analysis of multiple families with fragile X syndrome, Sherman et al. (1985) identified multiple special inheritance characteristics that differed from other X-linked traits. All mothers of affected sons were carriers; no new mutations were predicted. The risk of fragile X syndrome in offspring depended upon the sex and phenotype of the carrier parent. Daughters of nonpenetrant transmitting males were rarely affected, whereas daughters of normal carrier females had a higher chance of having affected daughters. Cognitively impaired females had a higher risk of having affected offspring. Mothers and daughters of transmitting males who were phenotypically similar had sons and daughters who expressed the gene differently. The gene seemed to be more penetrant in the offspring of daughters of transmitting males than in offspring of mothers of transmitting males. Sherman et al. (1985) suggested that a premutation exists which generates a definitive mutation only when transmitted by a female and that there is a submicroscopic rearrangement at Xq27.3 which per se causes no trouble but generates a significant genetic imbalance when involved in a recombinational event with the other X chromosome.

Pembrey et al. (1985) advanced a premutation hypothesis to explain unusual characteristics of the genetics of this disorder: transmission occurs through normal males; the heterozygous daughters of such males are never mentally retarded and have few or no fragile sites, and by contrast in the next generation, a third of heterozygous females are mentally subnormal with an average of 29% fragile sites (Sherman et al., 1985). This came to be called the Sherman paradox (Fu et al., 1991).

Winter and Pembrey (1986) analyzed linkage relationships of flanking genetic markers in daughters of normal transmitting males. There was a significant reduction in recombination in meioses giving rise to affected grandsons of normal transmitting males, as compared to families with no apparent normal transmitting males. One interpretation offered was interference related to a recombinational event leading to the full fragile X mutation.

Weaver and Sherman (1987) gave guidelines for counseling families with the Martin-Bell syndrome. Because of the peculiarities of the pedigrees, it is necessary to give different estimates for the risk among the sons and daughters of normal carrier mothers, mentally impaired carrier mothers, and normal transmitting males. Among the sons, the probability for mental impairment is 0.38, 0.5, and 0, respectively, and the chance of a son being a mentally normal carrier is 0.12, 0, and 0, respectively. Among the daughters, the risk of being a mentally impaired carrier is 0.16, 0.28, and 0, respectively, and the chance of being a mentally normal carrier is about 0.34, 0.22, and 1.0, respectively. Given a sporadic case in a male with no fragile X demonstrable in the mother, the estimates for occurrence in a brother of the proband vary from 9 to 27%, depending on the theoretical model used; the estimated risk in first cousins varies from 0.01 to 0.05.

Having excluded a mutation rate in male germ cells of the magnitude required by an exclusive mutation hypothesis to explain the high incidence of the fragile X syndrome, Vogel et al. (1990) proceeded to demonstrate an increased fitness of heterozygous females by a comparison with the reproductive performance of 'adequate' controls (mothers and grandparents of Down syndrome patients). Estimates ranged between 1.11 and 1.36. A higher incidence of dizygotic twinning suggested a biologic component for this increased fertility. On the other hand, the fragile X families had a significantly lower social status than the controls, suggesting a sociopsychologic component of their higher fertility.

Oberle et al. (1991) found that the transition from a premutation to a full mutation occurred only after passage through a female.

Yu et al. (1992) found that all individuals with the fragile X genotype had a parent with an amplified p(CCG)n repeat, indicating that few, if any, cases of fragile X syndrome are not familial.

Tabolacci et al. (2008) reported a 10-year-old boy with a normal CGG tract in the FMR1 gene and no fragile X syndrome phenotype; however, his 2 brothers were affected with fragile X syndrome due to an expanded allele. The mother carried a premutation allele of about 190 CGG. The 10-year-old unaffected boy was found to have an allele of 43 repeats with an unusual configuration detected using 2 different restriction enzymes, and the boy was not mosaic. Haplotype analysis proved that the rearranged allele originated from the maternal expanded allele, indicating contraction of the expanded CGG tract and reversion to a normal size FMR1 allele.

Imprinting

Laird (1987) proposed that abnormal chromosome imprinting is involved in inheritance of the fragile X syndrome. Two independent events are required for expression of the syndrome: the fragile X mutation and X chromosome inactivation in pre-oogonial cells. According to this model, the fragile X mutation leads to an imprint, or stable inactivation of a gene or genes at the fragile X site because the mutation prevents reactivation of a mutant fragile X chromosome that had been inactivated in a female for dosage compensation. This block to reactivation leads to mental retardation in progeny by reducing the level of products from the unreactivated region in the male's cells, and for a heterozygous female, in somatic cells in which the normal X chromosome has been inactivated. The basis of this localized block to complete reactivation of a fragile X chromosome was proposed to be late replication of DNA at the fragile site (Laird et al., 1987).

From an analysis of data on fragile X, Laird et al. (1990) concluded that 2 progenitor cells for human oogonia may be present at the time of the initial event that leads to chromosome imprinting. The estimate was based on the fact that one-half of the female's primary oocytes would, on the average, be expected to show imprinting if X-chromosome inactivation is the initial step. The population genetic predictions of the 'X-inactivation imprinting' model indicate that the fraction of carrier males who are nonpenetrant (nonimprinted) would be about 0.5 at equilibrium (Sved and Laird, 1990). Sved and Laird (1988) suggested that this predicted fraction is higher than the reported fraction of 0.2 (Sherman et al., 1985) because of an unusual ascertainment bias.

Laird (1991) explained the cytogenetic disappearance of the fragile X site in the few daughters of affected males that have been reported as a consequence of erasure of the imprint when it is passed through males. Erasure of chromosome imprinting often occurs when the imprinted chromosome is passed through the parental gender opposite from the gender that established the imprint. Reimprinting apparently can occur, however, in primary oocytes of these daughters.

Follette and Laird (1992) examined the stability of the imprinted state, defining stability as 100% penetrance of the syndrome in sons who receive an imprinted chromosome from the mother. In a preliminary estimate, they concluded that the fragile X imprint was stable in 46 of 48 female meioses, giving a tentative estimate of about 96% for the stability of the imprint.

Kirkilionis et al. (1992) presented the pedigree of a large family that illustrated dramatically the Sherman paradox and was compatible with the predictions of the Laird X-inactivation imprinting model.

Zeesman et al. (2004) reported a family in which a fragile X mosaic male, with both premutation and full mutation alleles in his peripheral blood leukocytes, had a daughter with both premutation and partially methylated full mutation alleles and a significant developmental disability. The sperm cells in the father contained only alleles in the premutation range; because the daughter had both premutation and full mutation alleles, the expansion to full mutation must have occurred postzygotically. The authors believed this to be the first report of a paternally derived full mutation expressed in a female. Steinbach and Steinbach (2005) disputed the conclusion of Zeesman et al. (2004) of paternal transmission of fragile X syndrome, and Tassone et al. (2005) provided a response.

Mapping

In 6 of 18 informative Sardinian pedigrees with fragile X syndrome, Filippi et al. (1983) found close linkage with G6PD (305900) and deutan colorblindness (CBD; 303800), both linked to Xq28. The maximum likelihood estimate of recombination was 6% with 90% fiducial limits between 2.5 and 19.5% and odds favoring linkage of 428:1. There was no linkage between G6PD and the Renpenning form of X-linked mental retardation (RENS1; 309500) on Xp11.

Camerino et al. (1983) found close linkage between the factor IX locus (F9; 300746) on Xq27 and fragile X syndrome in a large affected family (lod score of 4.02 at a theta of 0.05 for Xq27). In addition, they demonstrated transmission of the disorder through a phenotypically normal male. They observed no meiotic recombination out of 17 opportunities.

Szabo et al. (1984) determined that the G6PD locus is distal to the fragile X locus on Xq27.3. Although both G6PD and F9 have been linked to fragile X, F9 has been shown to segregate independently from deutan and protan (CBP; 303900) color blindness in some families. Szabo et al. (1984) concluded that the Xq27 region is a 'hotspot' for meiotic recombination; that the microscopically detectable change in fragile X syndrome is probably a minute chromosomal aberration resulting from an inaccurate recombination event; and that recombination is suppressed at the Xq27.3 region in heterozygous females.

Using intragenic RFLPs of factor IX in the study of 3 families with the fragile X syndrome, Forster-Gibson et al. (1985) found a minimum of 4 recombinations in 9 meioses. A maximum lod score of 2.75 at theta 0.20 was estimated. The authors concluded that the genetic distance between fragile X and factor IX was too great for factor IX probes to be useful for carrier detection of fragile X syndrome.

Warren et al. (1985) reported a family in which 2 brothers with fragile X mental retardation had different factor IX RFLPs, indicating that a recombinational event occurred between the 2 loci. Brown et al. (1985) found that pedigrees with nonpenetrant males showed tight linkage to factor IX, whereas the linkage was loose in those pedigrees with full penetrance in males. Giannelli et al. (1987) found that families with nonpenetrant carrier males showed tighter linkage to factor IX than did the others and suggested the existence of 2 fragile X loci. In a multilocus linkage analysis of 147 families, Brown et al. (1988) found significant variation in the recombination distance between F9 and FRAXA. Heterogeneity testing showed that 20% of the families had tight F9-FRAXA linkage, whereas 80% demonstrated loose linkage, with an average recombination distance of 0.35. On average, the multipoint distances found were DXS51-F9, 6.9%; F9-FRAXA, 22.4%; FRAXA-DXS52, 12.7%; and DXS52-DXS15, 2.2%. In 14 families with fragile X and 9 normal pedigrees from the CEPH collection, Thibodeau et al. (1988) also observed genetic heterogeneity between the fragile X locus and the F9 locus, with recombination frequencies of DXS51-F9, 0%; F9-DXS52, 45%; DXS51-FRAXA, 15%; F9-FRAXA, 18%; DXS98-FRAXA, 36%; and DXS52-FRAXA, 15%. The authors proposed the relative order for the 5 loci as DXS51, F9, DXS98--FRAXA--DXS52.

Using a 275-kb fragment of human DNA isolated in a yeast artificial chromosome (YAC) and thought to span the fragile site, Yu et al. (1991) derived 2 probes that spanned the fragile site as demonstrated by in situ hybridization. Mapping delineated further the sequences that appeared to span the fragile site to about 15 kb. A 5-kb EcoRI fragment was found to contain fragile site breakpoints. When this fragment was used as a probe on the chromosomal DNA of normal and fragile X individuals, alterations in the mobility of the sequences were found only in fragile X DNA. These sequences were of an increased size and varied within families, indicating that the region was unstable. The results were consistent with those of Oberle et al. (1991).

Richards et al. (1991) used microsatellite markers to position the fragile X locus within the multipoint map of the X chromosome to a position 3.7 cM distal to DXS297 and 1.2 cM proximal to DXS296. They described 2 polymorphic microsatellite AC repeat markers, FRAXAC1 and FRAXAC2, physically located within 10 kb and on either side of the (CCG)n repeat responsible for the fragile site. The 2 markers showed strong linkage disequilibrium and have heterozygosity of 44 and 71%, respectively. No recombination was observed either between these markers in 40 CEPH pedigrees or with FMR1 in affected pedigrees.

Diagnosis

Jacky and Dill (1980) detected the fragile X chromosome in cultured lymphocytes and fibroblasts from affected patients. Glover (1981), Tommerup et al. (1981), and Jacobs et al. (1982) demonstrated that pharmacologic inhibition of thymidylate synthetase (TYMS; 188350) was effective in inducing the fragile X marker in cell cultures. Snyder et al. (1984) showed that culture conditions that promote expression of the fragile X site do not affect expression of lymphocyte HPRT but do cause a marked reduction in G6PD activity.

Sutherland (1989) indicated that there is a fragile site (FRAXD) located at Xq27.2, separate from the classic FRAXA site at Xq27.3 which is responsible for mental retardation. The FRAXD is inducible by high doses of aphidicolin. Ramos et al. (1992) concluded that the fragile site at Xq27.2 can be demonstrated in normal persons under the conditions of thymidylate stress routinely used for cytogenetic diagnosis of the fragile X syndrome. Furthermore, this fragile site is present at low levels (1-2%) in all persons who express it and, therefore, its expression is unlikely to cause false-positive diagnoses of the syndrome. Lesions at Xq26 are also seen at low levels in lymphocytes of persons without the syndrome.

Griffiths and Strachan (1991) described a technique, based on a culture system reported by Wheater and Roberts (1987), that enabled the cytogeneticist to do fra(X) screening and prometaphase banding on the same specimen.

Using restriction enzymes, Oberle et al. (1991) detected abnormally large-sized fragments and abnormal methylation around the fragile X site in affected males and carrier females. Some affected males appeared to be mosaics, with coexistence of a large methylated fragment and a smaller normal unmethylated fragment. Rare apparent false negatives were considered to be the result of genetic heterogeneity or misdiagnosis.

Rousseau et al. (1991) concluded that direct DNA diagnosis of the fragile X syndrome is efficient and reliable. Southern analysis of EcoRI and EagI digests of DNA distinguished clearly in a single test between the normal genotype, the premutation, and the full mutation. All 103 affected males and 31 of 59 females with full mutations had mental retardation. Fifteen percent of those with full mutations had some cells carrying only the premutation. All of the mothers of affected children were carriers of either a premutation or a full mutation. Because of the certainty of DNA diagnosis, this method replaced cytogenetic detection of the fragile X chromosome, which carries a rate of misdiagnosis of about 5% for both false positives and the more frequent false negative conclusion, and diagnosis by the linkage principle, which gives a probabilistic result rather than an absolute one. Jacobs (1991), however, stated that the cytogenetic marker still had an honorable role to play in the diagnosis of fragile X syndrome. It was reliable for virtually all males and for the majority of affected females and was the most efficient and cost effective methodology at that time.

Mandel et al. (1992) reported on the Fifth International Workshop on the Fragile X and X-Linked Mental Retardation held near Strasbourg, France, in August 1991. In addition to their summary, over 50 papers on the fragile X syndrome and 18 papers related to other X-linked mental retardation syndromes presented at the conference were published in the American Journal of Medical Genetics. Mandel et al. (1992) reviewed the hypothesis of Patricia Jacobs which postulated 3 mutations: a change from a normal insert (N) to a small insert that is at low risk of converting to a large insert (S); a change from that type of small insert to a small insert at high risk of converting to a large insert (S*); and a change from the high risk small insert to a large insert (L) which is associated with clinical abnormality. Cytogenetic screening of the mentally handicapped for the fra(X) was equivalent to testing for individuals with a large insert (L) as there was no evidence that a small insert (S) has a deleterious effect on the phenotype. The consensus was that in diagnostic laboratories cytogenetics is still the method of choice, with subsequent molecular investigation of those patients found or suspected of being fra(X) positive; no consensus was reached on the relative merits of cytogenetics and molecular techniques for screening. Mulley et al. (1992) reported a high success rate with the direct molecular diagnosis of fragile X using the pfxa3 probe which detects amplification of an unstable DNA element consisting of variable length CCG repeats.

Snow et al. (1993) found that PCR followed by DNA sequencing of the FMR1 gene allowed the most accurate determination of CGG repeat numbers up to approximately 130 repeats. Turner et al. (1996) suggested that the clinical definition of fragile X syndrome be redefined in males as a mental handicap associated with absolute or relative deficiency of the FMR1 protein. In the absence of a readily available protein test, analysis of the trinucleotide repeat size has been used for diagnosis. An increase in the size of the trinucleotide repeat over a particular value initiates methylation of the FMR gene promoter site and suppression of FMR1 gene transcription. Testing can identify individuals who lack FMR1 protein as a consequence of deletion of the gene but will not identify those individuals whose FMR1 protein is defective through mutation.

Willemsen et al. (1995, 1997) developed a diagnostic method using mouse monoclonal antibodies against the FMR1 protein that allowed for detection of the fragile X syndrome in a blood smear. This noninvasive test required only 1 or 2 drops of blood and could be used to screen large groups of mentally retarded persons and neonates. Willemsen et al. (1999) modified the antibody test for application to hair roots. Mentally retarded female patients with a full mutation showed FMR protein expression in only some of their hair roots (less than 55%), and no overlap with normal female controls was observed.

Storm et al. (1998) noted that incomplete EcoRI digestion may lead to false diagnosis of fragile X syndrome and suggested that HindIII digest be used instead of EcoRI to identify premutation vs normal fragment length in genomic DNA.

Abrams et al. (1999) examined olfactory neuroblasts from 2 mentally retarded, autistic brothers with fragile X expansion mutations in leukocytes. Olfactory neurons were chosen for study because they are accessible neurons that undergo regeneration and are closely linked to the brain. In both subjects, the genotype in neuroblasts was highly, but not perfectly, consistent with that observed in leukocytes. The results suggested that FMR1 mutation patterns in leukocytes are a good, albeit potentially fallible, reflection of such patterns in the brain and demonstrated the feasibility of using olfactory neuron samples to evaluate FMR1 mutations in humans in vivo.

Stoll (2001) presented 11 children under the age of 8 years and noted the difficulties in diagnosis of fragile X syndrome at this age. The author emphasized the importance of fragile X DNA testing in all children with mental retardation, autism, or significant developmental delay without a clear etiology.

MacKenzie et al. (2006) reported a 46-year-old male patient with a typical fragile X syndrome phenotype who was found to be a somatic mosaic for the FMR1 repeat expansion. Analysis of peripheral blood detected a premutation allele of 58 CGG repeats, whereas skin fibroblasts yielded a full mutation allele of 500 CGG repeats. The authors suggested that the proband may have inherited a full mutation that has undergone selective contraction, given his age at molecular diagnosis. MacKenzie et al. (2006) concluded that testing of ectodermally derived tissues may provide improved diagnosis for fragile X syndrome.

Coffee et al. (2009) reported the development of an assay for newborn screening of fragile X syndrome. The assay showed 100% specificity and 100% sensitivity for detecting FMR1 methylation on dried blood spots, thus successfully distinguishing normal males from those with the full mutation. The assay could also detect excess FMR1 methylation in 82% of females with full mutations, although the methylation status did not correlate with intellectual disability. With amelogenin PCR used for detecting the presence of a Y chromosome, this assay also detected males with Klinefelter syndrome (47,XXY). Among 64 males with FMR1 methylation, 7 were found to have full-mutation fragile X syndrome and 57 had Klinefelter syndrome. In their study of 36,124 newborn males, Coffee et al. (2009) estimated the incidence of fragile X syndrome to be 1 in 5,161 newborn males, and that of Klinefelter syndrome to be 1 in 633.

Carrier Females

Toledano-Alhadef et al. (2001) tested 14,334 Israeli women of childbearing age for fragile X carrier status between 1992 and 2000. These women were either preconceptional or pregnant and had no family history of mental retardation. They identified 207 carriers of an allele with more than 50 repeats, representing a prevalence of 1:69. There were 127 carriers with more than 54 repeats, representing a prevalence of 1:113. Three asymptomatic women carried the full-mutation allele. Among the premutation and full-mutation carriers, 177 prenatal diagnoses were performed. Expansion occurred in 30 fetuses, 5 of which had an expansion to the full mutation. The authors recommended wide-scale screening to identify female carriers.

In 34 female full mutation carriers and unaffected female control relatives, Willemsen et al. (2003) found a correlation between cognitive function and the percentage of hair roots that expressed the FMRP protein. Cognitive function in the female carriers was much more strongly determined by the absence of FMRP than by genetic background.

Angkustsiri et al. (2008) described a 23-year-old woman with the full fragile X mutation who had no dysmorphic features and above-average intelligence combined with significant impairments due to anxiety and learning disability. Her brother had fragile X syndrome, her mother was a premutation carrier, and her maternal grandfather was the first patient diagnosed with the fragile X tremor/ataxia syndrome (FXTAS; 300623 and Hagerman et al., 2001). Angkustsiri et al. (2008) concluded that women with fragile X syndrome can present primarily with learning and emotional problems and that clinicians should consider the diagnosis in these women regardless of their IQ, particularly if there are physical features or a family history consistent with fragile X syndrome.

Prenatal Diagnosis

Jenkins et al. (1982) detected the fragile X marker in cultured amniocytes, enabling successful prenatal diagnosis. Jenkins et al. (1984) described prenatal diagnosis of 3 cases of fragile X syndrome based on cytogenetic analysis of cultured amniocytes. The testes of 2 positive fetuses appeared large for gestational age.

Sutherland et al. (1991) reported prenatal diagnosis of fragile X syndrome in a male fetus using direct analysis of an unstable sequence in DNA obtained by chorionic villus sampling. They used a probe referred to as pfxa3 to detect an abnormal 2.3-kb band in the fetus. Normal carrier males usually have a fragile X band that is between 1.1 and 1.6 kb. Yamauchi et al. (1993) used the diagnostic DNA probe pPCRfx1 to confirm that an at-risk fetus was a heterozygous female carrier.

Dreesen et al. (1995) approached preimplantation testing for the fragile X syndrome by genotyping the polymorphic DXS548 AC-repeat locus, which is closely linked to the FMR1 gene, in unfertilized oocytes and extruded polar bodies. They concluded that a PCR procedure could be performed within 16 hours after blastomere biopsy and that for carrier females heterozygous at the DXS548 locus, preimplantation testing with DXS548 is a possible alternative to prenatal testing.

Molecular Genetics

Nomenclature of Expanded Trinucleotide Repeats

The repeat involved in the fragile X syndrome is variously referred to here as (CGG)n or (CCG)n. The identical repeat found in the cloned FRAXE gene (309548) was referred to as (GCC)n by Knight et al. (1993). There are only 10 different trinucleotide repeats, but each can be written in a number of ways. Sutherland (1993) favored the convention that lists the motif in alphabetical order in the 5-prime to 3-prime direction. Consistent with this, he uses the (CCG)n designation. He preferred, furthermore, the designation (AGC)n for the other clinically significant dinucleotide repeat found in myotonic dystrophy (DM1; 160900), Huntington disease (143100), Kennedy disease (SMAX1; 313200), and SCA1 (164400); (CAG)n is the designation most often used. Sutherland (1993) suggested that the same convention can apply to dinucleotides. He wrote: 'It must be very confusing for newcomers to the literature to find (AC)n, (CA)n, (GT)n, and (TG)n repeats, when the cognoscenti know these are synonyms.'

Fragile X Syndrome

Kremer et al. (1991) demonstrated that the presence of an unstable expanded trinucleotide repeat sequence, designated p(CGG)n (309550.0004), in the FMR1 gene is the basis of fragile X syndrome. The authors showed that normal X chromosomes have about 40 +/- 25 copies of p(CCG)n and that within these limits the sequence is a stable DNA polymorphism. The fragile X genotype was characterized by an increased amount of unstable DNA that maps to the repeat.

Pieretti et al. (1991) found absence of FMR1 mRNA in lymphoblastoid cell lines and leukocytes derived from patients with fragile X syndrome, whereas it was normally expressed in normal controls and carriers.

Devys et al. (1992) noted that there are 2 main types of mutations involved in fragile X syndrome. Premutations, which do not cause mental retardation, 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 a 600 bp or more amplification, often with extensive somatic heterogeneity and abnormal DNA methylation.

De Boulle et al. (1993) identified a missense mutation in the FMR1 gene (I304N; 309550.0001) in a patient with a severe form of fragile X mental retardation, confirming that abnormality of the FMR1 gene underlies fragile X syndrome.

Russo et al. (1998) described a female with borderline cognitive impairment who was compound heterozygous for a full FMR1 mutation and a premutation. The parents came from the same small village in Italy. The proband's mother and aunt reported that they had undergone premature ovarian failure at 35 years of age (see POF1, 311360). Mila et al. (1996) reported a compound heterozygous Spanish female. Linden et al. (1999) reported a 15-year-old girl with fragile X syndrome who was compound heterozygous for a full expansion (363 repeats) and a premutation (103 repeats) in the FMR1 gene. Both parents carried premutations (98 repeats in the father, 146 repeats in the mother). Cognitively, this woman was functioning in the mid range of involvement for fragile X females. She attended regular classes and received supplemental assistance for her learning disabilities. She experienced behavioral characteristics typical of females with fragile X syndrome including severe shyness, anxiety, panic episodes, mood swings, and attention deficits. She responded well to appropriate treatment including fluoxetine for anxiety, methylphenidate for attention problems, and educational therapy.

Gronskov et al. (2011) identified a truncating mutation in the FMR1 gene (S27X; 309550.0005) in a man with classic features of fragile X syndrome. He had mental retardation, early-onset seizures, poor language development, and autistic tendencies. Dysmorphic features included an elongated face, high and broad forehead, low-set large ears, prognathia, and enlarged testes. Neurologic examination showed hypotonia and hypermobility, with hyperextensible joints. Western blot analysis of patient lymphoblastoid cells showed no FMRP protein expression. His mother, who also carried the mutation, had mild to moderate intellectual disability, hypermotor behavior, and automatisms. Gronskov et al. (2011) noted that the frequency of point mutations in the FMR1 gene is unknown, since most screening techniques look for the expanded repeat.

Reviews

D'Hulst and Kooy (2009) provided a review of fragile X syndrome, with a focus on molecular genetics.

Population Genetics

Jacobs (1982) indicated that a reasonable estimate of frequency of fragile X syndrome is 0.5 per 1,000 males. Although many of the cases first ascertained were of northern European descent, affected males have since been found in most ethnic groups.

In Sweden, Blomquist et al. (1982) found that 6 of 96 Swedish boys with IQ less than 50 born between 1959 and 1970 had fraXq28. Blomquist et al. (1985) found the fragile X in 13 (16%) of 83 boys but none of 129 girls with infantile autism.

Webb et al. (1986) performed a population study of school children in the city of Coventry, England, and, using cytogenetic studies, gave an overall prevalence for fragile X syndrome in males and females of 1:952. Morton et al. (1997) reevaluated the 29 children diagnosed with fragile X syndrome by Webb et al. (1986) and confirmed the presence of the FMR1 gene expansion in only 7 of the children, giving a revised prevalence of 1:2,720 to 1:5,714, depending on whether the 4 children lost to follow-up are included. On the basis of molecular genetic analysis, Turner et al. (1996) reported that a prevalence of 1:4,000 or 2.4:10,000 was more realistic than the 1:1,000 reported by Webb et al. (1986).

Filippi et al. (1991) reported findings in a very large Sardinian kindred spanning 6 generations and including 13 patients with Martin-Bell syndrome, several instances of normal transmitting males or females, and the G6PD Mediterranean (305900.0006) mutant segregating in some of its branches. All the fragile X patients and the 15 obligate heterozygous women could be traced through their X-chromosome lineage to a woman in the first generation who must have been heterozygous for a silent premutation at the fragile X locus. Filippi et al. (1991) concluded that this premutation had been converted into a full mutation at least 9 times during the gametogenesis of this ancestor's X-related descendants, of whom 4 were males.

Morton and Macpherson (1992) proposed a model in which the fragile X mutation is postulated to occur as a multistep process. This attractive model provides a framework in which the seemingly contradictory observations of a mutation old enough to establish a founder effect and an apparently high new mutation rate are united. Morton and Macpherson (1992) suggested that 4 types of alleles occur in the fragile X syndrome (see table in the review by Chakravarti, 1992). The 4 types of alleles were as follows: N = normal, with a frequency of 0.9751; S = stable insert with a frequency of 0.0225 and a mean age of about 90 generations; Z = unstable insert with a frequency of 0.0014 and a mean age of 2 generations; and L = mutation with a frequency of 0.0010 and a mean age of 1.4 generations. Thus myotonic dystrophy (DM1; 160900) and fragile X appear to share both the phenomenon of anticipation and the phenomenon of founder effect.

Richards and Sutherland (1992) referred to the amplification mutation involving (CCG)n in the fragile X syndrome and the trinucleotide repeats in myotonic dystrophy and Kennedy disease as 'dynamic mutations.' In studies using 2 polymorphic CA repeats located close to the 'mutation target' for the fragile X syndrome, Oudet et al. (1993) observed significant differences in allelic and haplotypic distributions between normal and fragile X chromosomes, indicating that a limited number of primary events may have been at the origin of most present-day fragile X chromosomes in Caucasian populations. They proposed a putative scheme with 6 founder chromosomes from which most of the observed fragile X-linked haplotypes can be derived directly or by a single event at one of the marker loci. Such founder chromosomes may have carried a number of CGG repeats in an upper-normal range, from which recurrent multistep expansion mutations have arisen. The diversity of haplotypes at the fragile X locus may reflect genetic heterogeneity but may also be explained by mutations in the markers themselves.

Richards et al. (1992) presented haplotype evidence for a founder effect in the fragile X mutation. They found clear evidence of linkage disequilibrium between fragile X and 2 polymorphic microsatellite markers that flank FMR1 and are within 10 kb of the (CCG)n repeat. These markers have 5 to 7 alleles, show no recombination with each other, and define 15 haplotypes. In an analysis of 134 fragile X chromosomes from unrelated affected individuals in Australia and the United States, they found that 58% of the fragile X mutations occurred on the 3 backgrounds that account for 18% of normal chromosomes. Correspondingly, the single most common normal haplotype, which has a frequency of 50%, carries only 18% of fragile X mutations. The data argued for the expected occurrence of multiple, independent mutations, but also indicated the unexpectedly long history of some of these fragile X mutations. Using the FRAXAC1 polymorphic marker in the study of a large number of patients, Hirst et al. (1993) found its allele distribution to be strikingly different on fragile X chromosomes, confirming earlier observations and giving further support to the suggestion of a fragile X founder effect (Richards et al., 1992).

Haataja et al. (1994) presented evidence for a founder effect of fragile X syndrome in Finland arising from a common ancestor in the 16th century.

In a study of 122 Israeli families affected with the fragile X syndrome diagnosed in 7 genetic centers, Dar et al. (1995) found that Tunisian Jews, who comprise only 4% of the general population, accounted for 21% of the fragile X families, suggesting founder effect.

Rousseau et al. (1995) reported a population frequency of 1 in 259 for female carriers of an allele of more than 54 repeats.

The CGG repeat, which is normally polymorphic in length, is frequently interrupted by AGG triplets, which are believed to stabilize the repeat. The absence of AGG triplets, leading to long tracts of perfect CGG repeats, may give rise to predisposed alleles. Kunst et al. (1996) determined the repeat length of 345 chromosomes from 9 populations from various parts of the world and used automated DNA sequencing to assess 14 of them. They found that the FMR1 alleles were very heterogeneous, although the level of variation correlated with the age and/or genetic history of a particular population. Native American alleles, interrupted by 3 AGG repeats, exhibited marked stability over 7,000 years. However, in older African populations, parsimony analysis predicted the occasional loss of an AGG, leading to more perfect CGG repeats.

Studies of (CGG)n repeat structures of selected human populations showed a high degree of conservation of the canonical (CGG)9AGG interruption pattern in different populations and confirmed the proposed stabilizing effect of AGG interruptions (Eichler and Nelson, 1996). In the native population of Greenland, Larsen et al. (1999) found a narrow distribution of (CGG)n allele sizes, similar to that reported for Asian populations. DNA sequencing of alleles with 36 CGG repeats revealed an AGG(CGG)6 insertion previously reported exclusively in Asian populations and a high frequency of 2 other sequence patterns. The data confirmed the Asian origin of the Greenlandic (Eskimo) population and indicated that some (CGG)n alleles have remained stable for 15,000 to 30,000 years, since the population of the New World arrived from Asia via the Bering Strait. The findings added new evidence for the 'out of Asia' theory of the colonization of the New World (Cavalli-Sforza et al., 1994). Studies in Native Americans (Amerinds) had not shown the (CGG)6AGG insertion. This may be due to the relatively small sample sizes in these studies, but may also be caused either by a later migration of the Eskimo population compared with the Amerind and the Na-dene populations (as proposed in the '3 migrations theory' Greenberg et al., 1986 or by genetic bottlenecks during the population of the New World (Wallace and Torroni, 1992)).

Goldman et al. (1997) reported that the prevalence of FRAXA syndrome among institutionalized South African blacks was similar to that reported in the literature for institutionalized white populations. Crawford et al. (1999) found that the prevalence of the FRAXA full mutation in African American males was approximately the same as that in Caucasian American males.

Beresford et al. (2000) reported molecular analysis of 177 males with mental handicap and 1,226 random alleles from Guthrie newborn screening samples in Nova Scotia. No FMR1 premutations or mutations were found. Beresford et al. (2000) also noted that only 1 case of fragile X had been reported in this region since 1980, in an individual who had moved from elsewhere in Canada. Beresford et al. (2000) concluded that the fragile X syndrome was rare in Nova Scotia, a phenomenon they found remarkable given the high prevalence of other rare heritable disorders in the region and that the population has tens of thousands of founders from multiple founding groups.

Larsen et al. (2001) analyzed the AGG interspersion pattern of the (CGG)n repeat and the haplotype distribution of 2 closely located microsatellite markers in 3 circumarctic populations: Norwegians, Saami, and Nenets. The data indicated the existence of chromosomes of Asian origin in the Saami and Nenets populations. Haplotype analysis of Norwegian fragile X males compared to other populations showed that the fragile X founder haplotypes may vary between populations and that the CGG expansion associated with fragile X syndrome may originate from subpopulations of unstable alleles within the normal population.

Several population-based studies in Caucasians of mostly northern European descent established that the prevalence of the fragile X syndrome is probably between 1 in 6,000 and 1 in 4,000 males. Crawford et al. (2002) presented the final results of a 4-year study in the metropolitan area of Atlanta, Georgia, establishing the prevalence of the fragile X syndrome and the frequency of CGG repeat variants in a large Caucasian and African American population. They found that one-quarter to one-third of the children identified with the fragile X syndrome attending Atlanta public schools were not diagnosed before the age of 10 years. Also, a revised prevalence for the syndrome revealed a higher point estimate for African American males (1/2,545; 95% CI 1/5,208-1/1,289) than reported previously, although confidence intervals included the prevalence estimated for Caucasians from this and other studies (1/3,717; 95% CI 1/7,692-1/1,869).

Mingroni-Netto et al. (2002) studied the distribution of CGG repeats and DXS548/FRAXAC1 haplotypes in normal South American populations of different ethnic backgrounds. They found that some rare alleles that seem nearly absent in Europe occurred in higher frequencies among African Brazilians, which suggested a general trend for higher genetic diversity among Africans. Thus, the rarer alleles could be African in origin and would have been lost or possibly not present in the groups that gave rise to Europeans.

Dombrowski et al. (2002) screened 10,572 independent French Canadian males for premutation-size FMR1 alleles and identified 13 who carried alleles of more than 54 repeats, which corresponded to a population frequency of 1 in 813. Haplotype analysis of the 13 identified male carriers revealed that the prevalence of the major fragile X mutation-associated haplotype was increased among FMR1 alleles of 40 to 54 repeats. Although sequencing of highly unstable premutation alleles from fragile X families revealed only pure CGG tracts, 48 of 49 males from the general population with 40 or more triplets had 1 to 2 AGG interruptions. This suggests that the loss of an AGG interruption in the triplet repeat array may not be necessary for expansion of normal alleles of 29 to 30 triplets to intermediate size. The authors concluded that loss of AGG interruptions appears to be a late event that may lead to greatly increased instability and may be related to the haplotype background of specific FMR1 alleles.

Biancalana et al. (2004) reported the molecular diagnosis of fragile X syndrome in France during the 5-year period from 1997 to 2001: 477 families were diagnosed with fragile X syndrome, representing 2.8% of tested male probands and 1% of tested female probands.

Fatima et al. (2014) found the FMR1 full mutation in 15 (6.5%) of 229 Pakistani boys and in 1 (0.9%) of 104 girls with intellectual disability. The overall frequency among intellectually disabled children in this population was 4.8%.

History

The X-linked mental retardation reported by Martin and Bell (1943) is the same as the fragile X syndrome. Opitz et al. (1984) referred to this disorder as the 'Martin-Bell syndrome' on the assumption that the family reported from the Queen Square Hospital in London by J. Purdon Martin and Julia Bell (Martin and Bell, 1943) had that disorder. Although macroorchidism was not mentioned by Martin and Bell (1943), one of the patients was described as having a 'big face and jaw;' furthermore, at least 9 of the affected males were maternal grandsons of 2 unaffected brothers. All but 1 of the mothers of affected males were daughters of these 2 brothers, the other being their sister. Martin and Bell (1943) hypothesized that some controlling factor caused suppression of the disease in the 2 grandfathers without affecting their liability to transmit it. For a superb biography of Julia Bell (1879-1979), see Bundey (1996).

Richards et al. (1981) followed up on the Martin-Bell kindred, demonstrating that it was indeed the fragile X syndrome. The original index patient was then aged 56. All 4 affected males who had adequate karyotyping showed the fragile X syndrome in 17 to 50% of their cells. The other major contribution of Julia Bell was in the defining of many hereditary disorders, such as the forms of brachydactyly, on the basis of massive collections of pedigrees in the famous Treasury of Human Inheritance. She also collaborated with J. B. S. Haldane in the first estimation of linkage in the human, that of colorblindness and hemophilia (Bell and Haldane, 1937).

According to Opitz and Sutherland (1984), Escalante, a graduate student with Frota-Pessoa in Sao Paulo, Brazil, and Drs. Bryan and Gillian Turner in Sydney, Australia, independently noted the association of macroorchidism with X-linked mental retardation in the late 1960s. Escalante et al. (1971) published their findings (which they had reported in 1969 at the Warsaw Congress of the International Association for the Scientific Study of Mental Deficiency) in the Journal de Genetique Humaine. While visiting the Drs. Turner in Sydney, McKusick (1970) examined several patients with mental retardation and macroorchidism.

Antibiotics such as Bactrim (Roche) and Septra (Burroughs Wellcome) contain trimethoprim, which can lower folate levels by inhibition of dihydrofolate reductase. Hecht and Glover (1983) urged avoidance of trimethoprim and other folate antagonists in pregnant women who are at risk for having a child with the fragile X syndrome. Lejeune et al. (1982) described severe clinical regression of psychomotor development in a 2-year-old boy with the fragile X syndrome while on trimethoprim.

Froster-Iskenius et al. (1984) raised the possibility of an autosomal suppressor system to account for the transmission of the marker X syndrome by unaffected males. Steinbach (1986) and Israel (1987) also postulated an autosomal suppressor gene or modifier to explain the occurrence of mentally normal males who transmit the fragile X gene to their daughters and the fact that while about one-third of all female carriers have mental impairment, mothers and daughters of these mentally normal transmitting males are rarely, if ever, mentally impaired. Steinbach (1986) designated the modifier genes as D and N, the equivalent of Israel's s and S, respectively. Essentially the models were identical: normal male transmitters of fragile X were suggested to be homozygous for S (or N), while normal female transmitters are either homozygous for S or heterozygous, Ss (ND). Sherman (1987) found the Israel model attractive.

Thode et al. (1988) were unable to corroborate the existence of a form of the Martin-Bell syndrome with no detectable fragile X. They identified 32 men with the phenotype who were fragile X negative but concluded that they did not fit the full criteria.

Klauck et al. (1997) concluded from molecular genetic studies of 141 patients from 105 simplex and 18 multiplex families that an association of autism with fragile X is nonexistent and that the Xq27.3 region is not a candidate for autism.

Animal Model

For a discussion of animal models of fragile X syndrome, see 309550.