Friedreich Ataxia 1

A number sign (#) is used with this entry because one form of Friedreich ataxia (FRDA1) is caused by mutation in the gene encoding frataxin (FXN; 606829), which has been mapped to chromosome 9q. The most common molecular abnormality is a GAA trinucleotide repeat expansion in intron 1 of the FXN gene: normal individuals have 5 to 30 GAA repeat expansions, whereas affected individuals have from 70 to more than 1,000 GAA triplets (Al-Mahdawi et al., 2006).

Description

Friedreich ataxia is an autosomal recessive neurodegenerative disorder characterized by progressive gait and limb ataxia with associated limb muscle weakness, absent lower limb reflexes, extensor plantar responses, dysarthria, and decreased vibratory sense and proprioception. Onset is usually in the first or second decade, before the end of puberty. It is one of the most common forms of autosomal recessive ataxia, occurring in about 1 in 50,000 individuals. Other variable features include visual defects, scoliosis, pes cavus, and cardiomyopathy (review by Delatycki et al., 2000).

Pandolfo (2008) provided an overview of Friedreich ataxia, including pathogenesis, mutation mechanisms, and genotype/phenotype correlation.

Genetic Heterogeneity of Friedreich Ataxia

Another locus for Friedreich ataxia has been mapped to chromosome 9p (FRDA2; 601992).

Clinical Features

In FRDA, the spinocerebellar tracts, dorsal columns, pyramidal tracts and, to a lesser extent, the cerebellum and medulla are involved. The disorder is usually manifest before adolescence and is generally characterized by incoordination of limb movements, dysarthria, nystagmus, diminished or absent tendon reflexes, Babinski sign, impairment of position and vibratory senses, scoliosis, pes cavus, and hammertoe. The triad of hypoactive knee and ankle jerks, signs of progressive cerebellar dysfunction, and preadolescent onset is commonly regarded as sufficient for diagnosis. McLeod (1971) found abnormalities in motor and sensory nerve conduction.

Cardiac Manifestations

Cardiac manifestations are conspicuous in some cases (Boyer et al., 1962). Hewer (1968) found that one-half of 82 fatal cases of Friedreich ataxia died of heart failure and nearly three-quarters had evidence of cardiac dysfunction in life. Twenty-three percent had diabetes and 4 developed diabetic ketosis terminally. One case had an affected parent. Age at death varied from the first (3 cases) to the eighth (1 case) decade with a mean of 36.6 years. Muscular subaortic stenosis has been described in cases of Friedreich ataxia (Elias, 1972; Boehm et al., 1970). Ackroyd et al. (1984) reviewed cardiac findings in 12 children, aged 6 to 16 years, with FA. In 10, EKG abnormalities were found. All had abnormalities of the echocardiogram in the form of symmetric, concentric, hypertrophic cardiomyopathy.

Casazza and Morpurgo (1996) reviewed a 15-year experience with a large series of patients with Friedreich ataxia to determine the prevalence of hypokinetic cardiomyopathy and to define which patients among those with and without initial left ventricular hypertrophy are most likely to progress to the hypokinetic-dilated form. They concluded that a transition from the hypertrophic to the hypokinetic-dilated form is not rare. The presence of Q waves identified a subgroup of patients with wall motion abnormalities prone to develop a hypokinetic-dilated left ventricle; these patients have a poor prognosis.

Visual System Manifestations

Fortuna et al. (2009) studied in detail possible involvement of visual system pathways in 26 Italian patients with FRDA between 15 and 45 years of age. Twenty-one patients were completely asymptomatic, but visual field examination showed 1 of 3 different patterns of visual field defect: severe visual field impairment with general and concentric reduction of sensitivity, mild reduction of sensitivity and a concentric superior or inferior arcuate defect, and very little depression and only an isolated small paracentral area of reduced sensitivity. Optical coherence tomography showed reduced retinal nerve fiber layer (RNFL) thickness in all patients and reduced number of axons, and approximately half of patients had abnormal visual evoked potentials. Two of the 26 patients presented with sudden bilateral loss of central vision at 25 and 29 years of age, respectively, similar to Leber hereditary optic neuropathy (LHON; 535000). Two additional patients had had severe ophthalmologic features, close to those with the LHON-like visual loss, manifest by consistent reduction of RNFL thickness, bilateral central visual field involvement, and reduced visual acuity. Overall, the findings indicated that visual field involvement can occur in FRDA, resulting from a slowly progressive degenerative process involving the optic nerve and optic radiations. Fortuna et al. (2009) concluded that loss of central vision associated with poor visual acuity may occur late in the course of FRDA, with a predilection for patients who are compound heterozygotes.

Late-onset Form

De Michele et al. (1994) found age of onset greater than age 20 in 19 of 114 of their patients with the classic form of FA as defined by autosomal recessive or sporadic occurrence, progressive unremitting ataxia of limbs and gait, and absence of knee and ankle jerks. Each of the described patients had at least one of the following signs: dysarthria, extensor plantar response, and echocardiographic evidence of hypertrophic cardiomyopathy. Linkage analysis was performed for 16 patients and 25 healthy members from 8 of the 17 affected families studied. No recombination was found (maximum lod score of 5.17 at theta = 0.0) with the extended MLS1-MS-GS4 haplotype. This suggests that late-onset FA is likely to be an allelic disorder with classic FA for which the upper limit for age of onset was given as 20 years by Geoffroy et al. (1976) and as 25 years by Harding (1981). Eleven of the patients reported by De Michele et al. (1994) had onset after age 25; 2 of these patients had onset after age 30. The only significant differences between these late-onset patients and the more typical early-onset patients were a lower occurrence of skeletal deformities in the late-onset groups and normal visually evoked potentials which were abnormal in 69% of individuals presenting with FA before age 20. The disease progression was slower in the late-onset group.

The Ataxia Study Group (Pujana et al., 1999) in Spain found no spinocerebellar ataxia (see 164400) or DRPLA (125370)-type mutations (unstable CAG repeat expansions) in 60 late-onset sporadic cases of spinocerebellar ataxia. One of the 60 cases carried a homozygous GAA repeat expansion in the FRDA gene. In this case, the disease began with vertigo episodes at 30 years of age, whereas the onset for gait ataxia was 35 years, with progression of other signs such as dysarthria, areflexia, pes cavus, and reduced motor and sensory conduction velocity. Magnetic resonance imaging (MRI) showed moderate cerebellar cortical atrophy.

Lhatoo et al. (2001) reported a case of 'very late onset' Friedreich ataxia, confirmed by genetic testing, in a man who presented with a history of lower limb spasticity beginning at age 40. The features were unusual in that he did not have ataxia (although he did have a spastic gait), nystagmus, areflexia, or sensory neuropathy, and brain scans were normal.

Friedreich Ataxia with Retained Reflexes (FARR)

Harding (1981) described absence of lower limb tendon reflexes as an absolute criterion for the diagnosis of Friedreich ataxia, setting aside early-onset cerebellar ataxia with retained tendon reflexes (EOCA; 212895) as a separate category. Palau et al. (1995) presented 6 sibships in which 2 affected probands fulfilled all of Harding's criteria for the diagnosis of Friedreich ataxia, except for the preservation of deep tendon reflexes in the lower extremities. In 3 of these sibships, affected sibs were discordant for the presence or absence of deep tendon reflexes. They considered the presence of cardiomyopathy by ECG or echocardiogram as an essential criterion for this diagnostic category, which they described as Friedreich ataxia with retained reflexes (FARR). A maximum lod score of 3.38 at a recombination fraction theta equal to 0.00 was obtained, suggesting that FARR is an allelic variant of Friedreich ataxia.

Coppola et al. (1999) found that among 101 patients homozygous for GAA expansion within the FRDA gene, 11 from 8 families had FARR. These patients had a lower occurrence of decreased vibration sense, pes cavus, and echocardiographic signs of left ventricular hypertrophy than did the 90 Friedreich ataxia patients with areflexia. Furthermore, the mean age at onset was significantly later (26.6 years vs 14.2 years) and the mean size of a smaller allele was significantly less (408 vs 719 GAA triplets) in FARR patients. The neurophysiologic findings were consistent with milder peripheral neuropathy and milder impairment of the somatosensory pathways in FARR patients.

Marzouki et al. (2001) described 3 Tunisian families with early-onset cerebellar ataxia with retained tendon reflexes in which Friedreich ataxia, vitamin E deficiency ataxia (AVED; 277460), and known forms of autosomal dominant cerebellar ataxia were excluded by linkage analysis.

Chorea

Hanna et al. (1998) described 2 patients with a generalized chorea in the absence of cerebellar signs who were homozygous for the trinucleotide repeat expansion in intron 1 of the FXN gene that is typical of Friedreich ataxia. Chorea as a rare manifestation of Friedreich ataxia had previously been controversial. This was the first report of chorea in patients confirmed to have the FA genetic abnormality. One patient was a 21-year-old student in whom the diagnosis of idiopathic structural thoracic scoliosis was made at the age of 10 years. The scoliosis was treated surgically at age 14 years by insertion of Harrington rods. Neurologic symptoms developed at age 19 years. He noticed that his gait had become abnormal. He described involuntary jerks of his legs interfering with normal gait and causing occasional falls. Similar involuntary movements of his upper arms had stopped him from playing the guitar. His father described him as generally 'twitchy.' Neurologic examination revealed facial and generalized chorea but no cerebellar signs. Eye movements, speech, and optic discs were normal. He was areflexic. Genetic analysis showed repeat sizes of 500 and 800 repeats in the 2 alleles of the FXN gene. The second case was that of a 13-year-old boy who at the age of 10 years developed recurrent palpitations and was found to have ventricular arrhythmias secondary to a mild hypertrophic cardiomyopathy. His parents described him as generally twitchy and clumsy over the past year, but there was no history of gait disturbance. Neurologic examination revealed mild generalized chorea involving particularly his head, neck, and shoulders. Eye movements, speech, and optic discs were normal. Although he was generally mildly clumsy, there were no unequivocal cerebellar signs. Genetic analysis confirmed that he was homozygous for the FA intron 1 expansion with both alleles measuring 4.5 kb corresponding to a repeat size of approximately 1,000 repeats.

Diagnosis

In 20 childhood cases (mean age of onset of symptoms, 6.1 years), Ulku et al. (1988) demonstrated the usefulness of abnormal sensory nerve conduction velocities in confirming the diagnosis. Motor nerve conduction velocities are usually normal or show a mild reduction.

To investigate the genetic background of apparently idiopathic sporadic cerebellar ataxia, Schols et al. (2000) tested for CAG/CTG trinucleotide repeats causing spinocerebellar ataxia types 1, 2 (SCA2; 183090), 3 (SCA3; 109150), 6 (SCA6; 183086), 7 (SCA7; 164500), 8 (SCA8; 608768), and 12 (SCA12; 604326), and the GAA repeat of the frataxin gene in 124 patients, including 20 patients with the clinical diagnosis of multiple system atrophy. Patients with a positive family history, atypical Friedreich phenotype, or symptomatic (secondary) ataxia were excluded. Genetic analyses uncovered the most common Friedreich mutation in 10 patients with an age of onset between 13 and 36 years. The SCA6 mutation was present in 9 patients with disease onset between 47 and 68 years of age. The CTG repeat associated with SCA8 was expanded in 3 patients. One patient had SCA2 attributable to a de novo mutation from a paternally transmitted, intermediate allele. Schols et al. (2000) did not identify the SCA1, SCA3, SCA7, or SCA12 mutations in this group of idiopathic sporadic ataxia patients. No trinucleotide repeat expansion was detected in the multiple system atrophy subgroup. This study revealed the genetic basis in 19% of apparently idiopathic ataxia patients. SCA6 was the most frequent mutation in late-onset cerebellar ataxia. The authors concluded that the frataxin trinucleotide expansion should be investigated in all sporadic ataxia patients with onset before age 40, even when the phenotype is atypical for Friedreich ataxia.

Prenatal Diagnosis

Using the anonymous DNA marker MCT112 (D9S15), which shows tight linkage to FRDA (lod score = 36.1 at theta = 0.0), Wallis et al. (1989) achieved prenatal diagnosis in a family with 1 affected child; the fetus was affected. Monros et al. (1995) described experience using new flanking markers which they claimed increased the confidence of prenatal diagnosis to almost 100%.

Clinical Management

Peterson et al. (1988) observed improvement when amantadine hydrochloride was orally administered.

Rustin et al. (1999) assessed the effect of idebenone, a free-radical scavenger, in 3 patients with Friedreich ataxia. Their rationale for the study was based on the fact that the frataxin gene is involved in the regulation of mitochondrial iron content. Rustin et al. (1999) used an in vitro system to elucidate the mechanism of iron-induced injury and to test the protective effects of various substances. The in vitro data suggested that both iron chelators and antioxidant drugs that may reduce iron are potentially harmful in patients with Friedreich ataxia. Conversely, preliminary findings in patients suggested that idebenone protects heart muscle from iron-induced injury.

Carroll et al. (1980) referred to a Friedreich's Ataxia Association in England, a voluntary organization of patients and their families and friends.

Lynch et al. (2002) reviewed the genetic basis, diagnostic considerations, therapy, and usefulness of genetic testing for Friedreich ataxia.

Jauslin et al. (2002) developed a cellular assay system that discriminates between fibroblasts from FRDA patients and unaffected donors on the basis of their sensitivity to pharmacologic inhibition of de novo synthesis of glutathione. Supplementation with selenium effectively improved the viability of FRDA fibroblasts, suggesting that basal selenium concentrations may not be sufficient to allow an adequate increase in the activity of certain detoxification enzymes, such as glutathione peroxidase (GPX; see 138320). Idebenone, a mitochondrially localized antioxidant that ameliorates cardiomyopathy in FRDA patients, as well as other lipophilic antioxidants, protected FRDA cells from cell death. Jauslin et al. (2002) suggested that small-molecule GPX mimetics have potential as a treatment for Friedreich ataxia and presumably also for other neurodegenerative diseases with mitochondrial impairment.

Fahey et al. (2007) found that the 25-foot walk test velocity was an accurate measure of ambulation reflecting daily activity as measured with a step activity monitor accelerometer in patients with FRDA.

In a proof-of-concept study, Boesch et al. (2007) found that treatment of 11 FRDA patients with recombinant erythropoietin for 8 weeks resulted in a mean increase of 27% in frataxin levels in lymphocytes of 7 patients compared to baseline levels. All patients also showed a reduction of oxidative stress markers, although there was no significant clinical neurologic improvement.

Inheritance

Friedreich ataxia is inherited as an autosomal recessive disorder (Andermann et al., 1976; Montermini et al., 1997).

Mapping

Unlike one form of dominant ataxia (SCA1; 164400), Friedreich ataxia does not show linkage to HLA and other chromosome 6 markers. Chamberlain et al. (1987) excluded chromosome 19 as the site of the abnormality in this disorder. Keats et al. (1987) studied linkage between FA and 36 polymorphic blood group and protein markers in 3 patient populations: 16 families from the inbred Acadian population from Louisiana, 21 French-Canadian families from Quebec, and 9 apparently unrelated British families. No evidence of linkage heterogeneity was found among the populations. The negative lod scores excluded the FA locus from more than 20% of the genome.

Chamberlain et al. (1988) assigned the gene mutation responsible for Friedreich ataxia to 9p22-cen by genetic linkage to an anonymous DNA marker and to the interferon-beta gene probe (INFB; 147640). The anonymous probe, called MCT112 (D9S15), had been assigned to proximal 9q by multipoint linkage analysis; INFB maps to 9p21. With INFB, the maximum lod score was 2.98 at a male recombination fraction of 0.10. The maximum lod score between FRDA and MCT112, calculated for combined sexes, was 6.41 at a recombination fraction of 0.0 (0 to 5% confidence interval). No recombinants were observed between FRDA and the probe.

In studies of 33 families, Fujita et al. (1989) showed tight linkage with D9S15, which maps to 9q (HGM9); maximum lod = 6.82 at theta = 0.02. Close linkage was also found with D9S5; maximum lod score = 5.77 at theta = 0.00. Fujita et al. (1989) found less close linkage with INFB. Keats et al. (1989) established that the disorder in persons of Acadian ancestry is determined by a gene at the same locus, inasmuch as the Acadian form showed linkage to the same DNA marker, D9S15, that has been found in other studies (maximum lod = 5.06 at theta = 0.0).

Fujita et al. (1989) concluded that the FRDA locus lies on the proximal portion of the long arm of chromosome 9, not on the short arm. They showed a close linkage to 2 DNA markers: maximum lod = 11.36 at FRDA = 0.00 for D9S15; maximum lod = 6.27 at beta = 0.00 for D9S5. D9S5 was mapped to 9q12-q13 by in situ hybridization. They suggested that the cluster is situated distal to the heterochromatic region, i.e., 9q13-q21. The linkage information was extended by Hanauer et al. (1990).

By in situ hybridization, Raimondi et al. (1990) also assigned the D9S5 locus, which has been found to be very tightly linked to Friedreich ataxia, to 9q12-q13. In studies in Italy, Pandolfo et al. (1990) obtained results which, when combined with those reported by others, indicated close linkage of the FRDA locus and markers D9S5 and D9S15. The linkage data were supported by close physical linkage of D9S5 and D9S15 by pulsed field gel electrophoresis.

Wallis et al. (1990) identified a hypervariable microsatellite sequence within the chromosome 9 marker MCT112, which is tightly linked to FRDA. The system of AC repeats detects 7 alleles ranging in size from 195 to 209 basepairs, substantially increasing informativity at the locus. The maximum lod score of FRDA versus MCT112 was 66.91 at a recombination fraction of theta = 0.00. Hypervariable AC repeats of this type were described by Weber and May (1989) and by Litt and Luty (1989).

Using marker D9S15 in in situ hybridization studies, Shaw et al. (1990) localized the FRDA locus to 9q13-q21.1. Wilkes et al. (1991) identified 11 CpG islands in the 1.7-megabase interval most likely to contain the Friedreich ataxia locus, based on its tight linkage to the anonymous DNA markers MCT112 and DR47. Each of these regions is considered a potential candidate sequence for the mutated gene in this disorder, since no precise localization of the disease gene relative to the markers can be obtained from recombinational events. Four of the CpG islands were identified by analysis of 3 YAC clones including the MCT112/DR47 cluster over a 700-kb interval.

In 11 Acadian families from southwest Louisiana, Sirugo et al. (1992) found evidence of strong founder effect: a specific extended haplotype spanning 230 kb between markers D9S5 and D9S15 was present on 70% of independent FA chromosomes and only once (6%) on the normal ones. In a linkage study of 3 large FA families of Tunisian origin, Belal et al. (1992) identified a meiotic recombination in an unaffected individual, which excluded a 150-kb segment, including D9S15, as a possible location for the FRDA locus. Rodius et al. (1994) constructed a YAC contig extending 800 kb centromeric to the closely linked D9S5 and isolated 5 new microsatellite markers from this region. Using homozygosity by descent and association with founder haplotypes in isolated populations, they identified a phase-known recombination and a probable historic recombination on haplotypes from Reunion Island patients, both of which placed 3 of the 5 markers proximal to FRDA. These were the first close FRDA flanking markers to be identified on the centromeric side. The other 2 markers allowed Rodius et al. (1994) to narrow the breakpoint of a previously identified distal recombination. Taken together, the results placed the FRDA locus in a 450-kb interval, small enough for direct search of candidate genes.

Mapping studies showed that the FRDA locus is most tightly linked to D9S5 and D9S15 which lie only 250 kb apart. A recombinant demonstrated by Chamberlain et al. (1993), as well as the analysis of FRDA-linked haplotypes in a population with a founder effect (Sirugo et al., 1992), suggested that the disease gene lies on the D9S5 side of the D9S15-D9S5 interval. The orientation of the 2 markers in relation to the centromere and to each other could not be determined, however. (The maximum lod score was 96.69 at theta = 0.01 for D9S15 and 98.22 at theta = 0.01 for D9S5.) Fujita et al. (1991) described evolutionarily conserved sequences around the D9S5 locus that might correspond to a candidate gene for FRDA.

Heterogeneity

Winter et al. (1981) found about twice as many first-cousin marriages among the parents of affected sibships as was expected; this suggested genetic heterogeneity to them. Genetic heterogeneity was sought by Chamberlain et al. (1989) who typed members of 80 families with the chromosome 9 marker MCT112, previously shown to be closely linked to the disease locus. No evidence of heterogeneity was discovered. The combined total lod score was 25.09 at a recombination fraction of 0.00.

Chromosome 9p Locus

Kostrzewa et al. (1997) provided strong evidence of a second FRDA locus on 9p. Studying 2 families, each with 2 sibs with FRDA, they could not detect a mutation in STM7/X25. Furthermore, haplotype analysis of the STM7/X25 region of chromosome 9 demonstrated that the relevant portion of that chromosome differed in the patients. Although the patients studied had typical FRDA, 1 sib pair had the uncommon symptom of retained tendon reflexes. In order to investigate whether retained tendon reflexes are characteristic of FRDA caused by the second locus, which they termed FRDA2 (601992), they studied an unrelated FRDA patient with retained tendon reflexes. The observation of typical mutations in STM7/X25 (GAA expansions) in this patient demonstrated that the 2 genetically different forms of FRDA cannot be distinguished clinically.

Genetic Heterogeneity

Bouhlal et al. (2008) reported an unusual, highly consanguineous Tunisian family in which 11 individuals had autosomal recessive ataxia caused by 3 distinct gene defects. Seven patients who also had low vitamin E levels were all homozygous for the common 744delA mutation in the TTPA gene (600415.0001), consistent with a diagnosis of AVED (277460). Two patients with normal vitamin E levels were homozygous for a mutation in the FXN gene (606829.0001), consistent with a diagnosis of FRDA. The final 2 patients with normal vitamin E levels carried a mutation in the SACS gene (604490), consistent with a diagnosis of ARSACS (270550). The clinical phenotype was relatively homogeneous, although the 2 patients with SACS mutations had hyperreflexia of the knee. One asymptomatic family member was compound heterozygous for the TTPA and FXN mutations. Bouhlal et al. (2008) emphasized the difficulty of genetic counseling in deeply consanguineous families.

Molecular Genetics

Delatycki et al. (1999) stated that 2% of cases of Friedreich ataxia are due to point mutations in the FXN gene (606829), the other 98% being due to expansion of a GAA trinucleotide repeat in intron 1 of the FXN gene (606829.0001). They indicated that 17 mutations had so far been described. Similarly, Lodi et al. (1999) cited data indicating that the GAA triplet expansion in the first intron of the FXN gene is the cause of Friedreich ataxia in 97% of patients.

Genotype/Phenotype Correlations

Filla et al. (1996) studied the relationship between the trinucleotide (GAA) repeat length and clinical features in Friedreich ataxia. The length of the FA alleles ranged from 201 to 1,186 repeat units. There was no overlap between the size of normal alleles and the size of alleles found in FA. The lengths of both the larger and the smaller alleles varied inversely with the age of onset of the disorder. Filla et al. (1996) reported that the mean allele length was significantly higher in FA patients with diabetes and in those with cardiomyopathy. They noted that there was meiotic instability with a median variation of 150 repeats. Isnard et al. (1997) examined the correlation between the severity of left ventricular hypertrophy in Friedreich ataxia and the number of GAA repeats. Left ventricular wall thickness was measured in 44 patients using M-mode echocardiography and correlated with GAA expansion size on the smaller allele (267 to 1200 repeats). A significant correlation was found (r = 0.51, p less than 0.001), highlighting an important role for frataxin in the regulation of cardiac hypertrophy.

In a study of 187 patients with autosomal recessive ataxia, Durr et al. (1996) found that 140, with ages at onset ranging from 2 to 51 years, were homozygous for a GAA expansion that had 120 to 1,700 repeats of the trinucleotides. About one-quarter of the patients, despite being homozygous, had atypical Friedreich ataxia; they were older at presentation and had intact tendon reflexes. Larger GAA expansions correlated with earlier age at onset and shorter times to loss of ambulation. The size of the GAA expansions (and particularly that of the smaller of each pair of alleles) was associated with the frequency of cardiomyopathy and loss of reflexes in the upper limbs. The GAA repeats were unstable during transmission. Thus, the clinical spectrum of Friedreich ataxia is broader than previously recognized, and the direct molecular test for the GAA expansion is useful for the diagnosis, prognosis, and genetic counseling.

Pianese et al. (1997) presented data suggesting that (1) the FRDA GAA repeat is highly unstable during meiosis, (2) contractions outnumber expansions, (3) both parental source and sequence length are important factors in variability of FRDA expanded alleles, and (4) the tendency to contract or expand does not seem to be associated with particular haplotypes. Thus, they concluded that FRDA gene variability appears to be different from that found with other triplet diseases.

Bidichandani et al. (1997) found an atypical FRDA phenotype associated with a remarkably slow rate of disease progression in a Caucasian family. It was caused by compound heterozygosity for a G130V missense mutation (606829.0005) and the GAA expansion of the FXN gene. The missense mutation G130V was the second mutation to be identified in the FXN gene and the first to be associated with a variant FRDA phenotype. This and the other reported missense mutation (I154F; 606829.0004) mapped within the highly conserved sequence domain in the C-terminus of the frataxin gene. Since the G130V mutation was unlikely to affect the ability of the first 16 exons of the neighboring STM7 gene to encode a functional phosphatidylinositol phosphate kinase, Bidichandani et al. (1997) questioned the role of STM7 in Friedreich ataxia.

McCabe et al. (2002) reported phenotypic variability in 2 affected sibs with compound heterozygosity for the G130V mutation and a GAA expansion. The first sib, a 34-year-old man, first presented at age 10 with leg stiffness and mild gait ataxia and later developed significant limb spasticity. His sister had onset of disease at age 15, with progressive ataxia and lack of limb spasticity.

Since Friedreich ataxia is an autosomal recessive disease, it does not show typical features observed in other dynamic mutation disorders, such as anticipation. Monros et al. (1997) analyzed the GAA repeat in 104 FA patients and 163 carrier relatives previously defined by linkage analysis. The GAA expansion was detected in all patients, most (94%) of them being homozygous for the mutation. They demonstrated that clinical variability in FA is related to the size of the expanded repeat: milder forms of the disease (late-onset FA and FA with retained reflexes) were associated with shorter expansions, especially with the smaller of the 2 expanded alleles. Absence of cardiomyopathy was also associated with shorter alleles. Dynamics of the GAA repeat were investigated in 212 parent-offspring pairs. Meiotic instability showed a sex bias: paternally transmitted alleles tended to decrease in a linear way that depended on the paternal expansion size, whereas maternal alleles either increased or decreased in size. All but 1 of the patients with late-onset FA were homozygous for the GAA expansion; the exceptional individual was heterozygous for the expansion and for another unknown mutation. All but 1 of the FA patients with retained reflexes exhibited an axonal sensory neuropathy. However, preservation of their tendon reflexes suggested that the physiologic pathways of the reflex arch remained functional. A close relationship was found between late-onset disease and absence of heart muscle disease.

Delatycki et al. (1999) studied FRDA1 mutations in FA patients from Eastern Australia. Of the 83 people studied, 78 were homozygous for an expanded GAA repeat, while the other 5 had an expansion in one allele and a point mutation in the other. The authors presented a detailed study of 51 patients homozygous for an expanded GAA repeat. They identified an association between the size of the smaller of the 2 expanded alleles and age at onset, age into wheelchair, scoliosis, impaired vibration sense, and the presence of foot deformity. However, no significant association was identified between the size of the smaller allele and cardiomyopathy, diabetes mellitus, loss of proprioception, or bladder symptoms. The larger allele size was associated with bladder symptoms and the presence of foot deformity.

Pathogenesis

Babcock et al. (1997) characterized the frataxin homolog in Saccharomyces cerevisiae, designated YFH1, which encodes a mitochondrial protein involved in iron homeostasis and respiratory function. They suggested that characterizing the mechanism by which YFH1 regulates iron homeostasis in yeast may help define the pathologic process leading to cell damage in Friedreich ataxia. The knockout of the YFH1 gene in yeast showed a severe defect of mitochondrial respiration and loss of mtDNA associated with elevated intramitochondrial iron (Babcock et al., 1997; Koutnikova et al., 1997; Wilson and Roof, 1997).

Cavadini et al. (2000) showed that wildtype FRDA cDNA can complement the YFH1 protein-deficient yeast (YFH1-delta) by preventing the mitochondrial iron accumulation and oxidative damage associated with loss of YFH1. The G130V mutation (606829.0005) affected protein stability and resulted in low levels of mature frataxin, which were nevertheless sufficient to rescue YFH1-delta yeast. The W173G (606829.0007) mutation affected protein processing and stability and resulted in severe mature frataxin deficiency. Expression of the FRDA W173G cDNA in YFH1-delta yeast led to increased levels of mitochondrial iron which were not as elevated as in YFH1-deficient cells but were above the threshold for oxidative damage of mitochondrial DNA and iron-sulfur centers, causing a typical YFH1-delta phenotype. Cavadini et al. (2000) concluded that frataxin functions like YFH1 protein, providing additional experimental support for the hypothesis that FRDA is a disorder of mitochondrial iron homeostasis.

Rotig et al. (1997) suggested that the frataxin gene plays a role in the regulation of mitochondrial iron content. They found a combined deficiency of a Krebs cycle enzyme, aconitase (100880, 100850), and 3 mitochondrial respiratory chain complexes in endomyocardial biopsy samples from 2 unrelated patients with FRDA. All 4 enzymes share iron-sulfur (Fe-S) cluster-containing subunits of mitochondrial respiratory complexes I, II, and III that are damaged by iron overload through generation of oxygen-free radicals. Disruption of the YFH1 gene resulted in multiple Fe-S-dependent enzyme deficiencies in yeast. Deficiency of Fe-S-dependent enzyme activities in both FRDA patients and yeast should be related to mitochondrial iron accumulation, especially as Fe-S proteins are remarkably sensitive to free radicals. Rotig et al. (1997) suggested that mutated frataxin triggers aconitase and mitochondrial Fe-S respiratory enzyme deficiency in Friedreich ataxia, which should therefore be regarded as a mitochondrial disorder.

Koutnikova et al. (1997) demonstrated that human frataxin colocalizes with the mitochondrial protein cytochrome-c oxidase in HeLa cells and concluded that Friedreich ataxia is a mitochondrial disease caused by mutation in the nuclear genome.

Wilson and Roof (1997) suggested that mitochondrial dysfunction contributes to FRDA pathophysiology. Gray and Johnson (1997) speculated that the progression of Friedreich ataxia and the association of hypertrophic cardiomyopathy, blindness, deafness, and diabetes mellitus are consistent with a mitochondrial disorder.

To test the hypothesis that Friedreich ataxia is a disease of mitochondrial oxidative stress, Wong et al. (1999) studied cultured fibroblasts carrying homozygous GAA repeat expansions. The FRDA fibroblasts were hypersensitive to iron stress and considerably more sensitive to hydrogen peroxide than were control cells. The iron chelator deferoxamine rescued FRDA fibroblasts more than controls from oxidant-induced death, but mean mitochondrial iron content was only 40% greater in FDRA fibroblasts. Treatment with apoptosis inhibitors rescued FDRA but not control cells from oxidant stress, and staurosporine-induced caspase-3 (600636) activity was higher in FDRA fibroblasts, consistent with the possibility that an apoptotic step upstream of caspase-3 is activated in FDRA fibroblasts.

Lodi et al. (1999) reported in vivo evidence of impaired mitochondrial respiration in skeletal muscle of FRDA patients. Using phosphorus magnetic resonance spectroscopy, they demonstrated a maximum rate of muscle mitochondrial ATP production below the normal range in all 12 FRDA patients and a strong negative correlation between that maximum rate and the number of GAA repeats in the smaller allele. These results showed that FRDA is a nuclear-encoded mitochondrial disorder affecting oxidative phosphorylation and provided a rationale for treatments aimed to improve mitochondrial function in this condition. Lodi et al. (1999) pointed out that skeletal muscle deficits are not clinically apparent in patients with Friedreich ataxia. It was not clear why the disease phenotype is so prominent in the nervous system and heart. These tissues have the greatest expression of frataxin and might be expected to show the greatest phenotype, but if frataxin affects mitochondrial function, why are other mitochondria-rich tissues, such as skeletal muscle, not clinically affected? One explanation, suggested by Lodi et al. (1999), is that because of their disorder Friedreich ataxia patients cannot exercise to the point at which a skeletal muscle defect is apparent. A second potential answer was provided by Esposito et al. (1999), who reported that cardiac and skeletal muscle show vastly different responses to deficits in ATP generation. They demonstrated that skeletal muscle can increase antioxidant defenses to a greater level than cardiac muscle, thus rendering the latter more susceptible to oxidant damage. A third answer is that skeletal muscle derives a significant amount of energy from glycolysis, whereas cardiac myocytes derive most of their ATP from the oxidation of free fatty acids. Mitochondrial defects would preferentially be seen in tissues that are most reliant on respiratory oxidation (Kaplan, 1999).

Tan et al. (2001) reported that lymphoblasts of FRDA compound heterozygotes were more sensitive to oxidative stress by challenge with free iron, hydrogen peroxide, and free iron plus hydrogen peroxide, consistent with a Fenton chemical mechanism of pathophysiology. After transfecting the FRDA gene into FRDA compound heterozygous cells, FRDA mRNA and protein were produced at near-physiologic levels, and sensitivity to iron and peroxide was reduced to control levels. The FRDA compound heterozygous cells had decreased mitochondrial membrane potential as well as lower activities of aconitase and ICDH (2 enzymes supporting mitochondrial membrane potential), and twice the level of filtrable mitochondrial iron. Iron challenge caused increased mitochondrial iron levels and decreased mitochondrial membrane potential, both of which resolved after transfection. Since free iron is toxic, the observation that frataxin deficiency (either directly or indirectly) causes an increase in filtrable mitochondrial iron suggests a hypothesis for the mechanism of cell death in Friedreich ataxia.

Using the differential display technique, Pianese et al. (2002) demonstrated downregulation of mitogen-activated protein kinase kinase-4 (MAP2K4; 601335) mRNA in frataxin-overexpressing cells. Frataxin overexpression also reduced c-Jun N-terminal kinase (see 601158) phosphorylation. Furthermore, exposure of FRDA fibroblasts to several forms of environmental stress caused an upregulation of phospho-JNK and phospho-c-Jun. A significantly higher activation of caspase-9 (CASP9; 602234) was observed in FRDA versus control fibroblasts after serum withdrawal. The authors suggested the presence, in cells from patients with FRDA, of a 'hyperactive' stress signaling pathway, and proposed that the role of frataxin in FRDA pathogenesis could be explained, at least in part, by this hyperactivity.

Friedreich ataxia is characterized by a variable phenotype which may also include hypertrophic cardiomyopathy and diabetes. Giacchetti et al. (2004) reported an influence of mtDNA haplogroups on the Friedreich ataxia phenotype. Patients belonging to mtDNA haplogroup U were found to have a delay of 5 years in the onset of manifestations and a lower rate of cardiomyopathy.

Mitochondrial ferritin (FTMT; 608847) is a nuclear-encoded iron-sequestering protein that is localized in mitochondria. Campanella et al. (2009) analyzed the effect of FTMT expression in HeLa cells after incubation with hydrogen peroxide (H2O2) and antimycin A, and after long-term growth in glucose-free media that enhanced mitochondrial respiratory activity. FTMT reduced the level of reactive oxygen species (ROS), increased the level of ATP and activity of mitochondrial Fe-S enzymes, and had a positive effect on cell viability. FTMT expression in fibroblasts from FRDA patients prevented the formation of ROS and partially rescued the impaired activity of mitochondrial Fe-S enzymes, caused by frataxin deficiency.

Coppola et al. (2009) performed microarray analysis of heart and skeletal muscle in a mouse model of frataxin deficiency, and found molecular evidence of increased lipogenesis in skeletal muscle, and alteration of fiber-type composition in heart, consistent with insulin resistance and cardiomyopathy, respectively. Since the peroxisome proliferator-activated receptor-gamma (PPARG; 601487) pathway is known to regulate both processes, the authors hypothesized that dysregulation of this pathway could play a key role in frataxin deficiency. They demonstrated a coordinate dysregulation of the Pparg coactivator Pgc1a (PPARGC1A; 604517) and transcription factor Srebp1 (SREBF1; 184756) in cellular and animal models of frataxin deficiency, and in cells from FRDA patients, who have marked insulin resistance. Genetic modulation of the PPAR-gamma pathway affected frataxin levels in vitro, supporting PPAR-gamma as a potential therapeutic target in FRDA.

Al-Mahdawi et al. (2008) found decreased FXN expression in brain and heart tissue, at 23% and 65% of normal levels, respectively, in postmortem specimens from 2 FRDA patients compared to normal controls. Bisulfite sequence analysis showed consistent hypermethylation of CpG sites upstream of the GAA repeat region and hypomethylation of CpG sites downstream of the repeat region. The upstream GAA DNA methylation changes in both FRDA brain and heart were consistent with their proposed roles in inhibition of FXN transcription. The methylation profiles of Fxn transgenic mouse brain and heart tissues resembled the profiles of human tissue, with cerebellar tissues most affected in the brain. Chromatin immunoprecipitation analysis showed histone modifications in human FRDA brain tissue, with overall decreased histone H3K9 acetylation, particularly downstream of the GAA repeat, and increased H3K9 methylation. The findings suggested a major role for DNA methylation and histone changes in the inhibition of FXN transcription in tissues affected by the disorder, as well demonstrating the importance of epigenetic changes that affect heterochromatin structure. Al-Mahdawi et al. (2008) proposed that histone deacetylase (HDAC) inhibitors may be of therapeutic use by increasing acetylation of histones and thereby increasing FXN transcription in FRDA cells.

Role of Trinucleotide Repeat

Using RNase protection assays, Bidichandani et al. (1998) showed that the GAA repeat per se interferes with in vitro transcription in a length-dependent manner, with both prokaryotic and eukaryotic enzymes. This interference was most pronounced in the physiologic orientation of transcription, when synthesis of the GAA-rich transcript was attempted. These results were considered consistent with the observed negative correlation between triplet-repeat length and the age at onset of disease. Using in vitro chemical probing strategies, they also showed that the GAA triplet repeat adopts an unusual DNA structure, demonstrated by hyperreactivity to osmium tetroxide, hydroxylamine, and diethyl pyrocarbonate. These results raised the possibility that the GAA triplet repeat expansion may result in an unusual yet stable DNA structure that interferes with transcription, ultimately leading to a cellular deficiency of frataxin.

Sakamoto et al. (1999) described a novel DNA structure, sticky DNA, for lengths of (GAA-TTC)n found in intron 1 of the frataxin gene of patients with Friedreich ataxia. Sticky DNA is formed by the association of 2 purine-purine-pyrimidine (R-R-Y) triplexes in negatively supercoiled plasmids at neutral pH. GAA-TTC repeats of more than 59 copies formed sticky DNA and inhibited transcription in vivo and in vitro. (GAAGGA-TCCTTC)65, also found in intron 1 of the frataxin gene, did not form sticky DNA, inhibit transcription, or associate with the disease. These results suggested that R-R-Y triplexes and/or sticky DNA may be involved in the etiology of Friedreich ataxia. The trinucleotide repeat expansion (TRE) reduces gene transcription (Ohshima et al., 1998), probably because it forces DNA to adopt a sticky conformation (Sakamoto et al., 1999).

To elucidate the mechanism by which sticky DNA inhibits transcription, Sakamoto et al. (2001) performed in vitro studies and showed that the amount of RNA synthesized decreased as the number of (GAA-TTC)n repeats