Friedreich’s ataxia – yesterday, today and tomorrow
The present review traces the origin of Friedreich’s Ataxia (FA) from the time of Nikolaus Friedreich in the mid-nineteenth century. The early hesitation on the part of the neurological community in accepting FA as a distinct entity, rather than a variant form of tabes dorsalis and multiple sclerosis, has been highlighted. Research within the last 6-7 years, has firmly established FA as a trinucleotide repeat disorder, the location of the offending gene, and the disease-related gene product, frataxin. Frataxin is now thought to interfere with the mitochondrial oxidative process and enhance iron accumulation. However, whether this iron accumulation is a primary causative event for symptom production is not clear and iron chelators are unlikely to be helpful in therapy. Of great promise is the use of free radical scavengers and antioxidants. One such agent idebenone, a short chain analogue of co-enzyme Q10, may have a future.
Friedreich's ataxia is an autosomal recessive disease with a prevalence of between 1 and 2 per 100,000, characterized by symptoms and signs including progressive ataxia, absent tendon reflexes in the legs, distal impairment of position and sense of vibration, Babinski reflexes, and dysarthria. Other signs may be present, such as pes cavus, scoliosis, diabetes mellitus, and cardiomyopathy. In most cases it starts between the age of 8 and 15, the patients becoming wheelchair-bound after approximately 14 years. During the nineteenth century it was generally distinguished from other types of ataxia. Nikolaus Friedreich played a major role in defining the disease.
Nikolaus Friedreich, the son of Johannes Baptist was born in Wurzburg in 1825. His grandfather Nikolaus Anton (1761-1836), Professor of surgery, described peripheral facial nerve paralysis in 1798, some 23 years before Sir Charles Bell. Nikolaus (Jr.) began studying medicine in his native city in 1844 and went to Heidelberg for 6 months in 1847. He graduated in 1850 and became first assistant of the blind physician Karl Friedrich Marcus (1802-1862).,,
In 1858, Friedreich was appointed chief of the medical clinic in Heidelberg, where he took the chair of pathology and therapy, remaining there for the rest of his life.
His main interest was diseases of the nervous system. One of the early works in this field was his study on progressive muscular atrophy, which he dedicated to his teacher Virchow. Friedreich, however, erroneously considered all muscular atrophy of myogenic origin. His most important contribution to neurology was the study of hereditary spinal ataxia. In addition to the eponymic Friedreich's ataxia, his name also became associated with paramyoclonus multiplex. He took detailed histories of patients and performed examinations with great precision. As such, he was an exemplary teacher and had a very extensive practice. According to Kussmaul, there was not a greater expert in physical examination than Friedreich:
Nobody could equal his art of percussion; one could find no better percussion hammer or plessimeter than Friedreich's finger. Even in the most obese person could he arouse clearly perceptible sounds. He taught many well-known physicians, including, Kussmaul, Friedrich Schultze (1848-1934), and Wilhelm Erb who succeeded him to the chair of pathology and therapy in Heidelberg in 1882.
Friedreich began his observations on ataxia in the 1850s and discussed the patients with his colleagues, including Kussmaul and Virchow. Gradually, he realized that these patients were not suffering from ordinary locomotor ataxia as described in 1858-59 by Guillaume Benjamin Amand Duchenne (1806-1875) but from a variant. Friedreich presented the first data at a meeting in Speyer, Germany in 1861. Subsequently, he published several papers on the hereditary type of ataxia between 1863 and 1877., In the introduction to the first paper, he expressed his expectation that he would delineate a characteristic type of spinal degeneration within the group of tabes dorsalis. In the 1863 papers, he described 6 patients in 2 families. He added 3 other patients in the 1876-1877 papers, resulting in a total of 9 patients from 3 families.
The disease could be distinguished from tabes dorsalis because of the hereditary occurrence, the early age of onset, the long duration, the fact that 7 of the 9 patients were females, and the absence of sensory loss, at least in the early stages of the disease. He found that the lower part of the spinal cord was involved at the beginning of the disease, later it spread to involve the medulla oblongata. The disease started with ataxia, initially present only in the legs, later in the arms. Dysarthria was found later, sometimes accompanied by nystagmus and scoliosis. Sensory loss in the legs was observed only after many years. Loss of joint sensation was present in only 1 patient, in such a way that walking in darkness or with the eyes closed was more difficult.
Autopsy on 4 patients demonstrated degeneration of the dorsal spinal columns, particularly in the lower part. Upon microscopic examination, nerve atrophy and demyelination were noticed as well as replacement by fine fibrillary tissue. The lateral columns were also involved, and degeneration of cells was found in Clarke's column as well as in the hypoglossal nuclei. Interestingly, Friedreich concluded the article by mentioning a remarkable fatty degeneration of the cardiac muscle in 3 cases, which he associated with the increasing tendency of the patients to collapse.,
His belief that he had discovered a new disease was not endorsed by the medical world at first. Several colleagues considered it a variant of tabes dorsalis. Charcot's pupil Bourneville, in Paris, believed it was multiple sclerosis, as he thought nystagmus and dysarthria did not belong to the syndrome of locomotor ataxia. The lack of sensory and visual disturbances was unusual as well. It is obvious that many physicians tried to fit the syndrome into one of the existing diagnoses, including tabes dorsalis or multiple sclerosis. Friedreich replied to Bourneville in his 1876-1877 papers, pointing to the pathologic-anatomic findings in the dorsal columns of the spinal cord.
In a chapter on tabes dorsalis, Wilhelm Erb mentioned “Friedreich's type of tabes” in 1878. William Gowers wrote on hereditary ataxia, or Friedreich's disease, in 1880, and again in the first volume of his well-known Manual (1886), stating that in approximately 65 cases distributed in 19 families, “the family tendency of the disease……shown by the affection……of brothers and sisters,” had been recorded, including 57 cases reviewed by a certain Dr. Erverett Smith (1885, he observed 6 cases and presented a review of 51, including 6 cases of Friedreich). One year later, Bury summarized data of 100 published cases.
In the meantime, Brousse of Montpellier dedicated a thesis to the “maladie de Friedreich” (1882). The approval of Friedreich's findings, however, had to await Charcot's work 2 years later. In his 1876 paper, Friedreich had already expressed his wish that Charcot should find similar cases among his patients. Indeed, Charcot demonstrated a young patient suffering from a hereditary type of ataxia, not fitting the diagnosis of tabes or multiple sclerosis, in 1884.,
The New Genetics
Many patients with cerebellar ataxia, since the time of Friedreich, even with autosomal dominant transmission, were diagnosed as FA until Anita Harding introduced recessive inheritance as a major diagnostic criterion. She defined specific symptoms as essential or secondary to the diagnosis of FA, even at an early stage. The clinical spectrum of the disease has however changed since that time with the discovery of the causative gene and the documentation of the age of onset as late as 62 years.
In about 98% of patients with FA, the causative mutation is an expansion of a GAA trinucleotide repeat located in the first intron of the gene on chromosome 9q13. This is the first disease-causing, autosomal recessive GAA expansion found in an intron. Most autosomal dominant cerebellar ataxias are due to CAG expansions in coding regions. However, 2 recently identified forms, SCA10 and SCA12 are caused by non-coding expansion, ATTCT and CAG respectively. But these are autosomal dominant conditions.
The size of pathological expansion in FA is variable, ranging from 90 to 1700 units. Most normal alleles carry 6 to 9 repeats and never exceed 34 repeats.,, As in other trinucleotide-repeat diseases, expansions are unstable when transmitted. Male transmission is almost always associated with contractions, whereas both expansions and contractions are observed during female transmission.,,, Studies of sperm showed that the number of repeats can change prezygotically and postzygotically, which suggests that the expansion is unstable during both meiotic and mitotic division.,, As in the fragile X syndrome, alleles in the range between normal and expanded alleles are premutations, which are prone to expansion into the pathological range.,, Haplotype studies have shown that mutations (and premutations) originate from a small pool of large-normal alleles (14-34 repeats). All large-normal alleles could have evolved from a single event that transformed a small-normal (6 to 12 repeats) allele into a large-normal allele. Large-normal alleles are a pool for accidental expansions, and replace pathological expansions that disappear because of the reduced reproductive fitness of patients with FA with reversions of expansions to normal repeat length.
As in other diseases caused by triplet-repeat expansion, normal variation of the number of repeats includes a subset of alleles with a large but non-pathogenetic number of units. These large-normal alleles are prone to further expansion and constitute a reservoir for the disease in the populations in which they are present. Linkage disequilibrium analysis showed that more than 90% of large-normal alleles originated from a single founder event, which explains why FA seems to be more prevalent in the white population. FA should also be considered a possible diagnosis in adult-onset cerebellar ataxia. It is more common among sporadic cases than autosomal dominant cerebellar ataxias (5.2% as compared with 4.4%).
The exact incidence of FA in India is not known. Jagannathan mentioned FA in his Presidential Oration on Cerebellar Ataxia during the Annual Conference of the Neurological Society of India in 1980. But this was before the diagnostic criteria were clearly defined and the genetic basis became known. In a recently conducted study on primary cerebellar degenerations amongst ethnic Bengalees in West Bengal, India, Chakravarty and Mukherjee encountered 12 patients from 8 families showing characteristic clinical features of FA and autosomal recessive inheritance. However, genetic studies confirmed GAA expansion only in 1 family studied. All 8 families were Muslims-a community in which consanguinous marriages are common. These workers also encountered 4 patients from 2 families with features of FA but with retained reflexes. This subgroup was recognized by Filla et al in 1990, though mentioned earlier by Anita Harding herself in her classification in 1983. With the description of a shorter length of expansion (vide infra) more interest has been generated regarding this subgroup, which may occur in adult life causing spastic paraparesis. Mukerjee et al from India, described the genetic analysis of 3 families with FA. All were found to be homozygous for GAA repeat expansions. The GAA repeat in the normal population showed a bimodal distribution with 94% of alleles ranging from 7-16 repeats. The low frequency of large-normal alleles (6%) could indicate that the prevalence of the disease in the Indian population is likely to be low. However, such studies need to be performed in different ethnic populations of the country.
A small percentage of early onset cerebellar ataxia resembling FA, may result from a point mutation in the FA gene locus. A very different point mutation may lead to early onset ataxia resembling FA but with retained reflexes. This ataxia (ARSACS - Autosomal Recessive Spastic Ataxia Charlevoix Saguenay) occurs in French Canadians living in the Charlevoix-Saguenay region of Quebec, Canada. Similar cases have also been reported from Tunisia showing linkage to the ARSACS region on Chromosome 13q.
Frataxin and FA
The expansion causes a severe decrease in the expression of frataxin, a 210 amino-acid protein that is localized in the mitochondrial matrix.,,, Experiments in vitro and in vivo have shown that the reduction in amounts of frataxin mRNA depends on the length and orientation of the expansion and is due to the disruption of transcription rather than splicing. Sakamoto and co-workers hypothesize that the GAA-rich sequence is “sticky” DNA, which self-associates to form triple-helix structures. The formation of such a purine-purine pyramidine DNA triplex behind an advancing RNA polymerase could trap the enzyme and suppress frataxin gene expression. The formation of such triplex structures is inhibited by the presence of GGA repeats within the GAA expansion.
Some patients with FA are compound heterozygous; they carry a GAA expansion on one allele and a point mutation on the other. A patient with two point mutations, but no expansion, has not been identified yet. Perhaps because carriers of point mutations are very rare in the general population or because double point mutations are lethal. Embryonic lethality is observed in transgenic mice that lack frataxin.
There are both truncating and mis-sense mutations. All mis-sense mutations affect the carboxyl-terminal moiety of the protein.
FA: Genotype - Phenotype Correlation
The clinical spectrum of FA was greatly broadened by the screening of patients with ataxia for mutations in the frataxin gene., The clinical criteria established by Harding (autosomal recessive inheritance, onset before age 25 years, absence of leg reflexes, and pyramidal tract involvement) remain very specific. Only a few patients with this typical phenotype do not carry the GAA expansion in the frataxin gene. By contrast, many patients who do not fulfill these criteria have confirmed FA. Linkage studies have shown that the same locus is involved in patients with onset after age 25 and retained reflexes.,,,
The disease generally develops around puberty, but late-onset cases in the sixth or even the seventh decade of life have been reported.,, The most common symptoms at onset are progressive gait instability or general clumsiness. In early-onset cases, scoliosis or cardiomyopathy can precede ataxic gait. Very atypical presentations, spastic ataxia in late-onset FA or pure sensory ataxia, have also been observed. In some instances, unusual signs such as ophthalmoplegia, myoclonus, or chorea are present. As the disease progresses patients are significantly hampered not only by the ataxia but also by hypotonia, which combined with ataxia results in difficulties with fine motor activities, walking, standing, and sitting down without support. Dysarthria is progressive with slow, jerky speech. Pyramidal tract involvement with extensor plantar responses is present in most cases, but exceptions are well known. Sensory neuropathy results in a loss of the sense of vibration in the ankles and areflexia, and is one of the constant features in the disease. Sensory neuropathy can help to distinguish FA from other ataxia. In contrast to autosomal dominant ataxias such as SCA3, sensory neuropathy is nearly always clinically symptomatic in FA patients and associated with severe reduction or loss of sensory action potentials without involvement of motor conduction velocities.
In rare cases, sensory neuropathy can be absent even after a long duration of the disease. In extreme cases, signs of pyramidal tract involvement are isolated or prominent, and spasticity can be the presenting sign in patients homozygous for small GAA expansion. The association with signs suggestive of mitochondriopathies such as optic atrophy and hearing loss are observed in patients at an advanced stage of the disease. Ocular symptoms are rare, however, nystagmus is present in less than half of the patients, but fixation instability with square-wave jerks is almost always present. Neuroimaging reveals thinning of the cervical spinal cord and cerebellar atrophy in more advanced stages of the disease.
Two clinical aspects of the disease need special mention and both these may be linked to the genetic defect. These include cardiomyopathy and impaired glucose tolerance.
In Friedreich's original reports 3 patients died with fatty infiltration of the heart. However, it was only in 1938 that Loiseau confirmed the association of cardiac muscle involvement in FA. Cardiomyopathy, usually its hypertrophic form, is a cardinal feature of FA. The main and more frequent clinical manifestations indicative of myocardial involvement are rhythm disturbances and myocardial insufficiency terminating in congestive cardiac failure. The combination of electrocardiography and echocardiography detected one or more abnormalities in 95% of patients in Child et al's series. It appears that there are two fundamentally different types of cardiac disease in FA., Firstly, a common dystrophic form manifested usually by electrocardiographic initial force deformation without detectable echographic wall motion abnormalities, but occasionally by extension throughout the left ventricle with global hypokinesia and reduced QRS voltage. The second is a hypertrophic form represented by symmetric or asymmetric left ventricular hypertrophy with normal cavity size and ventricular function. Cardiac abnormalities are found more commonly in those with GAA expansion than in those without. The relationship of this cardiomyopathy to frataxin deficiency and the resultant mitochondrial dysfunction is mentioned in the next section of this review.
There is an increased incidence of diabetes in FA patients together with normal or sometimes increased stimulus-coupled insulin secretion in FA patients and in some of their first-degree relatives.,, Evidence of the possible relation of this to the GAA expansion and frataxin has been provided by two recent studies. Ristow et al demonstrated that the X25/frataxin GAA polymorphism is associated with non-insulin dependent diabetes mellitus in a frequency higher than any other mutation heretofore described. Hebinck et al demonstrated that a heterozygous expansion of the X25/frataxin GAA repeat in healthy individuals is associated with insulin resistance and might be considered a genetic co-factor in the pathogenesis of mitochondrial subtypes of diabetes.
Most patients with heterozygous point mutations present a typical clinical phenotype. However, owing to the limited number of patients with the same point mutation, no firm relation between the genotype and phenotype can be established. The only exception to this is the G130V mutation, which arose from a single ancestral mutation event, and is now found in a heterozygous state in several families. Most patients with the G130V mutation have early-onset FA with pyramidal tract involvement, which causes spasticity. Ataxia and dysarthria are mild or even absent in these patients. In many cases, fundoscopy reveals optic-disc atrophy. Surprisingly, despite early onset, the progression of the disease is slow in patients with the G130V mutation.
ARSACS patients present in early childhood with spastic gait ataxia. The disease progresses rapidly in young adults and patients are wheelchair-bound by their fifth decade. Sensory nerve conduction is abnormal and motor conduction is reduced. MRI shows cerebellar vermis atrophy. Retinal striations (prominent myelinated fibers) are a distinctive diagnostic feature. The Tunisian patients showed similar clinical features except for variable onset age, earlier loss of ankle jerks and absence of retinal striations.
Is FA a mitochondrial disease?
The involvement of extraneural organs during the course of the disease led to the idea of mitochondrial involvement in the pathogenesis of FA almost 25 years ago. That frataxin is a mitochondrial protein was confirmed by colocalization studies with mitochondrial markers,, and by electron microscopy studies that showed its localization at mitochondrial membranes and crests. The next important step took advantage of the existence of a yeast homologue of frataxin. When the yeast homologue was deleted, inhibited growth and reduced mitochondrial respiration were observed. Moreover, the mutant yeast was more sensitive to oxidative stress, and iron accumulated in the mitochondria. The human frataxin gene can complement the deletion yeast homologue.
Therefore, the yeast model suggested that iron accumulation in the mitochondria of patients with FA would result in hypersensitivity to oxidative stress. Consistent with this hypothesis is the observation of iron deposits in cardiomyocytes, of increased iron detected in the mitochondria of fibroblasts of FA patients, and in the dentate nucleus of FA patients analyzed by MRI. The final proof that FA is a mitochondrial disorder came from studies in vitro and in vivo. Rotig and co-workers studied cardiac biopsy samples from 2 patients with cardiomyopathy who turned out to have FA. They detected deficiencies of respiratory-chain complexes I, II and III and aconitase; these deficiencies correspond to deficits of iron-sulfur enzymes. These results were confirmed both by autopsy samples from 9 patients with FA and in vivo by quantification of phosphate and ATP concentrations, by use of phosphorus magnetic resonance spectroscopy, in the patients' skeletal muscle. Patients with FA have a lower than normal rate of mitochondrial ATP production after exercise. The decreased oxidation activity shows a strong negative correlation with the number of GAA repeats in the smallest allele. This result confirms that FA is a mitochondrial disorder and that the GAA expansion is the cause of the disease.
Frataxin may be directly involved in iron metabolism, but murine models support the hypothesis that iron deposits are not the initial cause of the disorder. This notion does not exclude a direct role of frataxin in iron metabolism, and frataxin may be an iron-binding protein.
Hope for a therapy?
There is yet no curative treatment for FA. The recent advances in the understanding of frataxin function and the consequences of its greatly reduced expression in patients with FA, and models of the disorder led to a different therapeutic approach. The observation of iron accumulation in both the human disease and the yeast models suggested that iron chelators could be useful in the treatment of FA. However, several lines of evidence argue against this approach. Firstly, desferrioxamine or phlebotomy can decrease plasma iron concentrations, which are normal in FA patients, but not the concentration of iron in mitochondria. Secondly, studies in vitro have shown that desferrioxamine can protect respiratory-chain complex II activity and lipids from oxidation by iron but reduces aconitase activity; suggesting that iron chelators could displace rather than protect against the toxic effects of ferrous iron. Thirdly, in the previously described mouse model, increased iron concentration is a secondary event and not the causative pathological mechanism.
Several lines of evidence support the hypothesis of an early impairment of oxidative defenses in FA. This mechanism is supported by similar clinical presentation, except for cardiomyopathy found in ataxia caused by inherited deficiency of vitamin E. This hypothesis led to the idea that treatments aimed at reducing the load of free radicals could slow disease progression. Rustin and co-workers used their in vitro model to show that among the known antioxidants, ascorbic acid is not protective but idebenone, a short-chain analogue of coenzyme Q10, does protect against cardiac hypertrophy. Their demonstration was supported by preliminary results indicating a striking reduction of cardiomyopathy after several months of treatment in 3 patients. The reduction of cardiomyopathy has now been confirmed. However, echocardiographic assessment of 9 patients with FA treated for 6 weeks with idebenone (360 mg/day) and of 10 patients treated with coenzyme Q10 (400 mg/day) and vitamin E (2100 IU/day) for 6 months did not show significant improvement. The lack of significant improvement might be due to the small number of patients included in both studies. In addition, phosphorus magnetic resonance spectroscopy revealed a 139% increase in the maximum rate of ATP production by skeletal-muscle mitochondria in one study but not in another. The study by Lodi and colleagues provides evidence of partial reversal of a surrogate biochemical marker, which occurred as early as 3 months after therapy and which was sustained in FA patients treated with antioxidants. Artuch et al have also reported some encouraging results. Cerebellar improvement was notable in mildly affected patients after the first 3 months of therapy when assessed by the International Cooperative Ataxia Rating Scale (ICARS). Idebenone treatment at an early stage of the disease seemed to reduce progression of cerebellar manifestation.
More recently, the studies by Mariotti et al and Buyse et al have demonstrated improvement in echocardiographic parameters in FA patients treated with idebenone 5 mg/kg daily. Filla and Moss commented that one needs to know whether idebenone and related agents could reverse the metabolic derangements caused by frataxin deficiency and whether these agents in higher doses and for longer periods may be more beneficial. Answers to these questions and many more and also a safety profile are needed before free radical scavengers can be recommended for routine use in patients with FA.
FA is a slowly progressive disease which lingers for long, there is no alternative but to be patient. While the last millennium witnessed clinical and genetic characterization of the disease, the present one is sure to find a way to modify the disease course.