Article Text

Download PDFPDF

Review
Autosomal recessive cerebellar ataxias: the current state of affairs
  1. S Vermeer1,
  2. B P C van de Warrenburg2,
  3. M A A P Willemsen2,
  4. M Cluitmans1,
  5. H Scheffer1,3,
  6. B P Kremer4,
  7. N V A M Knoers1,5
  1. 1Department of Human Genetics, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  2. 2Department of Neurology, Donders Institute of Brain, Cognition and Behaviour, Centre for Neuroscience, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  3. 3Institute for Genetic and Metabolic Disease, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  4. 4Department of Neurology, University Medical Center Groningen, Groningen, The Netherlands
  5. 5Department of Medical Genetics, University Medical Centre Utrecht, Utrecht, The Netherlands
  1. Correspondence to S Vermeer, Department of Human Genetics, Radboud University Nijmegen Medical Centre, P.O. Box 9101, Nijmegen 6500 HB, The Netherlands; s.vermeer{at}antrg.umcn.nl

Abstract

Among the hereditary ataxias, autosomal recessive cerebellar ataxias (ARCAs) encompass a diverse group of rare neurodegenerative disorders in which a cerebellar syndrome is the key clinical feature. The clinical overlap between the different cerebellar ataxias, the occasional atypical phenotypes, and the genetic heterogeneity often complicate the clinical management of such patients. Despite the steady increase in newly discovered ARCA genes, many patients with a putative ARCA cannot be genotyped yet, proving that more genes must be involved. This review presents an updated overview of the various ARCAs. The clinical and genetic characteristics of those forms with a known molecular genetic defect are discussed, along with the emerging insights in the underlying pathophysiological mechanisms.

  • Autosomal recessive cerebellar ataxia
  • phenotype
  • molecular pathogenesis
  • movement disorders (other than Parkinson's)
  • neurology
  • neurosciences

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Ataxia due to involvement of the cerebellum, brainstem and/or spinocerebellar long tracts is the key feature of the spinocerebellar degeneration. Clinically, patients are characterised by gait impairment, poor balance with frequent falls, disturbed upper limb coordination, dysarthria, and eye movement abnormalities. Acquired and hereditary forms of degenerative spinocerebellar ataxias are distinguished. Based on the mode of inheritance or on the identified genetic defect, hereditary ataxias can be further divided into autosomal dominant, autosomal recessive, X-linked, and maternally inherited (mitochondrial) disorders.

This review provides an updated clinical, genetic, and pathophysiological overview of the currently known autosomal recessive cerebellar ataxias (ARCAs) (tables 1–3). In the past few years, several novel forms of ARCA have been recognised, based on identified novel loci and genes. Some of these are, however, very rare, restricted to isolated populations, or have been described in single families only (table 2). For some of the ARCAs only a locus is known (table 3). We will focus on those ARCAs in which progressive cerebellar ataxia is a prominent feature, which are not restricted to single families, and for which the molecular genetic defect is known. The classic and previously extensively reviewed ARCAs will be discussed only briefly. The various recessive metabolic diseases in which cerebellar ataxia is a clear associated feature are listed in table 4 and will not be described in detail.

Table 1

Autosomal recessive cerebellar ataxias

Table 2

Autosomal recessive cerebellar ataxias so far only described in one family or in isolated populations

Table 3

Autosomal recessive cerebellar ataxias for which only a locus is known

Table 4

Metabolic diseases with cerebellar ataxia as an associated feature

Clinical prologue

The clinical and genetic heterogeneity among the hereditary ataxias complicates their clinical recognition and their diagnostic work-up. In ARCAs (tables 1–3) progressive cerebellar ataxia is usually, yet not invariably, the salient feature. However, often other neurological or non-neurological symptoms coexist. The overall estimated prevalence is approximately 2.2–7 per 100 000, but varies per population.7

As the different phenotypes are often not so well defined and may show overlapping features, it can be difficult to make a correct clinical diagnosis or request the appropriate mutation analysis. For instance, Friedreich's ataxia (FA) usually presents as an early onset cerebellar ataxia with neuropathy, while a (very) late onset FA can be associated with marked spasticity of the lower limbs. This may resemble autosomal recessive spastic cerebellar ataxia of Charlevoix–Saguenay (ARSACS) or spastic paraplegia type 7 (SPG7), the latter being a type of hereditary spastic paraplegia (HSP) that can also feature a prominent cerebellar ataxia.8

Different classifications of the recessive cerebellar ataxias based on either clinical phenotype, neuropathology, or pathogenesis have been proposed. However, from a clinician's initial point of view, we suggest to look at the prominent features of the different ARCAs and then determine which ARCA to consider first. Table 5 gives an overview of the different prominent features that can be distinguished among the ARCAs discussed in this review. This table can be used as a general guideline, and allows rational selection of an initial genetic screen.

Table 5

Which autosomal recessive cerebellar ataxias to consider based on prominent features

Given the phenotypic heterogeneity alluded to earlier, a more detailed evaluation of the phenotype, including paraclinical studies such as laboratory tests and neuroimaging (in particular the presence of cerebellar atrophy is a helpful distinguishing feature figure 1), will be helpful for further differentiation.

Figure 1

Tranversal, T2 weighted MRI showing midline cerebellar atrophy.

Friedreich's ataxia

This is the most common and best characterised recessive form of cerebellar ataxia. Typical FA presents as a progressive gait and limb ataxia with onset before or around puberty, dysarthria, sensory neuropathy, deep sensory impairment, and signs of pyramidal tract involvement. An excellent overview of the clinical picture of FA was given by Pandolfo.9

The estimated prevalence in the western world is about 1 in 29 000, with an estimated carrier frequency of 1:85.10

FA is the result of a homozygous intronic expansion of a GAA trinucleotide repeat located in the first intron of FXN, located on chromosome 9q13.11 Normal alleles usually carry 6–9 GAA repeats. Repeats at the upper end of normal (14–34 repeats) can expand during transmission into an intermediate range of allele length (up to 90 repeats) or into clearly disease causing alleles containing GAA expansions of 90–1700 repeats.12 In a minority (2–5%) of cases, the disease is caused by a heterozygous point mutation on one allele combined with a GAA expansion on the other allel.13 The mutations, including splice site mutations and deletions, occur throughout the whole gene.

FA is caused by partial deficiency of the mitochondrial protein frataxin, which is encoded by FXN. The function of the protein is still disputed, but it is known to be involved in cellular iron homeostasis. Frataxin binds ferrous iron, promoting the mitochondrial synthesis of iron-containing molecules, in particular iron–sulphur clusters (ISCs)14 and haem.15 It controls the ability of iron to perform redox chemistry.16 Frataxin deficiency leads to a dysfunction of the respiratory chain complexes and Krebs cycle components, due to inappropriate iron–sulphur cluster synthesis, thus provoking bioenergetics failure and subsequent cell death.17 Complete absence of frataxin is incompatible with life in higher organisms, as demonstrated by embryonic lethality seen in systemic gene knock-out models.18

Potential disease modifying therapeutic strategies aim at reducing the load of free radicals in order to slow disease progression. Trials with idebenone have shown a reduction in cardiac hypertrophy,19 but no neurological benefit from this treatment has been observed. However, based on results of the NICOSIA trial, in which higher doses of idebenone suggested some effect on neurological function,20 idebenone is now being used for symptomatic treatment of FA patients in Canada. In Europe and the USA, further trials with idebenone so far have not confirmed the results of the NICOSIA study.21 Another potential therapeutic strategy is aimed at increasing frataxin expression by treatment with recombinant human erythropoietin22 or histone deacytelase inhibitors.23 It has also been shown that peroxisome proliferator-activated receptor gamma (PPARγ) co-activator 1α, a transcriptional master regulator of mitochondrial biogenesis and antioxidant responses, is downregulated in most cell types from FA patients and animal models. This downregulation may play a role in the blunted antioxidant response observed in cells from FA patients. This response can be restored by PPARγ agonists, suggesting yet another potential therapeutic approach to FA.24

Hereditary ataxia with vitamin E deficiency

This disorder resembles FA clinically, but the biochemical hallmark of ataxia with vitamin E deficiency (AVED) is a very low plasma concentration of vitamin E (usually <3 μmol/l, normal 7–33 μmol/l) in the absence of intestinal fat malabsorption and/or abetalipoproteinaemia. The disease starts mostly before 20 years, with progressive cerebellar ataxia, dysarthria, reduced or absent deep tendon reflexes, and vibration sense impairment.25 Decreased visual acuity or retinitis pigmentosa may be an early finding. Cardiomyopathy is less frequent in AVED compared to FA. Patients seem to have more head titubation, less neuropathy, and a slower disease course compared to FA.26 Mild cerebellar atrophy is present in almost 50% of cases.27 The phenotypic variability of AVED can range from a severe FA-like phenotype28 to mild neurological impairment,29 very late disease onset,30 or myoclonic dystonia as the initial presentation.31

The gene involved is called α-tocopherol transfer protein (TTPA).32 Genotype–phenotype correlation studies suggest that truncating TTPA mutations correlate with a more severe phenotype than missense mutations.27

The α-tocopherol transfer protein (α-TTP) mediates the incorporation of vitamin E into circulating lipoproteins, and mutations presumably lead to reduced vitamin E (α-tocopherol) concentrations. Studies in mice indicate that α-TTP may be critical for the uptake of α-tocopherol in the brain and that next to its anti-oxidant function, α-tocopherol can modify mitochondrial reactive oxygen species production.33 Further studies are necessary to unravel the exact pathogenic mechanism.

Importantly, it is known that (early) supplementation with vitamin E appears to stop progression of the disease,26 and even may slightly improve cerebellar ataxia.34

The exact incidence of AVED is not known. Most patients originate from North Africa where it may be as frequent as FA.35

Autosomal recessive spastic cerebellar ataxia of Charlevoix–Saguenay

ARSACS was first described in the Charlevoix-Saguenay region of Northeastern Québec in Canada.36 Clinically, ARSACS is characterised by a triad of cerebellar ataxia, peripheral neuropathy, and lower leg pyramidal tract involvement. Disease onset is usually in early childhood, although onset in adulthood has been described.37 Oculomotor abnormalities and dysarthria are often mild. Slight dysphagia and urinary dysfunction (predominantly urge incontinence) are fairly common. Apart from mild dystonic features observed in a few patients, extrapyramidal features are absent.38 Increased demarcation of the retinal fibres embedding parts of the vessels near the disc has been noted.38 Cerebellar superior vermis atrophy, linear hypointensities in the pons, and atrophy of the cerebellar hemispheres and spinal cord can be seen on MRI.39

After identification of the SACS gene involved in ARSACS,40 increasing numbers of patients outside Canada were genetically characterised. By now, over 70 different mutations—truncating, missense, small and large deletions—have been identified. SACS encodes the large protein sacsin, of unknown function. The protein contains both a functional DnaJ domain, which is able to stimulate Hsp70 ATPase activity, as well as Hsp90-like ATPase domains. Furthermore, it harbours an N-terminal ubiquitin-like (UbL) domain. The presence of molecular chaperone domains and a UbL domain suggest that sacsin may play a specific cellular role in protein folding or complex assembly.41

Ataxia telangiectasia

Classical ataxia telangiectasia (AT) is characterised by a progressive cerebellar ataxia starting around 2–3 years of age. Most patients are wheelchair users by the age of 10 years. Oculocutaneous telangiectasias (figure 2), which usually manifest before the age of 6 years, are very characteristic of the disease. Oculomotor apraxia (OMA), which is almost always present in these patients, and dysarthria are early features. Deep tendon reflexes become diminished or absent in older patients. Chorea and dystonia are often present. Besides the characteristic telangiectasia, other non-neurological features like immunodeficiency, leading to recurrent infections and endocrine disturbances (growth failure and diabetes mellitus) complicate the clinical picture. Furthermore, in approximately one third of patients, cancer, usually of lymphoid tissues, occurs.42 Intelligence is thought to be near normal in patients with AT.43 In a very early stage of the disease, MRI of the brain may be normal, but over time severe atrophy of the cerebellum is present. In all patients with AT, serum α-fetoprotein concentrations (AFP) are elevated. Additional abnormal laboratory findings are mildly elevated serum transaminases and IgG subclass deficiency. Cytogenetic characteristics of AT are DNA radiosensitivity, particularly observable in lymphocytes,44 and chromosomal aberrations such as t(7;14) translocations. Patients with classical AT generally die in the second or third decade of life due to malignancies or respiratory failure. Some AT patients display a milder phenotype, designated ‘variant AT’. This variant phenotype is characterised by either a milder classical phenotype with a late onset45 or by predominant extrapyramidal signs, such as chorea-athethosis or resting tremor. In these variant cases lower motor neuron involvement is often present and cerebellar ataxia develops later on in the course of the disease or does not occur at all. In some variant cases oculocutaneous telangiectasias and immunodeficiency are absent and a normal cerebellum is seen on MRI, even at an older age.46

Figure 2

Conjunctival telangiectasias in a patient with ataxia telangiectasia. Telangiectasias can also be present on sun exposed skin areas.

The incidence of AT has been estimated at about 1 per 300 000 live births.47 Including AT variant cases, the estimated incidence in the Netherlands is 1 per 180 000.48

The gene involved in AT is the ATM (AT mutated) gene. It encodes a protein which is a member of the phosphatidylinosytol-3 kinase family of proteins that respond to DNA damage by phosphorylating key substrates involved in DNA repair and/or cell cycle control.49 Insights in genotype–phenotype correlations is growing: AT patients with a milder phenotype have at least one ATM missense mutation with preservation of at least some ATM protein kinase activity.46

The ATM protein kinase is involved in early cellular responses to DNA double-strand breaks (DSBs) generated during metabolic processes or by DNA-damaging agents, like radiomimetic chemicals or ionising radiation. DNA DSBs are primary sensed by the Mre11-Rad50-Nbs1 (MRN) complex. This complex then recruits ATM to the break, inducing ATM autophosphorylation, which leads to an active monomeric ATM. Once activated by the MRN complex, ATM phosphorylates members of the MRN complex and a variety of other proteins involved in cell cycle control and DNA repair. If the DNA damage is irreparable, the network eliminates affected cells by inducing apoptosis.50 ATM can also be activated through chromatin restructuring51 and it functions directly in the repair of chromosomal DNA DSBs by maintaining DNA ends in repair complexes generated during lymphocyte antigen receptor gene assembly. The latter mechanism provides a possible molecular explanation for the increase in lymphoid tumours (with translocations) in patients with AT.52 Thus, ATM helps to maintain the integrity of the genome and to minimise the risk of cancer and neurodegeneration.

The use of (routine) radiologic investigations should be minimised, while radiotherapy and some radiomimetic chemotherapeutic agents should be avoided because of the increased sensitivity of cells from individuals with AT to ionising radiation.

Ataxia telangiectasia-like disorder

Ataxia telangiectasia-like disorder (ATLD) is a very rare disorder with some clinical features that resemble those of AT but with a milder clinical course. ATLD is characterised by moderate radiosensitivity, later onset and slower progression of the disease, with longer survival and no tumour development.53 However, Oba et al recently described two siblings with ATLD who both died of lung adenocarcinoma, but it remains to be elucidated whether this is truly associated with ATLD.54 Patients with ATLD show no telangiectasias, have no immunodeficiency, and serum AFP concentrations are normal.55

ATLD is caused by mutations in the meiotic recombination 11 (MRE11) gene.56 To date, three truncating and four missense mutations in 18 patients from seven different families have been identified.

MRE11 forms the core of the multifunctional MRN complex that detects DNA DSBs, activates the ATM checkpoint kinase, and initiates homologous recombination repair of DSBs.57 Furthermore, the MRN complex plays a role during replication. It prevents replication fork associated damage during both ‘normal’ replication and under conditions of stress, and it likely has a scaffold function to maintain the replication fork in a competent conformation to resume progression during pauses in replication.58 MRE11 interacts with itself and both Rad50 and Nbs1. NBS1 is mutated in Nijmegen breakage syndrome, which shares chromosome instability and radiosensitivity with AT, but in which cerebellar pathology is absent. By using mouse models for ATLD and NBS, Shull et al demonstrated that DNA damage signalling in the nervous system is different between ATLD and NBS and likely explains their respective neuropathology.59

Ataxia with oculomotor apraxia type 1

Ataxia with oculomotor apraxia type 1 (AOA1) has some neurological similarity to AT, but in contrast to AT, AOA1 patients are not immune deficient and do not display elevated serum AFP concentrations.60 AOA1 is characterised by early onset progressive cerebellar ataxia with cerebellar atrophy on MRI, chorea, and severe axonal sensorimotor neuropathy. OMA here designated as eye–head dissociation during voluntary lateral head movements—the hallmark of the disease—occurs in approximately 80% of cases. At onset cerebellar ataxia and chorea are usually the most prominent features, whereas OMA and neuropathy are usually absent. Later, severe neuropathy dominates the phenotype. After a disease duration of 10–15 years, hypoalbuminaemia and hypercholesterolaemia are seen in most (83%) patients. The mean disease duration before patients become wheelchair users is 11 years (range 5–20). Different degrees of cognitive impairment have been described that range from severe mental retardation, cognitive deficits characteristic of a subcortical syndrome, to normal cognitive function. In France, AOA1 had a frequency of 9.1% among individuals with an early onset progressive cerebellar ataxia in whom FA has been excluded.61

AOA1 is caused by mutations in APTX that encodes the aprataxin protein of which two isoforms exists.62 The longest isoform of aprataxin contains three functional domains: an N-terminal forkhead associated (FHA) domain, a central histidine triad (HIT) domain, and a C-terminal zinc finger domain. The forkhead associated domain mediates complex formation with XRCC1-DNA ligase IIIα, suggesting a role for APTX in DNA single strand break (SSB) repair. The forkhead associated domain of aprataxin also interacts with phosphorylated XRCC4 mediating complex formation with XRCC4-DNA ligase IV, suggestive of an additional role in double strand break (DSB) repair.63 Most mutations are located in the nucleotide binding HIT domain. Cells derived from AOA1 patients are sensitive to DNA single strand break inducing agents, and show accumulation of single strand breaks under conditions of oxidative stress64 causing specific neuronal cell death. Furthermore, a recent study has demonstrated that aprataxin localises to mitochondria and preserves mitochondrial function.65 Aprataxin is expressed ubiquitously, among other structures in the caudate nuclei, thus providing a compelling explanation for chorea as part of the phenotype.66

Ataxia with oculomotor apraxia type 2

Ataxia with oculomotor apraxia type 2 (AOA2) is characterised by an onset between 3–30 years and cerebellar ataxia associated with axonal sensorimotor neuropathy.67 OMA is present in only approximately half, or even fewer, of the cases. In a study of 19 patients with AOA2, convergent strabismus was more frequent than OMA.68 Dystonic posturing of the hands, choreic movements, and head or postural hand tremor can be observed. In contrast to AOA1, the movement disorder in AOA2 usually persists and does not disappear over time. Mild cognitive impairment has been observed in some patients.68 Pronounced cerebellar atrophy on MRI is a constant and probably early feature. Increased serum AFP concentration is a good biological marker for AOA2. As elevated AFP also occurs in AT, this condition should be considered in the differential diagnosis. The study of Anheim et al, which has been performed in a part of the French population, suggests a prevalence of 1/900 000 for AOA2.69

The gene mutated in AOA2, SETX, encodes the nuclear protein senataxin, a putative DNA/RNA helicase which shares high homology to the yeast Sen1p protein.70 It plays a role in DNA repair in response to oxidative stress.71 Mutations include nonsense, missense and frame shift mutations, as well as large gene rearrangements. Most missense mutations cluster within the N-terminus and helicase domains. Missense mutations located in the N-terminal domain or the C-terminal helicase domain cause a less severe AOA2 phenotype and are more frequently associated with pyramidal signs and dystonia.72 Remarkably, heterozygous mutations in SETX may cause a dominantly inherited form of juvenile amyotrophic lateral sclerosis (ALS4).73 In addition to its role in DNA repair, senataxin plays a role in coordinating transcriptional events. Misregulation of transcription and aberrant pre-mRNA processing might contribute to the oxidative stress and the neurodegeneration observed in AOA2.74

Sensory ataxic neuropathy, dysarthria and ophthalmoparesis (SANDO) syndrome, and mitochondrial recessive ataxia syndrome (MIRAS)

Two ataxia syndromes—sensory ataxic neuropathy, dysarthria and ophthalmoparesis (SANDO) syndrome and the mitochondrial recessive ataxia syndrome (MIRAS)—have been described that share a common underlying molecular pathology. SANDO is characterised by adult onset, severe sensory ataxic neuropathy with dysarthria, ataxic gait and progressive external ophthalmoplegia (PEO).75 MIRAS refers to a syndrome with sensory ataxia, decreased or absent deep tendon reflexes of the lower limbs, epilepsy and neuropathy. Cognitive impairment is mild to moderate and psychiatric symptoms together with involuntary movements have been described. MRI of the brain often displays white matter changes and thalamic lesions, but in some patients only minor cerebellar atrophy or subtle high signal changes dorsally from the dentate nucleus occur (figure 3). The age of onset varies between 5–41 years.76 Both SANDO and MIRAS are caused by mutations in POLG1. Two mutations (W748S and A467T) are specifically associated with MIRAS. MIRAS has been reported as the most prevalent cause of recessive ataxia in Finland due to a founder effect.

Figure 3

Symmetric hyperintense signal changes in both cerebellar hemispheres just dorsal of the dentate nucleus, on axial T2 weighted image in a patient with sensory ataxic neuropathy, dysarthria and ophthalmoparesis (SANDO) syndrome.

POLG1 mutations have been linked to a wide variety of other inherited neurodegenerative phenotypes, such as isolated chronic progressive external ophthalmoplegia, infantile Alpers syndrome, hepatocerebral syndromes, and spinocerebellar ataxia with epilepsy (SCAE) (http://www.tools.niehs.nih.gov/polg/). Schicks et al identified rather frequent POLG1 mutations (11.3%) in a cohort of 80 ataxia patients. They suggest sequencing of POLG in non-SCA and non-FA ataxia patients with progressive external ophthalmoplegia, psychiatric comorbidities, and/or axonal neuropathy.77

POLG1 encodes the mitochondrial DNA (mtDNA) polymerase-γ and is involved in mtDNA replication. The pathogenesis of POLG related diseases involves secondary damage of mitochondrial DNA in the form of tissue specific multiple deletions and/or quantitative depletion.78 This probably leads to energy failure and subsequent neuronal cell death.79 Further studies are needed to clarify the complex pathomechanism of this disorder. Being aware of this molecular defect as a potential cause of the syndrome is crucial as the use of sodium valproate for epilepsy in patients with POLG1 mutations may trigger liver insufficiency.

Marinesco–Sjogren syndrome

The classic features of the rare Marinesco–Sjogren syndrome (MSS) are cerebellar ataxia with cerebellar atrophy, early onset cataract, psychomotor delay, hypotonia, and progressive muscular weakness and atrophy often with elevated creatine kinase values. Additional findings are hypergonadotropic hypogonadism, skeletal abnormalities, short stature, and strabismus.80

MSS is caused by mutations in the SIL1 gene which encodes a protein that interacts with the ATPase domain of GRP78 (HSPA5), a molecular chaperone functioning mainly in the endoplasmic reticulum (ER) and required for the proper folding of proteins.81 Not all patients with classical MSS harbour mutations in SIL1 suggesting genetic heterogeneity.82

Autosomal recessive cerebellar ataxia type 1

Autosomal recessive cerebellar ataxia type 1 (ARCA1) presents as a pure cerebellar ataxia with a relatively late onset (17–46 years) and slow progression. The disorder was identified in a French Canadian population, of which 53 individuals from 26 families were affected. Imaging revealed diffuse pure cerebellar atrophy.83

SYNE1 involved in ARCA1 was the first identified gene responsible for a recessively inherited, pure cerebellar ataxia.83 Currently, seven different truncating disease causing mutations have been identified. These different mutations result in a relatively homogeneous clinical phenotype. SYNE1 is one of the largest genes in the human genome, comprising 147 exons which encode a 8797 amino acid long protein, called Spectrin repeat containing Nuclear Envelope one. The protein contains two N-terminal actin binding regions which consists of tandem paired calponin homology domains, a transmembrane domain, multiple spectrin repeats, and a C-terminal Klarsicht domain. In humans, SYNE1 is predominantly expressed in cerebellum. In the peripheral nervous system, SYNE1 is involved in anchoring specialised myonuclei underneath the neuromuscular junctions. The protein is part of the spectrin family of structural proteins that share a common function of linking the plasma membrane to the actin cytoskeleton. Other proteins of the family are dystrophin/DMD (Duchenne's and Becker muscular dystrophies), β-III-spectrin/SPTBN2 (spinocerebellar ataxia type 5), and puratrophin-1/PLEKHG4 (spinocerebellar ataxia type 31). It is speculated that the ARCA1 associated loss of function mutations in SYNE1, resulting in loss of function of the brain specific larger isoforms, may disrupt cerebellar architecture.83 SYNE1 mutations which do not lead to a complete loss of function are associated with other diseases without cerebellar ataxia. Four different missense mutations in SYNE1 have been associated with Emery–Dreifuss muscular dystrophy.84 Recently, a splice site mutation leading to lack of the C-terminal transmembrane domain KASH has been identified in a family with autosomal recessive arthrogryposis multiplex congenital.85

Autosomal recessive cerebellar ataxia type 2 with coenzyme Q10 deficiency

Autosomal recessive cerebellar ataxia type 2 with coenzyme Q10 deficiency (ARCA2) is a rare ARCA subtype, characterised by childhood onset gait ataxia and cerebellar atrophy with slow progression. Exercise intolerance in childhood and slightly elevated serum lactate were present in three out of the seven patients described by Lagier-Tourenne et al.86 All seven patients did not display coenzyme Q10 (CoQ10) deficiency in muscle, in contrast to the four patients described by Mollet et al.87 Clinically, these latter patients were characterised by developmental delay, elevated serum lactate, seizures, and cerebellar atrophy.

ARCA2 is caused by mutations in the AARF domain containing kinase 3 (ADCK3), also known as chaperone activity of BC1 complex-like (CABC1), which encodes a mitochondrial protein required for coenzyme Q biosynthesis.86 Recently, the first nonsense mutations in CABC1 have been identified in five patients originating from two Dutch families.88 These patients clinically resemble the patients described by Mollet et al.87 So far it has not been proven that treatment with CoQ10 provides any clinical benefit in patients with ARCA2.

Autosomal recessive cerebellar ataxia type 3 caused by mutations in ANO10

Autosomal recessive cerebellar ataxia type 3 caused by mutations in ANO10 (ARCA3) has recently been described by us as a relatively pure recessive cerebellar ataxia, caused by mutations in anoctamin 10 (ANO10). So far, mutations have been identified in eight patients from three different families originating from the Netherlands, Serbia, and France. Clinically ARCA3 is characterised by impaired coordination of limbs and gait with onset between 15–45 years. Brain imaging shows cerebellar atrophy. Additionally, lower motor neuron involvement has been observed in three patients originating from two different families.89

ANO10, also known as TMEM16K (transmembrane 16K), is a member of the human anoctamin (ANO) family, which comprises at least nine other proteins, all exhibiting eight transmembrane domains and a DUF590 domain of unknown function.90 ANO10 may code for a calcium regulated chloride channel and it was suggested that altered calcium signalling in Purkinje cells due to dysfunctional or absent ANO10 may play a role in the pathophysiology of ARCA3. Functional studies are necessary to confirm this.

Putative pathogenesis

Different presumed pathogenesis involved in ARCAs can be distinguished. In this review we only discuss the ones involved in the ARCAs described here.

Six possible disease mechanisms can be distinguished: (1) mitochondrial dysfunction; (2) disorder of DNA repair; (3) protein misfolding and chaperone dysfunction; (4) mislocalisation of synaptic myonuclei; (5) altered function of calcium mediated chloride channels; and (6) altered vesicular trafficking (box 1).

Box 1

Classification based on molecular pathogenesis

  • Mitochondrial dysfunction

    • FA

    • AVED

    • IOSCA

    • SANDO or MIRAS

    • ARCA2

  • DNA repair dysfunction

    • AT

    • ATLD

    • AOA1 or EAOH

    • AOA2

    • SCAN1

  • Protein misfolding and chaperone dysfunction

    • ARSACS

    • MSS

  • Mislocalisation of synaptic myonuclei

    • ARCA1

  • Altered function of calcium mediated chloride channel

    • ARCA3

  • Altered vesicular trafficking

    • Salih ataxia

AOA1, ataxia with oculomotor apraxia type 1; AOA2, ataxia with oculomotor apraxia type 2; ARCA1, autosomal recessive cerebellar ataxia type 1; ARCA2, autosomal recessive cerebellar ataxia type 2 with coenzyme Q10 deficiency; ARCA3, autosomal recessive cerebellar ataxia type 3 caused by mutations in ANO10; ARSACS, autosomal recessive spastic cerebellar ataxia of Charlevoix–Saguenay; AT, ataxia telangiectasia; ATLD, ataxia telangiectasia-like disorder; AVED, ataxia with vitamin E deficiency; FA, Friedreich's ataxia; IOSCA, infantile onset spinocerebellar ataxia; MIRAS, mitochondrial recessive ataxia syndrome; MSS, Marinesco–Sjogren syndrome; SANDO, sensory ataxic neuropathy, dysarthria and ophthalmoparesis; SCAN1, spinocerebellar ataxia with axonal neuropathy.

Increased oxidative stress leading to specific neuronal cell death, caused by either mitochondrial dysfunction or operating through DNA repair dysfunction, may be a common disease mechanism in ARCAs. Inactivating mutations result in the physical absence or the functional loss of the salient proteins. This affects the control of energy output, control of DNA maintenance, and cell cycle control.

Although different pathogenetic mechanisms are involved in the various cerebellar ataxias, it is likely that the products of both dominant and recessive cerebellar ataxia genes are all linked in a single or multiple interconnected functional network,91 comparable to the Usher interactome.92 For example, the protein encoded by SYNE1 is a member of the spectrin family of which the genes PLEKHG4 and SPTBN2, both involved in autosomal dominant spinocerebellar ataxia (SCA), are also members. A recent study by Asai et al demonstrated an interaction between the mutated protein kinase C gamma, involved in SCA14, and the aprataxin protein involved in AOA1.93 As recent studies have shown that both mutated aprataxin and TDP1 lead to mitochondrial dysfunction, this may turn out to be a common pathophysiological mechanism in the different ARCAs.65 94

Future perspectives

At present, molecular testing is still expensive and in many countries the availability of large scale mutation analysis is rather poor. Therefore, focused testing is still recommended. In all patients with a suspected ARCA mutation analysis of the FA gene should be considered. On the other hand, Salih ataxia, SCAN1, and IOSCA (table 2) are examples of ARCA that so far have only been described in specific (isolated) populations, and therefore are not thought of or searched for routinely.

Simultaneous testing of ARCA genes by next generation sequencing, including exome sequencing, is already available. It is expected that within 1 or 2 years simultaneous analysis of ARCA genes will be cost effective, and given the extensively clinical and genetic heterogeneity and overlap with ARCAs, it will be the initial molecular genetic test of choice. Also, these new techniques will greatly facilitate the identification of novel ARCA genes. New genes will lead to a better diagnosis in more patients, which is helpful for determining prognosis and for genetic counselling. Eventually, the use of this new technique will make classifications based on a clinical phenotype, presently still used as a diagnostic guideline, redundant. The ultimate ambition is to understand the biology of these devastating disorders and to develop targeted treatments.

References

Footnotes

  • Funding This work was supported by a grant from the Netherlands Organization of Health Research and Development (Zon MW RM000085) to NK.

  • Competing interests None.

  • Patient consent The figures do no show any recognisable features that can be traced back to the patients.

  • Provenance and peer review Not commissioned; externally peer reviewed.