Autosomal Recessive Cerebellar Ataxia Type 3 Due to ANO10 Mutations: Delineation and Genotype-Phenotype Correlation Study

Importance  ANO10 mutations have been reported to cause a novel form of autosomal recessive cerebellar ataxia (ARCA). Our objective was to report 9 ataxic patients carrying 8 novel ANO10 mutations to improve the delineation of this form of ARCA and provide genotype-phenotype correlation.

Observations  The ANO10 gene has been sequenced in 186 consecutive patients with ARCA. The detailed phenotype of patients with ANO10 mutations was investigated and compared with the 12 previously reported cases. The mean age at onset was 33 years (range, 17-43 years), and the disease progression was slow. Corticospinal tract signs were frequent, including extensor plantar reflexes and/or diffuse tendon reflexes and/or spasticity. No patient in our series had peripheral neuropathy. Magnetic resonance imaging of the brains of our patients revealed marked cerebellar atrophy. The most frequent mutation, a mononucleotide expansion from a polyA repeat tract (c.132dupA) that causes protein truncation, was never observed in homozygosity. Only 2 truncating mutations were reported in homozygosity, one of which (c.1150-1151del) was associated with juvenile or adolescent onset and mental retardation, whereas we show that the presence of at least 1 missense or in-frame mutation is associated with adult onset and slow progression.

Conclusions and Relevance  An ANO10 mutation is responsible for ARCA that is mainly characterized by cerebellar atrophy and lack of peripheral neuropathy. We therefore suggest naming this entity autosomal recessive cerebellar ataxia type 3 (ARCA3).

Autosomal recessive cerebellar ataxias (ARCAs) correspond to a heterogeneous group of inherited neurodegenerative disorders that affect the cerebellum and/or the peripheral nerves.¹ A new classification of ARCA has been proposed including 3 groups: ARCA with pure sensory neuropathy, ARCA with sensorimotor axonal neuropathy, and ARCA without neuropathy.² Recently, a new form of ARCA due to mutations in the ANO10 gene has been reported.³ To which group of the classification does this new entity belong and whether genotype-phenotype correlation may be proposed have not yet been elucidated. We report 9 unrelated patients carrying 8 novel ANO10 mutations, and we compared these patients with the 12 previously reported cases in order to provide the delineation of the disease and genotype-phenotype correlations.

Between 2010 and 2013, a total of 186 unrelated index patients were consecutively recruited from 4 tertiary centers for inherited neurodegenerative disorders. Inclusion criteria were the combination of (1) sporadic or recessive progressive cerebellar ataxia; (2) age at onset of younger than 50 years; and (3) unknown etiology despite extensive investigations, including molecular analysis for Friedreich ataxia and obvious forms of recessive ataxias. Written informed consent was obtained from all participants. Participants were not financially compensated. The recessive ataxia research project was approved by the Strasbourg ethics committee. A detailed neurological examination, including scales of severity and magnetic resonance imaging of the brain, was performed for all patients.

Genetic analysis of ANO10 was performed for 44 patients by conventional Sanger sequencing. A total of 142 additional patients were directly analyzed by a targeted exon-capture strategy coupled with multiplexing and high-throughput sequencing of 57 genes causing ataxia when mutated, including ANO10. The splice-site mutations were evaluated with the SplicePort program (http://spliceport.cbcb.umd.edu/) and were analyzed by reverse transcription of RNA from patients’ fibroblasts followed by polymerase chain reaction from exon 8 to exon 11 (520–base pair [bp] fragment) and from exon 6 to exon 12 (809-bp fragment). Polymerase chain reaction fragments were visualized on agarose gels and sequenced by use of the Sanger method.

The patients from this and previous studies were divided into 2 groups according to the type of the ANO10 mutation: (1) at least 1 missense or in-frame mutation, or (2) 2 truncating mutations. The 2 groups were compared by use of the t test.

The clinical, radiological, and biological findings of the 9 patients are shown in Table 1 and Figure 1. Of the 186 patients, 9 (5%) had ANO10 mutations. All 9 of these patients carrying ANO10 mutations had obvious cerebellar atrophy detected on magnetic resonance imaging scans of the brain and a lack of peripheral neuropathy. There was a subgroup of 19 patients with recessive progressive cerebellar ataxia and 37 patients with sporadic progressive cerebellar ataxia for whom cerebellar atrophy and lack of peripheral neuropathy were documented. Of these 56 patients, 6 (11%) had ANO10 mutations.

The 9 patients carried 1 previously reported ANO10 mutation and 8 novel ANO10 mutations (Table 1). The c.132dupA mutation was identified in 4 unrelated patients. The mutation is an insertion of an A (adenine nucleotide) following a homopolymer of 9 A’s and can be considered as a mononucleotide expansion of a mononucleotide repeat tract. Indeed, the c.132dupA mutation is a frequent mutation with a heterozygous carrier frequency of 1/184 in different ethnic groups (Exome Variant Server [http://evs.gs.washington.edu/EVS]) and probably arose because of multiple independent homopolymer instability events. The c.132dupA mutation leads to early protein truncation (p.Asp45Argfs*9). Surprisingly, all 4 patients with this mutation were compound heterozygous, and no homozygous c.132dupA patient was identified. The 3 identified missense mutations, p.Phe171Ser, p.Gly229Trp, and p.Phe337Val, are likely pathogenic because they affect amino acid positions that are conserved in the orthologous sequences of almost all metazoan species (Figure 2). The p.Gly229Trp mutation is located at the C-terminal border of the first α-helical transmembrane segment, and the p.Phe337Val mutation is located at the C-terminal border of the third transmembrane segment. Reverse transcriptase–polymerase chain reaction analysis of the 2 donor splice-site mutations (Figure 3) revealed complete alteration of the splice sites. Surprisingly, skipping the mutated exon or using cryptic splice sites resulted in all cases in in-frame transcripts. The deletion mutation (deletion of exon 12), identified in 3 unrelated patients, is also in frame. A phenotype vs genotype comparison, including also previously reported patients (Figure 4), revealed that the age at onset was higher for patients with at least 1 missense or in-frame exon skipping mutation than for patients with 2 truncating mutations (homozygous c.1050_1051del and homozygous p.Tyr203*) (P < .01).

Although electrophysiological signs of lower motor neuron involvement in lower limbs were found in a few patients^3,4 and in case 3, none of the 21 reported patients had polyneuropathy, which classifies ANO10-linked ataxia among the group of cerebellar ataxias without neuropathy.³ Because ARCA1^5,6 and ARCA2⁷ are already representatives of this group, we propose to name this new entity autosomal recessive cerebellar ataxia type 3 (ARCA3) (Table 2).

The mean age at onset is late in our series (including 8 patients with onset at older than 25 years of age), but, overall, age at onset for those with ARCA3 appears highly variable, ranging from 6 to 43 years. Disease progression is slow, with patients remaining ambulatory for up to 25 years after the onset of the disease. However, it may be more severe, as in one of our patients (case 4), who needed a stick for walking after 7 years of disease duration. Lower limb reflexes appeared to be always increased, but extensor plantar reflexes, mental retardation, or cognitive decline were variably present (Table 1).^3,4,8 Two patients exhibited spasticity; therefore, ARCA3 should also be considered in case of spastic ataxia. One patient (case 9) reported paroxysmal episodes of ataxia and vertigo, which could be explained by the abnormal function of mutated anoctamin-10, which is a putative calcium-activated chloride channel.⁹ No extraneurological signs were found in our series.

The α₁-fetoprotein serum level, which is known to be a reliable biomarker for ataxia with oculomotor apraxia type 2¹⁰ and ataxia telangiectasia, was mildly increased in 2 cases. Whether α₁-fetoprotein could be a biomarker for ARCA3 remains to be elucidated by further large studies.

Overall, we report an ARCA3 phenotype in accordance with previous reports and, more precisely, define a core presentation that should be kept in mind when facing a patient with recessive or sporadic ataxia. Autosomal recessive cerebellar ataxia type 3 appears to be a common cause of recessive ataxias. While the frequency is accounted for, in part, by the high frequency of the major mutation, c.132dupA, which belongs to the group of unstable sequences, this mutation is underrepresented in patients with ARCA3 with respect to the other mutations that are at least 10-fold less frequent (http://evs.gs.washington.edu/EVS). The fact that all patients with this mutation were compound heterozygotes and that no homozygous c.132dupA patient was identified suggests that c.132dupA homozygotes are either not viable or have a more severe phenotype, similar to the p.Arg141His mutation of phosphomannomutase 2 in the congenital disorder of glycosylation type Ia.¹¹

In accordance with this correlation is the fact that all 6 previously reported patients with the homozygous truncating mutation c.1150-1151del^3,4 have a significantly earlier age at onset (juvenile or adolescent onset associated with mental retardation) compared with the patients with at least 1 missense or in-frame mutation (Figure 4) who have adult onset, slow progression, and persistence of ambulation over a long disease duration. However, a recently reported patient,⁸ homozygous for the truncating nonsense mutation p.Tyr203*, does not fit with this correlation because he had onset at 42 to 46 years of age. It is nevertheless interesting to note that the only 2 reported homozygous truncating mutations (p.Tyr203* and c.1150-1151del) are both located in exon 6, which is in frame and differentially spliced in vivo (absent from isoform 2 of ANO10 [http://www.uniprot.org/uniprot/]).

It is possible that these 2 mutations variably affect differential splicing (or cryptic splice-site activation, as in case 1 of this study) of exon 6 and result in production of in-frame transcripts, because the 2 mutations are located close to the acceptor and donor splice sites, respectively, of exon 6. Such a mechanism would explain the differential consequence of homozygous truncating mutations p.Tyr203* and c.1150-1151del and the absence of homozygous truncating mutations c.132dupA, despite the high frequency of heterozygote carriers (1/184). RNA studies of the p.Tyr203* and c.1150-1151del homozygous patients are warranted to solve this issue.

In conclusion, ARCA3, which has a major mutation (c.132dupA) that is a mononucleotide expansion from a mononucleotide repeat tract, is not a rare form of recessive cerebellar ataxia. It should be considered in cases of progressive cerebellar ataxia with cerebellar atrophy but without peripheral neuropathy, even in sporadic cases and whatever the age at onset.

Corresponding Author: Michel Koenig, MD, PhD, Laboratoire de Génétique des Maladies Rares, Institut National de la Santé et de la Recherche Médicale UMR_S 827, Institut Universitaire de Recherche Clinique, 641 Avenue du Doyen Gaston Giraud, 34093 Montpellier CEDEX 5, France (michel.koenig@inserm.fr).

Published Online: August 4, 2014. doi:10.1001/jamaneurol.2014.193.

Author Contributions: Drs Renaud and Koenig had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Anheim and Renaud contributed equally to this article as first authors. Drs Tranchant and Koenig contributed equally to this article as last authors.

Study concept and design: Renaud, Anheim, Redin, Muller, Scheffer, Tranchant, Koenig.

Acquisition, analysis, or interpretation of data: Renaud, Anheim, Kamsteeg, Mallaret, Mochel, Vermeer, Drouot, Pouget, Salort-Campana, Kremer, Verschuuren-Bemelmans, Muller, Durr, Tranchant, Koenig.

Drafting of the manuscript: Renaud, Anheim, Drouot, Muller, Durr, Tranchant, Koenig.

Critical revision of the manuscript for important intellectual content: Anheim, Kamsteeg, Mallaret, Mochel, Vermeer, Pouget, Redin, Salort-Campana, Kremer, Verschuuren-Bemelmans, Scheffer, Tranchant, Koenig.

Statistical analysis: Renaud, Muller.

Obtained funding: Muller, Scheffer, Koenig.

Administrative, technical, or material support: Mochel, Vermeer, Drouot, Pouget, Redin, Durr, Koenig.

Study supervision: Anheim, Tranchant, Koenig.

Conflict of Interest Disclosures: None reported.

Funding/Support: Dr Renaud was supported by a fellowship from the Journées de Neurologie de Langue Française. This study was supported by funds from the Institut National de la Santé et de la Recherche Médicale, the Centre National de la Recherche Scientifique, the Agence Nationale pour la Recherche-Maladies Rares and Maladies Neurologiques et Psychiatriques (grant ANR-09-MNPS-001-01 to Dr Koenig), and the ANR/E-rare JTC 2011 “Euro-SCAR” (grant 2011-RARE-004-01 to Dr Koenig).

Role of the Sponsor: The funding agencies had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.