Identification of the SCA21 disease gene: remaining challenges and promising opportunities

This scientific commentary refers to ‘TMEM240 mutations cause spinocerebellar ataxia 21 with mental retardation and severe cognitive impairment', by Delplanque et al. (doi: 10.1093/brain/awu202). For nearly one and a half centuries, physicians have observed that disorders of coordination often run in families, and have documented how variable these disorders can be in terms of their natural history and symptomatology (Friedreich, 1863; Menzel, 1891). Transmission of such afflictions from parent to child led to the realisation that they show an autosomal dominant pattern of inheritance. The disorders were named ataxias, from the Greek ‘taxis’ meaning order, and were associated with dysfunction of the cerebellum and its connections. For a century, they were classified mainly based upon their neuropathology, and were collectively known as olivopontocerebellar atrophies (Konigsmark and Weiner, 1970). Anita Harding refined this classification in the 1980s to account for the mode of inheritance and clinical features, and created a very useful classification scheme consisting of three types of autosomal dominant cerebellar ataxia (ADCA). According to the Harding ADCA classification system, patients with type I ADCA present with ataxia in combination with extracerebellar signs, type II ADCA patients display ataxia, extracerebellar signs, and also retinal degeneration, and type III ADCA patients exhibit a pure cerebellar ataxia without any extracerebellar signs (Harding, 1982). With the advent of molecular genetics, the ADCAs were further defined based upon their genetic linkage location and became known as the spinocerebellar ataxias, or ‘SCAs’. Today, there are more than 35 known SCA genetic loci, within which 22 causative genes have been identified (Matilla-Dueñas et al., 2012). The SCAs that belong to Harding’s type I ADCA category are a very heterogeneous group of diseases, with highly pleiotropic phenotypes and complex pyramidal and extrapyramidal symptoms, including cognitive decline, motor neuron disease, neuropathy, and depression. In this month’s issue of Brain, Delplanque and colleagues add another piece to the SCA puzzle by refining the clinical description and defining the genetic basis of SCA21 (Delplanque et al., 2014). SCA21 was initially identified by the very same group more than a decade ago (Vuillaume et al., 2002), and its genetic locus was erroneously mapped to chromosome 7. The original report described a large French family in which patients exhibited gradually progressive gait and limb ataxia in combination with akinesia, rigidity, tremor, and hyporeflexia. Several SCA21 patients also presented with mild to severe cognitive impairment. Re-analysis of this same family has now yielded evidence of additional clinical features including delayed acquisition of cognitive and motor skills, mild to severe mental retardation, variable onset of clumsiness, and examples of very slow disease progression (Delplanque et al., 2014). Although cognitive impairment has been observed in a few other SCAs, this phenotype is remarkably consistent in the large SCA21 pedigree described in the current work. Since the 2002 report, dramatic advances have occurred in molecular and genomics technology, permitting refinement of the genetic linkage analysis through the use of genome-wide single nucleotide polymorphism (SNP) microarrays. These new results restricted the chromosomal region of interest for SCA21 to a region on the short arm of chromosome 1. Extensive genomic sequencing of several family members further narrowed this region down to an interval containing the TMEM240 gene, and yielded a non-synonymous SNP in the coding region of the TMEM240 gene that co-segregated perfectly with the SCA21 disease phenotype. Importantly, this variant was absent from almost 1000 control subjects, leading the authors to screen 368 other French families with unidentified SCA phenotypes. Directed sequencing of the TMEM240 gene in these SCA families revealed seven unrelated families with linked mutations in the very same gene. Two of these families possessed the P170L mutation detected in the original large kindred, while the other five families displayed distinct putative mutations in the TMEM240 gene. Interestingly, five of the six mutations reside in the carboxyl terminal of the TMEM240 amino acid coding region. One of these mutations, T80M, is located between two transmembrane alpha helices, and occurs in a family whose phenotype does not include cognitive impairment. Despite the evidence for TMEM240 as the causal gene for SCA21, the mechanism by which the various mutations cause disease remains unclear. Of the six mutations identified, five are missense and one introduces a premature stop codon. The presence of a nonsense mutation among the six would seem to suggest that haploinsufficiency accounts for SCA21 disease pathogenesis; however, this nonsense mutation is located in the last exon and thus is unlikely to be targeted for nonsense-mediated decay. It is possible that a truncated protein product is generated; hence, it would be premature to conclude that the mutations cause a loss of function in the TMEM240 protein. On the other hand, too little is known about the TMEM240 protein and its normal function to support a gain-of-function mechanism. Indeed, the extent of our collective knowledge of this gene derives from whole brain transcriptome profiling indicating that it is widely expressed throughout the brain (Hawrylycz et al., 2012). Importantly, Delplanque et al. analysed the degree of conservation of the five mutated amino acid residues and found that all of these amino acids are fully conserved down to zebrafish. Consequently, future studies should focus on the normal function of the TMEM240 protein and how the different disease-causing mutations alter function. Indeed, when one considers the SCAs that are not caused by repeat expansion mutations, one finds very little commonality between the different disease gene products. For example, one group of SCA disease genes consists of regulatory factors and enzymes, while another group encodes proteins that localize to the plasma membrane. Among the plasma membrane proteins implicated in SCAs, mutations in the KCNC3 potassium channel and in the inositol triphosphate receptor ITPR1 cause SCA13 and SCA15, respectively. SCA5 is caused by mutations in β-III spectrin, which normally plays a role in stabilizing EAAT4, a glutamate transporter in Purkinje cells (Perkins et al., 2010). As the TMEM240 protein is a membrane-spanning protein, it too could be involved in modulating ion channel function at the neuronal cell membrane, akin to the role of β-III spectrin. Certainly, future research efforts should consider how the distinct plasma membrane-spanning proteins implicated in SCAs could be functionally linked: this underscores why the discovery of TMEM240 as the cause of SCA21 is so exciting and potentially important in advancing our understanding of ataxia disease biology. Of course, an understanding of TMEM240 protein function will be crucial not only for generating models of SCA21 disease pathogenesis, but also for developing strategies for treating SCA21 and related disorders. Our research on the spinocerebellar ataxias is supported by the National Institutes of Health (R01 EY014061) and by a Pioneer Award from the National Ataxia Foundation.