Spinocerebellar Ataxia 31
A number sign (#) is used with this entry because spinocerebellar ataxia-31 (SCA31) is caused by a 2.5- to 3.8-kb insertion containing pentanucleotide repeats including (TGGAA)n within an intron of the BEAN gene (612051) on chromosome 16q21-q22.
For a general discussion of autosomal dominant spinocerebellar ataxia, see SCA1 (164400).
See also SCA4 with sensory axonal neuropathy (600223), which also maps to chromosome 16q, but has a different phenotype.
Clinical FeaturesNagaoka et al. (2000) reported 6 Japanese families with a pure cerebellar syndrome, referred to as autosomal dominant cerebellar ataxia type III (ADCA III) in the clinical classification by Harding (1982). The average age at onset was 55.9 years (range 45 to 72 years). Affected individuals had gait ataxia, cerebellar dysarthria, limb ataxia, decreased muscle tone, and horizontal gaze nystagmus. Sensation was normal in all but 1 patient who was 85 years old. There were no signs of pyramidal tract involvement. MRI showed cerebellar atrophy. Nagaoka et al. (2000) noted that the phenotype in their families differed from that of SCA4 with sensory neuropathy in the families reported by Gardner et al. (1994).
Ishikawa et al. (2005) found that 6 of 14 (42.9%) families with pure cerebellar ataxia similar to that described by Nagaoka et al. (2000) developed audiologic evidence of mild to moderate bilateral sensorineural hearing loss, which may or may not have been related to the cerebellar ataxia.
Owada et al. (2005) reported a 5-generation Japanese kindred in which multiple members had autosomal dominant pure cerebellar ataxia. The average age at onset was 52.1 years, although 2 patients had onset before age 20 years. Clinical features were consistent with a pure cerebellar syndrome, including truncal ataxia, limb ataxia, dysarthria, and reduced muscle tone. Tendon reflexes were normal but reduced at the ankles in 29% of patients. Gaze nystagmus was not obvious, and there were no signs of peripheral nerve involvement. Six of 7 patients examined also had hearing impairment of cochlear origin. Neuropathologic examination of 1 patient showed moderate cerebellar atrophy with Purkinje cell degeneration, abnormal dendrites, and somatic sprouts of Purkinje cells. Some degenerating Purkinje cells were surrounded by synaptophysin (SYP; 313475)-immunoreactive amorphous material containing calbindin (CALB1; 114050)- and ubiquitin (UBB; 191339)-positive granules.
Ouyang et al. (2006) reported 20 patients from 9 Japanese families with 16q-linked ataxia, as well as 2 patients with sporadic disease. The most common clinical features included late age at onset (mean 62 years), gait ataxia, dysarthria, nystagmus, and cerebellar atrophy. Less common features included hearing loss, tremor, brisk tendon reflexes, and decreased vibration sense. Affected members of 1 family showed spasticity without extensor plantar responses. All patients had a -16C-T variation in the PLEKHG4 gene (609526.0001), but this was later excluded as the molecular cause of the disorder (Amino et al., 2007 and Sato et al., 2009).
Hirano et al. (2009) reported 45 heterozygous patients and 4 homozygous Japanese patients with SCA31, as defined by presence of the -16C-T linkage marker in the PLEKHG4 gene. One of the homozygous patients was born of consanguineous parents, both of whom had a history of ataxia. Three of the homozygous patients were sibs. Their mother reported a history of ataxia, but their father did not. Although the age of onset of homozygous patients was similar to heterozygous patients overall (59.3 years), the disease onset tended to be earlier for homozygous patients within each family. Common clinical features overall included late-onset pure cerebellar ataxia with brisk reflexes. Hearing loss was variable. Among a larger cohort of 116 Japanese families with ataxia, the overall prevalence of SCA31 was 27% in South Kyushu and Okinawa, which was higher than any other form of ataxia. The prevalence of SCA31 was highest in the Miyazaki (65%) and Kagoshima (24.6%) districts.
MappingNagaoka et al. (2000) mapped a locus responsible for what they characterized as a form of pure autosomal dominant cerebellar ataxia to chromosome 16q where the SCA4 locus had been mapped by Gardner et al. (1994).
Li et al. (2003) identified new polymorphic markers in the critical region of mapping. By typing these markers on 8 Japanese families with ADCA type III, including those reported by Nagaoka et al. (2000), they found that a common 'founder' haplotype was seen in a restricted area of 16q22.1.
By linkage analysis of 4 families from southern Japan with pure cerebellar ataxia, Hirano et al. (2004) refined the candidate disease locus to a 1.25-Mb interval between markers 17msm and CTTT01 on chromosome 16q22.1 (maximum 2-point lod score of 6.01 at D16S3141). Haplotype analysis suggested a founder effect for all 4 families.
Molecular GeneticsExclusion of Mutations in PLEKHG4 As Causative
In affected patients from 52 unrelated Japanese families with a pure form of cerebellar ataxia mapping to chromosome 16q, Ishikawa et al. (2005) identified a heterozygous variation in the PLEKHG4 gene (-16C-T; 609526.0001). One of the families was reported by Owada et al. (2005). Ohata et al. (2006) identified the -16C-T transition in the PLEKHG4 gene in 63 patients from 51 Japanese families with cerebellar ataxia. All families were from Nagano prefecture, which is relatively isolated by steep mountains, and 49 of the families shared a common haplotype. The phenotype was homogeneous, with adult onset and pure cerebellar ataxia without additional symptoms. However, 1 affected patient did not carry the -16C-T transition, even though her affected family members did have the change. In addition, this patient shared only a narrow part of the common haplotype, including a region centromeric to the -16C-T transition, suggesting that a true pathogenic mutation may be present in a different gene.
Amino et al. (2007) identified a Japanese family with autosomal dominant cerebellar ataxia linked to chromosome 16q who did not have the -16C-T transition, but carried the common haplotype centromeric to the PLEKHG4 gene identified in the patient reported by Ohata et al. (2006) who also did not carry the common -16C-T change. The findings redefined the disease locus to a 900-kb region between a polymorphism, which they called SNP04 that is centromeric to the PLEKHG4 gene, and the -16C-T transition. Sato et al. (2009) also stated that the -16C-T transition is in strong linkage disequilibrium with this disorder, but is not causative.
2.5- to 3.8-kb Insertion Containing Pentanucleotide Repeats Including (TGAAA)n on Chromosome 16q21-q22
By Southern blot analysis of a 900-kb critical region on chromosome 16q21-q22, followed by sequencing analysis, Sato et al. (2009) identified 2.5- to 3.8-kb insertions (612051.0001) in all 160 affected individuals from 98 families with SCA31, including 1 family reported by Amino et al. (2007). PCR amplification followed by sequencing showed that the insertion consisted of a preceding TCAC sequence followed by 3 pentanucleotide repeat components (TGGAA)n, (TAGAA)n, and (TAAAA)n in all patients tested. In a homozygous patient from whom the 900-kb critical region was derived, the authors found a (TGGAA)n sequence of greater than 110 repeats and a (TAAAATAGAA)n sequence of greater 112 repeats, both of which were too long to be read through. The (TGGAA)n and (TAAAATAGAA)n sequences were separated by a bridging sequence and (TAGAA)46. The insertions were located in introns of the BEAN (612051) and TK2 (188250) genes, which are on opposite strands and transcribed in opposite directions. These insertions were not identified in 99.77% of 800 Japanese and 60 American chromosomes, or in individuals with SCA4 (600223). However, 2 (0.23%) of 860 control chromosomes did carry similar smaller 1.5- or 2.0-kb insertions without (TGGAA)n sequences. Sato et al. (2009) concluded that the insertions in SCA31 patients exerted their toxicity either because of their length or the (TGGAA)n sequence, or because of both. The length of the SCA31 insertion was inversely correlated with the age at disease onset. Further analysis showed that the insertion site was identical for all insertions and was located at an Alu sequence. A single-nucleotide change in an intron of the TK2 gene segregated with SCA31 but was not considered to be pathogenic. The repeat insertions did not appear to cause splicing abnormalities or alterations in the expression levels of BEAN, TK2, or other nearby genes. Sato et al. (2009) demonstrated that the insertion transcribed in the direction of BEAN formed RNA foci in approximately 30 to 50% of Purkinje cell nuclei from SCA31 patients but not in cells from controls. RNA foci were not observed for antisense probes corresponding to TK2 transcripts in SCA31 or control brains. The splicing factors SFRS1 (600812) and SFRS9 (601943) were found to directly bind to (UGGAA)n, the transcribed sequence of (TGGAA)n, in vitro. In silico analysis showed that (TGGAA)n was abundant in centromeres of several human chromosomes, suggesting a role in heterochromatin or chromosomal structure.
Population GeneticsIn a nationwide survey of Japanese patients, Hirayama et al. (1994) estimated the prevalence of all forms of spinocerebellar degeneration to be 4.53 per 100,000. Of these, 7.5% were estimated to have pure cerebellar ataxia, with onset after young adulthood. Cerebellar atrophy was appreciable on brain imaging.
Ouyang et al. (2006) estimated that 16q-linked ADCA is the third most common form in Japan, after MJD (109150) and SCA6 (183086).
Among 113 Japanese families from the island of Hokkaido with autosomal dominant SCA, Basri et al. (2007) found that SCA6 was the most common form of the disorder, identified in 35 (31%) families. Thirty (27%) families had SCA3, 11 (10%) had SCA1, and 10 (9%) had 16q22-linked SCA. The specific disorder could not be identified in 16 (14%) families.
HistoryFamilial forms of pure cerebellar ataxia have been reported (see, e.g., Harding, 1982; Hoffman et al., 1970; Frontali et al., 1992).