Clinical and molecular features of spinocerebellar ataxia type 6

The autosomal-dominant cerebellar ataxias (ADCAs) are highly represented among the increasing number of neurodegenerative diseases caused by unstable mutations.¹ ADCA designates a clinically, neuropathologically, and genetically heterogeneous group of inherited neurodegenerative disorders divided into three types on clinical grounds.² In ADCA type I, by far the most frequent, cerebellar ataxia is variably associated with dysarthria, ophthalmoplegia, pyramidal and extrapyramidal signs, deep sensory loss, amyotrophy, and dementia. The association of a progressive macular degeneration with cerebellar ataxia is a particular feature of ADCA type II. ADCA type III is a"pure" cerebellar syndrome, usually with late onset after age 50.

Recent advances in genetics have given new insight into this classification. Unstable mutations, causing expansion of polyglutamine tracts, have been identified in three forms of ADCA type I, spinocerebellar ataxia (SCA) 1,³ SCA2,^4-6 and SCA3/Machado-Joseph disease (MJD),⁷ and in ADCA type II(SCA7).⁸ At least two other loci are implicated in ADCA type I: SCA4 on chromosome 16⁹ and another, not yet localized, for which no evidence of a polyglutamine expansion could be found.¹⁰ Ranum et al.¹¹ mapped the SCA5 gene to the centromeric region of chromosome 11 in a family with an ADCA type III phenotype, and genetic heterogeneity of this phenotype is already suspected.¹² Recently, a small CAG expansion in the gene of the alpha-1_A subunit of the voltage-dependent calcium channel was found in eight kindreds with mild but progressive cerebellar ataxia and neuropathologic lesions restricted to the cerebellum and inferior olives.¹³ Other mutations in this gene are responsible for episodic ataxia (EA) type 2 and familial hemiplegic migraine.¹⁴ In the present study, we analyzed a large series of patients with ADCA to determine the frequency of the SCA6 subtype and to analyze the molecular and clinical profile of the patients.

Methods. Patients and families. Ninety-one index patients with ADCA, in whom CAG expansions in the SCA1, SCA2, SCA3/MJD, and DRPLA genes were previously excluded, were screened for the SCA6 mutation. Sixteen families had ADCA type III, defined as the presence of pure progressive cerebellar ataxia without signs of pyramidal, extrapyramidal, or ophthalmologic involvement after 10 years of evolution, and three families had EA. We excluded families who presented retinal degeneration. Eighty-one percent of the families were French. In addition, we analyzed 146 patients with clinical features of isolated cerebellar ataxia or olivopontocerebellar atrophy (OPCA) without family histories.

Eye movements were recorded in two SCA6 patients by direct-current electrooculography. Horizontal visually guided saccades and smooth pursuit were studied as described.¹⁵

Genotyping. Genotypes at the SCA6 locus were determined by polymerase chain reaction (PCR) amplification with 200 ng of genomic DNA, 10 pmol of each primer,¹³ 200 µM dNTPs, 2% formamide, 5% dimethyl sulfoxide, 1 mM MgCl₂ in 1× PCR buffer (50 mM KCl, 10 mM Tris-HCl, pH 8.3, 0.01% gelatin), in a final volume of 25 µL. One unit of Taq DNA polymerase was added during the first denaturation step at 94 °C for 10 minutes. Samples then underwent 35 cycles of denaturation at 94 °C for 1 minute, annealing at 65 °C for 1 minute, and extension at 72 °C for 1 minute. The last extension step was lengthened to 2 minutes. Aliquots of the PCR products were then diluted 1:1 in formamide loading buffer, run on a 6% acrylamide/7 M urea gel, and blotted onto nylon membrane. Membranes were hybridized overnight in Amasino buffer¹⁶ at 42 °C with one of the primers, 5′-³²P labeled, and then washed 15 minutes in 2× SSC, 0.1% SDS at room temperature and exposed to x-ray film for autoradiography.

The distribution of the normal allele was determined in unrelated healthy subjects and patients with identified mutations in other spinocerebellar ataxia loci (n = 50).

Microsatellite D19S1150 was amplified by PCR as described¹⁴ to analyze for linkage disequilibrium.

Statistical analysis. Mean values are given with SDs. Comparisons of means were performed with the non-parametric Mann-Whitney U and Kruskall-Wallis tests. The correlation between age at onset and CAG repeat length was determined by linear regression.

Results. Frequency and characteristics of the mutation. An expanded CAG repeat in exon 47 of the alpha-1_A subunit of the voltage-dependent calcium channel gene was found in 15 patients and 3 at-risk individuals from 4 of 91 ADCA families, three of which were classified as type I and one as type III. SCA6, therefore, accounted for only 2% of this series, whereas SCA1 and SCA2 represented 15% each (n = 32) and SCA3/MJD, 33% (n = 73). Two SCA6 kindreds were French (AAD-143 and SAL-329), one was German (AAD-113), and one Welsh (SAL-399-88). SCA1, SCA2, SCA3/MJD, and SCA6 accounted for 15% (n = 22), 10% (n = 14), 32% (n = 47), and 1% (n = 2), respectively, of the 146 French families. The mutation was not observed in isolated patients with cerebellar ataxia, OPCA, or EA.

There were 22 (n = 5), 24 (n = 3), and 28 (n = 10) CAG repeats in 15 patients and 3 at-risk carriers from the four SCA6 families. In 588 normal chromosomes, including 100 from control subjects and 488 from non-SCA6 ataxic patients, the number of repeats varied from 4 to 15 (mean 11.4 ± 2.0 CAG repeats) (figure 1). The SCA6 locus was highly polymorphic (10 normal alleles) because 76% of the subjects tested were heterozygous at this locus. We did not observe instability of the expanded allele during transmission, and the expansion was of the same size in all patients and at risk carriers in the same family(figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Distribution of the CAG repeats on 588 normal and 18 pathologic chromosomes. Black bars indicate pathologic chromosomes from SCA6 patients (n = 18); white bars indicate normal alleles from control subjects (n = 100); and hatched bars indicate normal chromosomes from non-SCA6 ataxic patients (n = 488). The distribution of SCA6 alleles was similar in control subjects and non-SCA6 ataxic patients (p > 0.05).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. PCR-amplified CAG repeats at the SCA6 locus in three families. The local identification number is given for each subject. The number of CAG repeats indicated beside each band remains stable in the transmissions analyzed and is identical in all family members.

The same frequent (29%) allele of the intragenic marker D19S1150 was shared by the two French families. A different allele segregated in the German kindred and was also present in the only available patient from the Welsh family.

Clinical and neuroradiologic observations of 12 SCA6 patients. The mean age at onset was 45 ± 14 years (range, 24 to 67) and at examination was 55 ± 16 years (range, 24 to 75). The initial symptom was gait ataxia in all but three patients, who first experienced diplopia, dysarthria, or leg pain. After a mean disease duration of 11 ± 8 years, all patients, except four, needed help walking and two were confined to wheelchair 16 and 24 years after onset. There was no cerebellar dysarthria in five patients, the reflexes in the lower limbs were decreased in three, and four had difficulty swallowing. There was no pyramidal syndrome or limited eye gaze, except in the oldest examined woman, aged 75, who had limited upward eye movements. Eye movements were recorded in two patients from different families. Saccade velocity was normal in both. In patient SAL-329-27 (28 CAG repeats) with a 12-year disease duration, smooth pursuit was saccadic, slow eye movement was impossible, and gaze-evoked nystagmus was present. Patient AAD-113-13, with 22 CAG repeats, showed hypometric saccades after a disease duration of 3 years. Eye movements in both patients were similar to those observed in SCA3 patients: gaze-evoked nystagmus, saccadic pursuit, hypometric saccades, and normal saccade velocity (S. Rivaud, unpublished data). Brain MRI of three patients (AAD-113-13, AAD-113-16, and SAL-329-27) showed extensive atrophy of the cerebellum, including the vermis and hemispheres, whereas brainstem and cerebral cortex were spared (data not shown). EMG was normal in patient SAL-329-27 after a disease duration of 12 years.

Clinical correlations with the expanded CAG repeat. CAG repeat length was negatively correlated with age at onset (r = -0.87, p< 0.0005). The presence of horizontal nystagmus was significantly associated with larger CAG repeat size (p < 0.01), whereas the presence of dysarthria significantly correlated with longer disease duration(p < 0.05).

In 10 parent-child pairs, onset occurred 8.6 ± 9.7 years earlier(range, -20 to +10) in offspring than in their parents (p < 0.05), suggesting anticipation. The mean ages at onset, however, were similar among generations. The calculated anticipation might, therefore, result from an observation bias in the parent-offspring pairs, as already suggested.¹⁷

Discussion. This study shows that the SCA6 mutation is rare in France (1%) and in patients from various western Europe countries (2%), in contrast to others (10 to 30%).^18,19 The largest expansions in this series (28 repeats) and in a recent report (30 repeats)¹⁸ are smaller than the smallest expansions observed in SCA2 (34 repeats),²⁰ Huntington's disease(36 repeats),²¹ SCA7 (38 repeats),⁸ and SCA1 (39 repeats).²² This may be related to the fact that, unlike the other SCA expansions, the SCA6 mutation is stable during transmission; the expansions were the same size in parent and offspring and among members of the same family. So far, only one variation of the number of CAG repeats has been described.¹⁸ As for the other SCA mutations, however, the age at onset was strongly correlated with the size of the CAG repeat (r = -0.87), demonstrating its direct involvement in the disease.

Cosegregation of at least two different alleles of the intragenic D19S1150 marker with the expansion argues against the hypothesis of a single founder. In addition, given the stability of the expansions, the presence of the same frequent D19S1150 allele in the two French families with different numbers of repeats (n = 24 and n = 28), it is unlikely that mutations were inherited from a common ancestor.

The overall clinical picture of SCA6 resembles that observed in other ADCA. The mean age at onset (45 ± 14 years) is not different from that of 128 SCA3 patients (38 ± 12) but is significantly older than that of 80 SCA1 patients (34 ± 10, p < 0.05) and 119 SCA2 patients (35 ± 14, p < 0.05) (A. Dürr, unpublished results). Age at onset, therefore, does not distinguish SCA6 from other ADCA families, but its intrafamilial variability is lower in SCA6. The functional handicap was severe in 7 of 12 patients and therefore seems less benign than initially reported.¹³ In the family with the smallest CAG repeat (AAD-113), onset occurred after age 50 in all patients, which is compatible with type III ADCA. However, in the family with the largest CAG repeat (SAL-329), age at onset ranged from 24 to 45 years, and additional neurologic signs, reminiscent of ADCA I, were frequently found, as in one of the family in which the SCA6 mutation was initially identified.²³ Neuropathologic features of SCA6 (i.e., extensive loss of Purkinje cells with modest neuronal loss in the inferior olives and a normal pons²³) were clearly different from SCA3 patients where the inferior olives and Purkinje cells are preserved¹⁷ or in SCA2 patients where pontine atrophy is a major feature.²⁴ In SCA1, there is also major cerebellar atrophy, as in SCA6, but the pons and midbrain are also affected.^17,25 These differences are also observed by MRI, which shows degeneration in the cerebellum in SCA6, whereas other structures, particularly in the brainstem, are always affected in SCA1, 2, and 3.

In summary, SCA6, a rare form of ADCA in our series, is associated with a phenotype sharing features with both ADCA type I and type III. However, there is no pyramidal syndrome as in SCA1 or early gaze impairment as in SCA2, placing SCA6 at the pure end of the clinical spectrum of ADCA type I. This is consistent with a neurodegeneration limited to the cerebellum and inferior olives, which distinguishes it from SCA1, 2, and 3, where degeneration is more widespread. The clinical diagnosis of individual patients is ambigous, however, and molecular analysis of the SCA6 locus should be considered in patients presenting with ADCA I or III.

Acknowledgments

We are grateful to the families for participating in this study, to Profs. O. Lyon-Caen and J.-L. Mandel for referring some of the probands, and to Drs. B. Gaymard and I. Lagroua for clinical evaluation. We also thank Drs. M. Ruberg, A.S. Lebre, J. Bou, A. Camuzat, C. Penet, and Y. Pothin for their help.