A number sign (#) is used with this entry because of evidence that spinocerebellar ataxia-4 (SCA4) is caused by a heterozygous trinucleotide repeat expansion (GGCn) in the ZFHX3 gene (104155) on chromosome 16q22. The normal length of the trinucleotide repeat is 14 to 26 units, with the most common length being 21 repeat units, including interruptions. Disease-causing expanded repeats are greater than 40 (range from 42 to 74) units of pure GGC repeats without interruptions.

Spinocerebellar ataxia-4 (SCA4) is an autosomal dominant neurologic disorder characterized by the onset of balance disturbances and gait and limb ataxia usually in the fourth decade, although earlier onset in the teens or twenties has been reported. There is evidence of genetic anticipation within families. The disorder is slowly progressive, and most patients eventually become wheelchair-bound. Additional features include hypometric or slow saccades, sensory or sensorimotor axonal peripheral neuropathy, dysarthria, and autonomic dysfunction, including orthostatic hypotension and problems with bowel or bladder control. More severely affected individuals have dysphagia and significant unintended weight loss, which may contribute to premature death. Brain imaging shows cerebellar atrophy (Wallenius et al., 2024).

For a discussion of autosomal dominant spinocerebellar ataxia, see SCA1 (164400).

Gardner et al. (1994) described a large Utah kindred with a distinct form of autosomal dominant spinocerebellar ataxia. Flanigan et al. (1996) presented clinical and electrophysiologic data of the family reported by Gardner et al. (1994). The phenotype consisted of ataxia with the invariant presence of a prominent axonal sensory neuropathy. Disease onset was typically in the fourth or fifth decade, but age at onset ranged from 19 to 59 years, with a median age at onset of 39 years. The earliest symptom was usually a gait disturbance, followed by difficulty with fine motor tasks and often by dysarthria. Ataxia progressed over decades, often leading to wheelchair dependence. At presentation, most patients did not complain of symptoms of neuropathy, although evidence of a length-dependent neuropathy could invariably be demonstrated on examination: all had vibratory and joint position sense loss, and 95% had at least a minimal pinprick sensation loss. All patients had absent ankle-jerk reflexes; knee-jerk reflexes were absent in 85% and complete areflexia was seen in 25%. Wallenius et al. (2024) noted that the family from Utah reported by Gardner et al. (1994) and Flanigan et al. (1996) originated from Sweden, suggesting that they may have had SCA4 due to ZFHX3 repeat expansions as found in other SCA4 families, all Swedish, studied by them.

Hellenbroich et al. (2003) reported a German family in which more than 14 individuals spanning 5 generations were affected with autosomal dominant cerebellar ataxia. The mean age at onset was 38.3 years (range 20 to 61), and there was some suggestion of genetic anticipation. All patients had cerebellar ataxia with limb dysmetria, dysarthria, and cerebellar atrophy, as well as sensory neuropathy with hypo- or areflexia, decreased sensation, and absent sural sensory nerve action potentials.

Wallenius et al. (2024) reported 38 individuals in 5 multigenerational families from the region of Skane in southern Sweden with autosomal dominant spinocerebellar ataxia and autonomic neuropathy. Fifteen patients were examined by the authors. Two of the families had previously been reported by Wictorin et al. (2014), including 1 who had also been reported by Moller et al. (1978). The patients in the study of Wallenius et al. (2024) were all adults who developed gait and balance disturbances at a mean age of 37.6 years (range 15 to 60). The disorder was slowly but relentlessly progressive, and all had both gait and limb ataxia. Most patients lost the ability to stand without support and became wheelchair-bound, as well as losing handwriting ability. Affected individuals also had slow or hypometric ocular saccades, dysarthria, and distal sensory impairment with hypo- or areflexia. Electrophysiologic studies showed a sensory or sensorimotor axonal neuropathy. More variable features included torticollis, involuntary facial twitching, head tremor, involuntary leg jerks, restless legs, leg cramps, intention tremor, fasciculations, extensor plantar responses, dysmetria, dysphagia, and muscle wasting. One individual was noted to have mirror movements. Brain imaging showed cerebellar atrophy. Dysautonomia was common, mostly manifest as orthostatic hypotension and difficulties with bladder or bowel control. Several individuals with earlier onset had severe involuntary weight loss and muscle wasting due to dysphagia or swallowing difficulties. The severe weight loss in some patients was considered to have contributed to their premature deaths between 28 and 47 years of age. Patients with earlier onset had more severe additional symptomatology. Some patients died in their forties or fifties, although others survived into the late seventies; one patient was still alive at age 80. One patient died at age 28. Postmortem examination of the patient who died at 28 years of age showed mild cerebellar atrophy with neuronal loss and gliosis, a loss of pigmented cells in the substantia nigra, and moderate cell loss in the locus ceruleus. Nerve cells of the myenteric plexus in the esophagus contained p62 (SQSTM1; 601530)-immunoreactive inclusions. Alpha-synuclein (SNCA; 163890) immunoreactivity was seen in brainstem and medulla oblongata neurons, in the hippocampus, and in myenteric ganglion cells in the gastrointestinal tract. Lewy bodies were not observed. Of note, 2 children in family 1 (8 and 4 years of age) who did not carry the pathogenic repeat expansion in the ZFHX3 gene had a more severe and complex neurologic disorder since infancy, including delayed psychomotor development, hypotonia, delayed walking with unsteady gait, balance problems, myoclonic jerks, leg pain, joint hyperlaxity, behavioral problems, incontinence, and cerebellar atrophy on brain imaging.

Figueroa et al. (2024) reported numerous individuals from 8 families, including the large multigenerational Utah pedigree (UTA) previously reported by Gardner et al. (1994) and Flanigan et al. (1996), with SCA4 confirmed by genetic analysis. The Utah kindred could be traced back to a likely common ancestor born in southern Sweden around the start of the 19th century. The 7 newly described families were of German descent (Lubeck, Munchen, Magdeburg 1 and 2, Essen 1 and 2, and Hamburg). Haplotype analysis showed that several of the families shared a common repeat expansion haplotype. Most patients had onset of symptoms in adulthood (range 30 to 65 years), although some patients in later generations had earlier onset in the second or third decade, consistent with genetic anticipation. The presenting feature was usually gait ataxia and impaired balance, followed by dysarthria and peripheral sensorimotor neuropathy. Most had oculomotor defects, such as saccadic pursuit, saccadic intrusions, and gaze-evoked nystagmus. Additional more variable features included dysphagia, chronic cough, and autonomic dysfunction manifest as orthostatic hypotension, erectile dysfunction, and incontinence. Rare patients had exercise-induced dystonia, tongue fasciculations, tremor, myoclonic jerks, extensor plantar responses, and sleep disturbances. Brain imaging showed cerebellar atrophy, sometimes with spinal cord atrophy. Neuropathologic examination of a man who died 30 years after onset showed depletion of Purkinje cells in the cerebellum and neuronal intranuclear inclusions (NII) that were immunoreactive to ZFHX3, ubiquitin, and p62.

Paucar et al. (2024) reported 3 large multigenerational families from southern Sweden with SCA4. The mean age at symptom onset was 56.4 years (range 20 to 60 years), and genetic anticipation was observed in all families. Common features included gait ataxia, cerebellar atrophy, and sensorimotor neuropathy with areflexia. Additional features included dysautonomia with abnormal tilt test, motor neuron involvement (muscle weakness, Babinski sign), eye movement abnormalities (slow saccades, ophthalmoplegia), and movement disorders other than ataxia, such as dystonia and myokymia. Advanced disease was characterized by severe weight loss and recurrent pneumonias sometimes requiring gastrostomy. Moderate to severe loss of myelinated fibers was seen in 2 sural nerve biopsies. Neuroimaging showed progressive atrophy in the cerebellum, brainstem, and spinal cord, and PET scan showed brain hypometabolism in several subcortical regions and the cerebellum. Postmortem neuropathologic examination of 1 patient showed a moderate to severe loss of Purkinje cells in the cerebellum and of motor neurons in the anterior horns of the spinal cord, as well as pronounced degradation of the posterior tracts. Intranuclear, mainly neuronal, inclusions positive for p62 and ubiquitin were sparse, but widespread in the brain and upper spinal cord. Occasional polyQ-positive intranuclear inclusions were present in neurons.

The transmission pattern of SCA4 in the families reported by Wallenius et al. (2024) was consistent with autosomal dominant inheritance with genetic anticipation.

By linkage analysis of a Utah kindred with autosomal dominant SCA with sensory neuropathy, Gardner et al. (1994) identified a candidate disease locus, termed SCA4, on chromosome 16q.

Flanigan et al. (1996) provided full information on the mapping of the SCA4 locus identified by Gardner et al. (1994) to 16q22.1. The disorder was mapped in a 5-generation family with an autosomal dominant, late-onset spinocerebellar ataxia; the gene was tightly linked to the microsatellite marker D16S397 (lod = 5.93 at theta = 0.00).

By linkage analysis of a German family with autosomal dominant SCA, Hellenbroich et al. (2003) identified a 3.69-cM region on chromosome 16q22 between markers D16S3019 and D16S512 (maximum 2-point lod score of 4.48 at D16S3018). Analysis of 9 CAG/CTG tracts in this region showed no evidence for a repeat expansion.

In 2024, 3 different research groups (Wallenius et al., 2024, Figueroa et al., 2024, and Paucar et al., 2024) independently identified a heterozygous GGC(n) trinucleotide repeat expansion in the ZFHX3 gene (104155.0003) as the cause of spinocerebellar ataxia-4 (SCA4) that had been mapped to chromosome 16q22.

In 8 affected individuals from 5 Swedish families with SCA4, Wallenius et al. (2024) identified a heterozygous 3-bp (GGC) repeat expansion in the last coding exon (exon 10) of the ZFHX3 gene (104155.0003); the GGC repeat encoded a glycine residue. All families originated from Skane, the southernmost region of Sweden, and haplotype analysis indicated a founder effect. Two of the families had previously been reported (see Moller et al., 1978 and Wictorin et al., 2014). The repeat was expanded to greater than 40 repeats (range 42 to 74) in affected individuals, whereas the most common nonexpanded repeat length was reported as 21 repeats (range 14 to 26) in controls. The nonexpanded repeat in controls consisted of 20 glycine residues interrupted by a single serine. All nonexpanded alleles had interruptions within the GGC repeat; the interruptions were predominantly synonymous GGT and a nonsynonymous AGT (serine). Pathogenic expanded alleles did not contain interruptions: GGC was the only repeat unit. Genetic anticipation was observed, and there was a correlation between longer repeat expansions and earlier age at symptom onset. Long-read sequencing in a patient from family 1 who had onset at age 37 years showed 57 uninterrupted GGC repeats, whereas a patient in a later generation in family 1 who had onset at 15 years of age had 74 uninterrupted GGC repeats. Functional studies of the variant and studies of patient cells were not performed.

In affected members of 8 families with SCA4, Figueroa et al. (2024) identified a heterozygous a GGC(n) repeat expansion in the last coding exon (exon 10) of the ZFHX3 gene (104155.0003). The repeat, which was found by long-read genome sequencing, segregated with the disorder in the families in an autosomal dominant pattern of inheritance. The normal allele had 21 repeats, whereas the pathogenic GGC repeat was over 45 repeats (up to 61 repeats) and was translated into a polyG domain in-frame with the rest of the ZFHX3 protein as demonstrated in patient fibroblasts. There was a significant inverse correlation between age at onset and repeat expansion length. One of the families was a large multigenerational family from Utah previously reported by Gardner et al. (1994) and Flanigan et al. (1996) who could be traced to a common ancestor in southern Sweden around the start of the 19th century. The 7 other families were from Germany (Lubeck, Munchen, Magdeburg 1 and 2, Essen 1 and 2, and Hamburg). Haplotype analysis showed that several of the families shared a common repeat expansion haplotype. Fibroblast samples from 4 SCA4 patients in the Utah family showed evidence of abnormal autophagy compared to controls. Wildtype ATXN2 was upregulated in all patient samples. Induced pluripotent stem cells (iPSCs) generated from one SCA4 patient from the Utah family with 21/53 repeats contained increased polyG-expanded ZFHX3 protein levels, but became rapidly apoptotic upon induction of differentiation to neurons, whereas control iPSCs easily differentiated into neurons under similar conditions. Neuropathologic examination of 1 patient detected neuronal intranuclear inclusions (NII) in the cerebellum that were immunoreactive to ZFHX3, ubiquitin, and p62 (SQSTM1; 601530).

Using a combination of methods, including long-read sequencing, Paucar et al. (2024) identified a heterozygous GGC(n) repeat expansion in the coding region of the ZFHX3 gene (104155.0003) in affected member of 3 multigenerational Swedish families with SCA4. The pathogenic repeat ranged from 46 to 64 copies. The majority of normal alleles contained 21 copies, with a maximum of 26 copies. Functional studies of the variant were not performed, but neuropathologic studies of 4 patients showed intranuclear inclusions positive for ubiquitin and p62. Polyglycine-positive inclusions were found in neurons in 1 patient.

Exclusion Studies

Edener et al. (2011) excluded a pathogenic pentanucleotide repeat in the BEAN gene, which causes SCA31, as a cause of SCA4 in the family reported by Hellenbroich et al. (2003), indicating that SCA4 and SCA31 are not allelic disorders.