Collaborative Science Unites Researchers and a Novel Spastic Ataxia Gene

The hereditary ataxias are diverse and varied, comprising roughly 90 genes constituting the primary ataxias^{1, 2} and nearly 600 additional genes that can present with symptoms that include cerebellar ataxia.³ Despite this, at least one-third of patients remain undiagnosed by current technologies, including whole exome sequencing,^4–7 suggesting that additional genes await discovery.

In this issue, Seong et al. describe a collaborative international effort identifying 12 patients from 7 families with a novel autosomal recessive cerebellar ataxia caused by mutation in VPS13D.⁸ Their largest family, with 5 affected individuals, had been previously reported and designated phenotypically as SCAR4 (spinocerebellar ataxia recessive type 4) characterized by adult onset cerebellar ataxia, macrosaccadic intrusions, pyramidal signs, and neuropathy, and thus resolves the precise molecular basis of that disorder. Another two families presented somewhat similarly with adult onset spasticity or spastic ataxia. The other four families had onset ranging from birth to age 5. Three of these had varying degrees of developmental delay with the most severe having intellectual disability, seizures, and white matter changes on brain MRI.⁸ The collaborative effort uniting these cases was brought about in part through the use of GeneMatcher,⁹ one of a growing numbers of tools and initiatives that unite researchers interested in specific genes based on genomic and phenotypic data with the goal of facilitating resolution of complex diagnostic dilemmas, identifying the molecular basis of unsolved rare diseases, and discovering new disease genes.^{10, 11}

Separately, in this same issue, Gauthier et. al. report a family with 2 patients with spastic ataxia, chorea, and dystonia also harboring variants in VPS13D. Another 5 cases from 4 additional families were located at collaborative sites, also utilizing contributions from GeneMatcher.¹² In this cohort, all patients showed early onset before age 12, developmental delay, intellectual disability, axial hypotonia, and chorea, eventually progressing to spastic ataxia and dystonia in adulthood.¹² Two patients from different families also had microcephaly and seizures. Further complicating the phenotypic spectrum, one family presented with early tremor progressing to dystonia, spasticity and mild ataxia as adults. MRI of the brains of the cohort showed T2/FLAIR hyperintensities either in a pattern restricted to the putamen and caudate, or more diffusely involving the white matter, reminiscent of patterns described in Leigh syndrome.¹²

The VPS13 family consists of four genes, VPS13A, VPS13B, VPS13C, and VPS13D,^{3, 13} all of which have now been shown to cause recessive neurological disease. VPS13A causes a progressive neurodegenerative disorder, chorea-acanthocytosis, primarily characterized by progressive chorea but can include other movement problems such as parkinsonism or dystonia, frontal cognitive impairment, and, in some patients, seizures and/or myopathy.¹⁴VPS13B is associated with a clinically variable syndromic neurodevelopmental disorder, termed Cohen syndrome, generally associated with features of microcephaly, intellectual disability, motor delay, joint hypermobility, neutropenia, and progressive retinopathy, among other findings.¹⁵ Mutations in VPS13C lead to an early-onset rapidly progressive form of parkinsonism with pyramidal signs and early cognitive impairment.¹⁶VPS13D is now the final member of the VPS13 family to be associated with human neurological disease. The gene is highly intolerant to missense variation and even more so to loss of function variants.^{8, 12, 17} VPS13D has recently been shown to be essential for mitochondrial metabolism in Drosophila, including viability, autophagy, and clearance.¹⁸ The gene is also essential in human cell lines and its deletion leads to abnormal mitochondrial morphology.^{18, 19} While the precise cellular functions of the other family members are not yet fully established, mutation of VPS13C also appears to lead to mitochondrial dysfunction.¹⁶

Consistent with the proposed role of VPS13D in mitochondrial function,¹⁸ Seong et al. also noted that disruption of the Drosophila homologue was lethal and caused severely altered mitochondrial morphology.⁸ Large spherical mitochondria and impaired trafficking were seen when loss was restricted to motor neurons.⁸ Similar structural defects were observed in patient fibroblasts.⁸ In patients, muscle biopsy within one affected family showed subsarcolemmal mitochondrial aggregates and mild lipidosis supporting mitochondrial dysfunction, although enzymatic studies in patient fibroblast cell lines were normal.¹² Not surprisingly, disruption of mitochondrial function is not an uncommon theme in cerebellar ataxia.^2,20

These reports illustrate the clinical challenge of achieving a molecular diagnosis among disorders that lie on the phenotypic spectrum bridging disorders of ataxia and spasticity.^{4, 21} The genetics of this new disorder appear complex and both studies acknowledged difficulty in establishing clear correlations between genotype and phenotype. Seong et al. noted that patients with both a nonsense and a missense variant tended to have adult onset disease (8 of 9 patients) and earlier and more severe phenotypes were seen in patients with suspected splicing mutations.⁸ Gauthier et al. saw early onset in all cases, although with a variable range from birth to age 12. The majority of cases (3 of 5 families) had a missense variant coupled with a variable deletion or duplication disrupting the protein reading frame.¹² Interestingly, the remaining two families possessed only missense variants (one family homozygous and the other compound heterozygous), one of which had the most severe phenotype of the entire cohort. ¹² Considering the lethality of VPS13D knockout in flies,^{8, 18} it is reasonable to hypothesize that the presence of a hypomorphic allele in a haploinsufficient background is the minimum requirement for disease, but given the varying types of mutations, their location within the protein, and the diverse genetic backgrounds between the patients in these studies, more work is needed to ascertain the nature of the differences leading to more severe presentations.

The decreasing cost and widespread utilization of genomic sequencing technology has considerably broadened the search space for novel genetic disorders to span the entire human genome. This has enabled the creation of large and rapidly searchable research datasets for gene discovery. These studies clearly illustrate the power of linking such efforts as several of the cases were brought together using a combination of one such tool, GeneMatcher, in conjunction with the authors extensive networks of clinical collaborators.^{8, 12} As precision medicine health care models²² become more commonplace, a critical effort should be made to broadly unite the comprehensive patient genomic datasets being created with precise clinical and phenotypic information and subsequently create mechanisms to effectively link these data to further accelerate discovery and translate research back to patients more rapidly. The ongoing development of methods to efficiently clinically annotate these various datasets from the electronic health record while insuring privacy of the genetic and health information will be crucial. In the modern genomic era, as the low-hanging fruit rapidly disappears, leaving newly discovered genes that show ever-decreasing prevalence, the availability of centralized resources comprising unified well-phenotyped genomic datasets from patients with complex neurological disorders, such as cerebellar ataxia and/or spasticity, will greatly facilitate the identification of families who share mutations in suspected novel disease genes.