Novel CACNA1A Variant p.Cys256Phe Disrupts Disulfide Bonds and Causes Spinocerebellar Ataxia

Spinocerebellar ataxia (SCA) is a progressive, autosomal dominant neurodegenerative disorder typically associated with CAG repeat expansions.

We describe here the novel heterozygous missense mutation p.Cys256Phe (C256F) in a German SCA patient, proving its pathogenicity by demonstrating loss-of-function changes in transfected tsA-201 cells. A C256F homology model strongly suggests that the affected cysteine forms a disulfide bond that stabilizes a functionally relevant extracellular domain.

Patients and Methods
The proband was examined at the Department of Neurology, Medical Faculty of the RWTH Aachen University Hospital. Informed consent was obtained for molecular genetic analyses and further scientific use. A nextgeneration sequencing-based multigene panel was This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. analyzed at the Center for Genomics and Transcriptomics in Tübingen, Germany. For functional analyses, we introduced the C256F mutation into the human Cav2.1 α1-subunit and expressed it together with accessory β4eand α2δ2-subunits in tsA-201 HEK cells. Ca 2+ currents were measured using the whole-cell patch-clamp technique. 5 We generated a Cav2.1 α1 homology model in complex with the α2δ-subunit, using the advanced homology modeling tool in Maestro Release 2019-4. For details, see Supplementary Information.

Patient Phenotype
A 63-year-old woman had been suffering from progressive gait unsteadiness for more than 10 years. In the family history (Fig. 1A), the patient's father was reported with dementia and parkinsonism. Her three siblings were reportedly healthy, but the patient's 32-year-old daughter was affected by an unspecific motor disability reported as infantile cerebral palsy. In our clinical examination, the proband had a pronounced omni-directional, gaze-evoked nystagmus, a reduced suppression of the vestibulo-ocular reflex, ataxic finger-to-nose and heel-to-shin tests on both sides, and a scanning speech pattern. There were no signs of spasticity or muscle weakness, migraine, or episodic neurological symptoms. Except for mild vibratory loss (4/8 at the ankles), we did not observe any sensory involvement. The Scale for the Assessment and Rating of Ataxia score was 6/40 points. A cerebral MRI (magnetic resonance imaging) revealed a profound cerebellar atrophy with preponderance of the vermis area (Fig. 1B). Extensive laboratory tests did not provide any known acquired cause of ataxia in this patient.

Molecular Genetics
We first ruled out pathogenic repeat expansions in SCA1, SCA2, SCA3, SCA6 (13 CAG repeats on both alleles, normal <18 repeats, pathogenic >20 repeats in CACNA1A), SCA7, and SCA17. A diagnostic gene panel revealed the heterozygous variant c.767G>T: p. Cys256Phe/C256F in CACNA1A (NM_001127221) that has neither been described before nor has been identified in healthy control populations (GnomAD). It affects a highly conserved amino acid position ( Figure S1) within a DNA repeat region in exon 5. Two The variant C256F significantly decreased current densities, which is also shown by the boxplots (maximum current density; Student's unpaired t test, **P < 0.01) and representative current traces obtained from depolarization to −24 mV (activation threshold), 1 mV (the maximal voltage, V max ), and 11 mV. Note that the tail current in WT was cut for better visualization. For parameters and statistics, see Table S1. (D) Normalized steady-state activation (•) and inactivation (▲) curves of WT versus C256F (mean AE SEM). The mutation C256F had no effect on the voltage dependence of activation and inactivation. For parameters and statistics see Table S1. [Color figure can be viewed at wileyonlinelibrary.com] pathogenic variants have been previously described at adjacent positions. At the same position, the exchange from cysteine to arginine was classified as pathogenic (Uniprot) in a patient with EA-2. For C256F, in silico predictions were mostly deleterious (11 pathogenic vs. 2 benign).

Functional Characterization of the C256F Missense Variant
Heterologous expression of wild-type and mutant Cav2.1 channel complexes in tsA-201 cells revealed an approximately 45% and highly significant (P < 0.01, Student's unpaired t test) reduction in maximal Ca 2+ current (ICa) density (ICa normalized to the cell size; pA/pF) for the mutant compared to wild-type ( Fig. 1C; Table S1). When analyzed at comparable current amplitudes, no effects on the voltage dependence of activation and inactivation gating were found ( Fig. 1D; Table S1). Instead, we found a small but significant increase in current inactivation independent from current amplitude and after prespecified time points of 250 and 500 ms during a 5-second depolarizing pulse to the voltage of maximal inward current (V max ) ( Figure S2A,B). The reduced current amplitude and faster current inactivation are consistent with a loss of channel function. The observed smaller current densities could not be explained by a lower expression of mutant Cav2.1 α1-protein in Western blots compared to wild-type ( Figure S3).

C256F Homology Model
The recent elucidation of rabbit Cav1.1 heteromeric structures 6 allowed to generate a homology model of Cav2.1 α1 in complex with the α2δ-subunit (Fig. 2). C256 forms a disulfide bond with C281 and is located in close proximity to a second disulfide bond formed by C272 and C287 ( Fig. 2A,B, top inset). Notably, all four cysteines are strictly conserved among all 10 voltage-gated Ca 2+ channel α1-subunits ( Figure S1), suggesting they are of high functional importance. These covalent bonds may help to stabilize an otherwise flexible loop region lining the top of the channel pore. For reference, Figure 2 shows the Ca 2+ ions (light green, Fig. 2B) bound to the selectivity filter (red dots in Fig. 2A). In addition, this loop region forms part of the interaction interface with α2δ-subunits. These subunits are required for proper membrane targeting of the channel complex. 7,8 The C256F missense variant prevents the formation of the disulfide bond with C281, and modeling suggests that its bulky side chain cannot be accommodated in a similar manner as the smaller and polar cysteine side chain (Fig. 2B, bottom inset). This strongly suggests that the C256F mutant disrupts the structural integrity of the loop region via the loss of this disulfide bond, thus increasing its flexibility and/or facilitating altered pairing of cysteine residues.

Discussion
The gene CACNA1A has previously been associated with SCA type 6, typically manifesting with late-onset gait disorders. 1,2 Other reports describe a classic phenotype, including directional nystagmus, limb ataxia, and vibratory loss, reminiscent of our patient as well. 2,9 SCA6 diagnosis usually relies on trinucleotide repeat expansions in the CACNA1A gene, leading to the expanded polyglutamine tract in the Cav2.1 α1 C-terminus. Not only is the polyglutamine tract expressed with the full-length channel, but also exists in a separate peptide (α1ACT), which serves as a transcription factor and is generated by proteolytic cleavage or independently transcribed from a second cistron. 10 Heterozygous loss-of-function missense or stop variants are typically associated with episodic ataxia, a similar autosomal dominant syndrome manifesting by attacks of ataxia, nystagmus, vertigo, and sometimes further signs like dystonia. In contrast to progressive ataxia, this condition manifests in childhood and improves or stabilizes over time.
The patient described here represents one of the rare cases with progressive ataxia without (not even borderline) trinucleotide repeat expansions, but with a missense variant in CACNA1A. Only few other cases have been published with this extraordinary combination before, [11][12][13] and functional evidence on these variants is lacking.
This novel variant has not been identified in healthy controls (gnomAD) and affects a conserved position within a functionally relevant region. At the same position, another missense variant, C256R (also not listed in gnomAD), was observed in a family with EA-2. 14 Another missense mutation disrupting the second disulfide bridge in this loop between C272 and C287 ( Fig. 2; Figure S1), C287Y, has been identified in association with mild to moderate interictal and progressive ataxia. 15 Interestingly, heterologous expression of C287Y under comparable experimental conditions in COS-7 cells also revealed a strong reduction in current amplitude most likely due to a reduced membrane targeting. 15 This loss-of-function phenotype is consistent with other ataxia-associated missense mutations 16 and is very similar to the reduced current amplitudes of the C256F variant reported here. Together, these findings strongly suggest that the disruption of any of the two adjacent disulfide bridges in the extracellular pore loop of homologous repeat I reduces Cav2.1 Ca 2+ channel activity, explaining a loss-of-function mechanism. Taking this clinical, genetic, and functional evidence together, we therefore conclude that CACNA1A-encoded Cav2.1 channels do not tolerate disruption of these disulfide bonds and that the novel variant C256F is pathogenic. For the first time, we provide functional evidence on an SCA-related missense variant, showing that episodic ataxia and SCA both share a loss-of-function pathomechanism. We recommend the screening of CACNA1A for strategically important missense variants in CAG repeat-negative SCA cases.