Association of A Novel Splice Site Mutation in P/Q-Type Calcium Channels with Childhood Epilepsy and Late-Onset Slowly Progressive Non-Episodic Cerebellar Ataxia

Abstract

Episodic ataxia type 2 (EA2) is characterized by paroxysmal attacks of ataxia with typical onset in childhood or early adolescence. The disease is associated with mutations in the voltage-gated calcium channel alpha 1A subunit (Cav2.1) that is encoded by the CACNA1A gene. However, previously unrecognized atypical symptoms and the genetic overlap existing between EA2, spinocerebellar ataxia type 6, familial hemiplegic migraine type 1, and other neurological diseases blur the genotype/phenotype correlations, making a differential diagnosis difficult to formulate correctly and delaying early therapeutic intervention. Here we report a new clinical phenotype of a CACNA1A-associated disease characterized by absence epilepsy occurring during childhood. However, much later in life the patient displayed non-episodic, slowly progressive gait ataxia. Gene panel sequencing for hereditary ataxias led to the identification of a novel heterozygous CACNA1A mutation (c.1913 + 2T > G), altering the donor splice site of intron 14. This genetic defect was predicted to result in an in-frame deletion removing 44 amino acids from the voltage-gated calcium channel Cav2.1. An RT-PCR analysis of cDNA derived from patient skin fibroblasts confirmed the skipping of the entire exon 14. Furthermore, two-electrode voltage-clamp recordings performed from Xenopus laevis oocytes expressing a wild-type versus mutant channel showed that the genetic defect caused a complete loss of channel function. This represents the first description of distinct clinical manifestations that remarkably expand the genetic and phenotypic spectrum of CACNA1A-related diseases and should be considered for an early diagnosis and effective therapeutic intervention.

Keywords: CACNA1A mutation; P/Q-type calcium channel; absence epilepsy; cerebellar ataxia; next-generation sequencing.

Figure 1

The patient presents with cerebellar atrophy and epilepsy. (A) Brain MRI section showing marked atrophy of the cerebellar vermis (T1-weighted image). (B) Brain MRI section showing slight atrophy of the cerebellar hemispheres (T2-weighted image). (C) Interictal EEG showing a frontotemporal spike, most prominent in the temporal leads (FT10, T8, TP10).

Figure 2

The patient carries a c.1913 + 2T > G variant in the CACNA1A gene. Sequencing chromatograms of a partial cDNA sequence of a healthy subject and the affected patient showing (A) exon 13/14 and exon 14/15, and (B) the abnormal junction (AG) between the exon 13 and 15 of the patient. (C) Alignment of the wild-type (top) and mutant (bottom) Cav2.1 protein showing the amino acid sequence that is predicted to be deleted (dashed line) as a consequence of the in-frame mutation.

Figure 3

Localization of the identified splice-site mutation. Membrane topology of a Cav2.1 channel showing the predicted structure which includes four homologous domains (I–IV) with six transmembrane segments (S1–S6) in each domain. The position of the novel c.1913 + 2T > G variant is indicated as a red mark at the beginning of the S4–S5 linker of the II domain. Highlighted in red are shown the regions of the Cav2.1 channel that are predicted to be missed as a result of the mutation. The diagram also shows the localization of single nucleotide substitutions, truncating and splice-site mutations in the secondary structure of the Cav2.1α₁ subunit. All CACNA1A missense, nonsense, and splice-site mutations that are depicted in the figure were collected from the HGMD database.

Figure 4

The c.1913 + 2T > G variant deletes exon 14. The RT-PCR analysis of cDNA from patient and control fibroblast using primers flanking exon 14 detected a 259 bp product in the patient. The PCR product obtained from healthy control individuals was 391 bp. A 100 bp DNA ladder was used as a marker to estimate the amplicon size.

Figure 5

The mutation causes the total loss of function of the Cav2.1 channels. (A) Overlaid representative current traces recorded from a Xenopus laevis oocyte injected with wild-type or mutant cDNA for Ca_V2.1 and auxiliary subunits. Currents were elicited by a depolarizing ramp from –90 mV to +90 mV, lasting 1 s in duration. (B) Bar plot showing the peak whole-cell current amplitudes for the indicated channels. Notice that membrane currents could not be detected (N.D.) in any of the cells injected with mutant cDNA. (C) Bar plot showing the peak whole-cell current amplitudes recorded from oocytes injected with wild-type Cav2.1, co-injected with mutant cDNA (1:1 ratio), or injected with mutant cDNA, together with the accessory subunits. Three different batches of oocytes were collected, and the recorded currents were averaged (data are mean ± SE; for panel B, n = 30 for wild-type; n = 44 for mutant; for panel C, n = 36 for wild-type; n = 26 for wild-type+mutant; n = 59 for mutant; the statistical analysis of graph B revealed a significance of p < 00001, whereas the data reported in graph C were not statistically significant except for the mutant alone).

Figure 6

Structural modelling of Cav2.1 channels. Side view of the homology model of the domains II and IV of the alpha subunit (left). The front and back domains were removed for clarity. The crystal structure of the rabbit Cav1.1 alpha subunit (accession code: 5gjw) was used to show the corresponding structures that are affected by the mutation identified in the patient’s Cav2.1 channel. The relevant helices are highlighted using different colors: S4 (blue), S4–S5 linker (cyan), and S5 (red). On the right-hand side is the top view of the channel to further show the structural and functional relationships between the S4–S5 linker and the voltage-sensor S4, as well as the S5 and the ion conducting pore. The magenta spheres represent the Ca²⁺ ions located within the selectivity filter. Roman numbers mark the corresponding domains.