A Recurrent Mutation in CACNA1G Alters Cav3.1 T-Type Calcium-Channel Conduction and Causes Autosomal-Dominant Cerebellar Ataxia

Abstract

Hereditary cerebellar ataxias (CAs) are neurodegenerative disorders clinically characterized by a cerebellar syndrome, often accompanied by other neurological or non-neurological signs. All transmission modes have been described. In autosomal-dominant CA (ADCA), mutations in more than 30 genes are implicated, but the molecular diagnosis remains unknown in about 40% of cases. Implication of ion channels has long been an ongoing topic in the genetics of CA, and mutations in several channel genes have been recently connected to ADCA. In a large family affected by ADCA and mild pyramidal signs, we searched for the causative variant by combining linkage analysis and whole-exome sequencing. In CACNA1G, we identified a c.5144G>A mutation, causing an arginine-to-histidine (p.Arg1715His) change in the voltage sensor S4 segment of the T-type channel protein Cav3.1. Two out of 479 index subjects screened subsequently harbored the same mutation. We performed electrophysiological experiments in HEK293T cells to compare the properties of the p.Arg1715His and wild-type Cav3.1 channels. The current-voltage and the steady-state activation curves of the p.Arg1715His channel were shifted positively, whereas the inactivation curve had a higher slope factor. Computer modeling in deep cerebellar nuclei (DCN) neurons suggested that the mutation results in decreased neuronal excitability. Taken together, these data establish CACNA1G, which is highly expressed in the cerebellum, as a gene whose mutations can cause ADCA. This is consistent with the neuropathological examination, which showed severe Purkinje cell loss. Our study further extends our knowledge of the link between calcium channelopathies and CAs.

Figure 1

Neuropathological Examination of the Cerebellum H&E staining in (A and B) individual AAD-SAL-233-14 (III-9 in Figure 2) and (C and D) a control individual. (A and C) Granular layer of the cerebellum. The black arrows in (C) point to normal glomeruli; normal glomeruli cannot be identified in (A). (B and D) Purkinje cell (PC) layer. Four normal PCs are visible in (D), and one of them is indicated by a black arrow; PC loss is severe in (B) such that only the processes of the basket cells are visible (“empty baskets,” black arrows). Note the additional layer composed of Bergmann glia (white arrows). The asterisks in (B) and (D) indicate the molecular layer, which appears loosened in (B).

Figure 2

Segregation of the p.Arg1715His Change in ADCA-Affected Pedigrees and Alignment of Cav3.1 Orthologs and Paralogs (A) Pedigrees of ADCA-affected families with the p.Arg1715His change. The numbers of affected individuals tested are as follows: six in AAD-SAL-233, three in AAD-GRE-319, and one in AAD-SAL-454. All affected, but no unaffected, individuals harbor the variant in the heterozygous state. (B) Chromatograms show the mutation in individuals AAD-SAL-233-9 (III-3), AAD-GRE-319-12 (II-2), and AAD-SAL-454-10 (III-1). (C) A schematic representation of Cav3.1 shows its organization in four domains, each containing six transmembrane segments; segment S4 contains many positively charged amino acids, such as arginine, and is therefore the voltage sensor. The p.Arg1715His change is located in segment S4 of domain IV. (D and E) Alignment of orthologs (D) and paralogs (E) shows that the arginine residues in Cav3.1 are very highly conserved across all species, T-type channels, and domains.

Figure 3

Electrophysiological Analysis of WT and p.Arg1715His Cav3.1 Calcium Channels (A) Current traces obtained with wild-type (WT) and p.Arg1715His channels at various membrane potentials (−90, −80, −70, −65, −60, −55, −50, −45, −40, −35, −30, −25, and −20 mV) and from a holding potential of −100 mV. Notice the red trace (−50 mV), which shows a smaller current for the p.Arg1715His channel (36% of the maximum current) than for the WT (54%) (B) Averaged current-voltage relationships from traces in (A). The normalized conductance-voltage curve was fitted with a Boltzmann equation: I/I_max = G_max(V_m − E_rev)/(1 + exp(V_1/2 − V_m)/k). (C) Steady-state inactivation curves. The curves were fitted with I/I_max = 1/(1 + exp(V_m − V_1/2)/k). (D) Availability of calcium currents (mean steady-state activation and inactivation curves). The steady-state activation curves were fitted with a Boltzmann equation, G/G_max = 1/(1 + exp(V_1/2 − V_m)/k), where G was calculated as follows: G = I/(V_m − E_rev). (E and F) Time constant of inactivation (τ inact) and activation (τ act) kinetics. Fitting the traces showed in (A) with a double exponential function produced the values shown. (G) Recovery from short-term inactivation according to a two-paired-pulse protocol. (H) Deactivation kinetics (τ deact). (I) DCN neuron firing was simulated with the steady-state activation and inactivation values obtained for the WT (black) and the p.Arg1715His (red) channels. The DCN model used was developed by Luthman et al. with the NEURON simulation environment (see Web Resources) on the basis of the model originally implemented in GENESIS by Steuber et al. The NaP, HCN, and CaLVA conductances were changed to match the “Neuron 1” model described by Steuber et al. In the above-mentioned equations, V_1/2 represents either the half-activation potential (steady-state activation curve) or the half-inactivation potential (steady-state inactivation curve). Other parameters are V_m, membrane potential; E_rev, reversal potential; k, slope factor; G, conductance; G_max, maximum conductance; I, current at a given V_m; and I_max, maximum current. The extracellular solution contained 135 mM NaCl, 20 mM TEACl, 2 mM CaCl₂, 1 mM MgCl₂, and 10 mM HEPES (pH adjusted to 7.44 with KOH). Patch pipettes were filled with an internal solution (140 mM CsCl, 10 mM EGTA, 3 mM CaCl₂, 10 mM HEPES, 3 mM Mg-ATP, and 0.6 mM GTP [pH adjusted to 7.25 with KOH]) and had a typical resistance of 2–3 MΩ. In (B)–(H), WT values are represented by black circles, and p.Arg1715His values are represented by red squares. Data represent the mean ± SEM.