Ataxic phenotype with altered CaV3.1 channel property in a mouse model for spinocerebellar ataxia 42

概要

Introduction
Spinocerebellar ataxia 42 (SCA42) is a neurodegenerative disorder recently shown to be caused by c.5144G>A (p.Arg1715His) mutation in CACNA1G, which encodes the T-type voltage-gated calcium channel CaV3.1. (Coutelier et al., 2015; Morino et al., 2015). Voltage-gated calcium channels are classified into low-voltage-activated (T-type) and high-voltage-activated (L-, P/Q-, N-, and R-types) channels, based on their threshold for voltage-dependent activation. CACNA1G encodes one of the T-type voltage-gated calcium channels, CaV3.1, which is widely expressed in the brain, especially in neurons in the thalamus, hippocampus, and inferior olivary nucleus (ION), as well as cerebellar Purkinje cells (PCs). CaV3.1 plays an important role in regulating calcium entry and membrane potential (Zamponi et al., 2015), including rebound burst firing of neurons in the thalamus (Kim et al., 2001) and cerebellar nuclei (Perez-Reyes and Lory, 2006). Here, we describe a large Japanese family with SCA42. Postmortem pathological examination revealed severe cerebellar degeneration with prominent Purkinje cell loss without ubiquitin accumulation in an SCA42 patient. To determine whether this mutation causes ataxic symptoms and neurodegeneration, we generated knock-in mice harboring c.5168G>A (p.Arg1723His) mutation in Cacna1g, corresponding to the mutation identified in the SCA42 family.

Methods
Cacna1g Arg1723His-KI mice were generated using CRISPR/Cas9 technology. Sanger sequencing was performed to verify c.5168 G>A mutation in Cacna1g, with the mutation appropriately confirmed in homozygous, heterozygous, and WT strains. We evaluated the body weight and behavioral tests including accelerated rotarod test, the hang wire test and a footprint test. Also, we conducted both conventional pathological analyses including HE and Klüver–Barrera staining and immunohistochemical staining, using cerebellar slices from knock-in mice. The total number of PCs, the length of the PC layer, and the areas of the molecular layer and total cerebellum in the HE-stained sagittal section of the vermis were measured. Using acute cerebellar slices obtained from mice, we analyzed both voltage-clamp and current-clamp recordings from PCs or ION neurons by whole-cell patch-clamp electrophysiology. In addition, to determine whether synaptic function is impaired in PCs of KI mice, both excitatory postsynaptic currents (EPSCs) and spontaneous inhibitory postsynaptic current (sIPSC) were recorded from PCs and long-term depression (LTD) was induced by pairing parallel fiber stimuli in conjunction with PC depolarization.

Results
According to behavioral tests and pathological examinations, both heterozygous and homozygous Cacna1g-Arg1723His-KI mice developed an ataxic phenotype from the age of 11–20 weeks and showed Purkinje cell loss at 50 weeks old. Degenerative change of residual Purkinje cells and atrophic thinning of the molecular layer were conspicuous in homozygous knock-in mice. These results indicated that p.Arg1723His mutation triggers neurodegeneration of cerebellar cortex, especially in PCs. Electrophysiological analysis of Purkinje cells using acute cerebellar slices from young mice showed that the point mutation altered the voltage dependence of CaV3.1 channel activation and reduced the rebound action potentials after hyperpolarization, although it did not significantly affect the basic properties of synaptic transmission onto Purkinje cells including EPSC, sIPSC and LTD. Finally, we revealed that the resonance of membrane potential of neurons in the inferior olivary nucleus was decreased in knock-in mice, which indicates that p.Arg1723His Cav3.1 mutation affects climbing fiber signaling to Purkinje cells.

Discussion
We assessed the motor coordination of SCA42 model mice using the rotarod test and footprint test. Both heterozygous and homozygous KI mice demonstrated an adult-onset mild ataxic phenotype, recapturing the essential clinical feature of dominant inheritance of SCA42. These results indicate that the mutation causes an ataxic phenotype by a process distinct from a simple loss-of-function mechanism because previous reports show that Cacna1g knockout mice have no motor impairment (Park et al., 2010). Furthermore, transgenic mice overexpressing wild-type CaV3.1 do not show an ataxic phenotype either (Ernst et al., 2009). Our pathological evaluation of SCA42 model mice revealed degeneration of PCs, as seen in human SCA42. Considering that the ataxic phenotype was already apparent at 11–20 weeks of age, triggering of the ataxic phenotype in KI mice might be attributable to functional changes, rather than simple loss of PCs. We showed that the mutation indeed altered the voltage dependence of activation of T-type calcium currents toward positive in native PCs. We found a significant decrease of rebound firing of PCs after hyperpolarization in KI mice. These results suggest that even a modest change in voltage dependence of T-type calcium channels has a significant impact on the rebound firing of PCs. It is generally recognized that T-type calcium channels in ION neurons play crucial roles in amplifying oscillation of the subthreshold membrane potential at their resonant frequency and thereby facilitating the generation of action potential. The altered ION property revealed in the present study may underlie the pathogenesis of SCA, given the proposed importance of oscillatory activity of the olivo-cerebellar circuit in timing control and the generation of complex temporal patterns. In conclusion, our study shows not only that a point mutation in CACNA1G causes an ataxic phenotype and Purkinje cell degeneration in a mouse model, but also that the electrophysiological abnormalities at an early stage of SCA42 precede Purkinje cell loss.