The Kv3.3 mutations are responsible for both early-onset spinocerebellar ataxia type 13 (SCA13) in a French family and adult-onset, dominantly inherited ataxia in a Filipino family. First author Michael Waters and colleagues made the connection when they narrowed the search for the offending mutations, which had been mapped to overlapping segments of chromosome 19, to a small region of DNA. Although the candidate site contained around 40 genes, the authors focused on Kv3.3 because it is expressed in the affected area of the brain, the Purkinje neurons of the cerebellum.

Waters and colleagues found that though the Filipino and French families harbored different mutations, they were both in the Kv3.3 gene. The mutations result in single amino acid substitutions, a histidine for arginine in the Filipino family (R420H) and a leucine for phenylalanine (F448L) in the French pedigree. These substitutions are in highly conserved positions which play crucial roles in the operation of the voltage-gated channel. Histidine 420 is found in the voltage sensing domain, while phenylalanine 448 is in a region of the protein that regulates the opening of the channel pore. Waters and colleagues found that the R420H mutation rendered the channel inactive, while the F448L mutation shifted the activation potential for the channel about 13 mV more negative. Though this shift did not seem to affect the activation of the channel, Waters found that deactivation of the F448L channel progresses much more slowly than the wild type.

Curiously, knockouts of Kv3.3 result in no obvious phenotype in mice. The authors suggest that this is because Kv3.1 and Kv3.2 family members may compensate, and indeed Kv3.1/Kv3.3 double knockouts show signs of tremor and ataxia (see Espinosa et al., 2001).

Why do the reported mutations lead to ataxia? Waters and colleagues found the R420H mutant acts in a dominant-negative fashion, forming heteromultimeric channels with other Kv3 family members and inactivating them. On the other hand, because the F448L mutant channel closes more slowly than the wild type, it is likely to delay repolarization of neurons and therefore increase the interval between action potentials. This spells particular trouble for Purkinje neurons of the cerebellum, which fire at high frequency. The 448 mutation is, therefore, expected to be more severe, and probably explains the early onset of disease in the French family.

But the kinetics of the F448L mutant are interesting for another reason—they suggest a potential mechanism for neurodegeneration. Because these channels are required to repolarize neurons, shutting off sodium influx in the soma and calcium influx in the dendrites of Purkinje cells, the slowly rectifying current of the F448L mutant could leave cells open to increased calcium influx, which in turn could trigger apoptotic pathways and cell death. The author suggest that “these channels be examined as mutational and therapeutic targets in bursting neurons in the hippocampus and substantia nigra, especially in conjunction with neurodegenerative diseases such as Parkinson and Alzheimer disease.”—Tom Fagan.

Reference:
Waters MF, Minassian NA, Stevanin G, Figueroa KP, Bannister JPA, Nolte D, Mock AF, Evidente VGH, Fee DB, Müller U, Dürr A, Brice A, Papazian DM, Pulst SM. Mutations in voltage-gated potassium channel KCNC3 cause degenerative and developmental central nervous system phenotypes. Nature Genetics. March 5, 2006. Advanced online publication. Abstract