Voltage-gated potassium (K+) channels display a wide variety of conductances and gating properties in vivo. This diversity can be attributed not only to the presence of many K(+)-channel gene products, but also to the possibility that different K(+)-channel subunits co-assemble to form heteromultimeric channels in vivo. When expressed in Xenopus oocytes or transfected cells, K(+)-channel polypeptides assemble to form tetramers. Certain combinations of Shaker-like subunits have been shown to co-assemble, forming heteromultimeric channels with distinct properties. It is not known, however, whether K(+)-channel polypeptides form heteromultimeric channels in vivo. Here we describe the co-localization of two Shaker-like voltage-gated K(+)-channel proteins, mKv1.1 and mKv1.2, in the juxtaparanodal regions of nodes of Ranvier in myelinated axons, and in terminal fields of basket cells in mouse cerebellum. We also show that mKv1.1 and mKv1.2 can be coimmunoprecipitated with specific antibodies that recognize only one of them. These data indicate that the two polypeptides occur in subcellular regions where rapid membrane repolarization may be important and that they form heteromultimeric channels in vivo.

Mice lacking the voltage-gated potassium channel alpha subunit, K(V)1.1, display frequent spontaneous seizures throughout adult life. In hippocampal slices from homozygous K(V)1.1 null animals, intrinsic passive properties of CA3 pyramidal cells are normal. However, antidromic action potentials are recruited at lower thresholds in K(V)1.1 null slices. Furthermore, in a subset of slices, mossy fiber stimulation triggers synaptically mediated long-latency epileptiform burst discharges. These data indicate that loss of K(V)1.1 from its normal localization in axons and terminals of the CA3 region results in increased excitability in the CA3 recurrent axon collateral system, perhaps contributing to the limbic and tonic-clonic components of the observed epileptic phenotype. Axonal action potential conduction was altered as well in the sciatic nerve--a deficit potentially related to the pathophysiology of episodic ataxia/myokymia, a disease associated with missense mutations of the human K(V)1.1 gene.