Search

Here we present evidence that the kinesin, Kif5B, is involved in the transportation and axonal targeting of Kv1 channels. We show that a dominant negative variant of Kif5B specifically blocks localization to the axon of expressed, tagged versions of Kv1.3 in cultured cortical slices. In addition, the dominant negative variant of Kif5B blocks axonal localization of endogenous Kv1.1, Kv1.2, and Kv1.4 in cortical neurons in dissociated cultures. We also found evidence that Kif5B interacts with Kv1 channels. Endogenous Kv1.2 colocalized with Kif5B in cortical neurons and coimmunoprecipitated with Kif5B from brain lysate. The T1 domain of Shaker K(+) channels has been shown to play a critical role in targeting the channel to the axon. We have three pieces of evidence to suggest that the T1 domain also mediates interaction between Kv1 channels and Kif5B: Addition of the T1 domain to a heterologous protein, TfR, is sufficient to cause the resulting fusion protein, TfRT1, to colocalize with Kif5B. Also, the T1 domain is necessary for interaction of Kv1.3 with Kif5B in a coimmunoprecipitation assay. Finally, dominant negative variants of Kif5B block axonal targeting of TfRT1, but have no effect on dendritic localization of TfR. Together these data suggest a model where Kif5B interacts with Kv1 channels either directly or indirectly via the T1 domain, causing the channels to be transported to axons.

Shaker K+ channels play an important role in modulating electrical excitability of axons. Recent work has demonstrated that the T1 tetramerization domain of Kv1.2 is both necessary and sufficient for targeting of the channel to the axonal surface [Gu, C., Jan, Y.N. & Jan, L.Y. (2003) Science,301, 646-649]. Here we use a related channel, Kv1.3, as a model to investigate cellular mechanisms that mediate axonal targeting. We show that the T1 domain of Kv1.3 is necessary and sufficient to mediate targeting of the channel to the axonal surface in pyramidal neurons in slices of cortex from neonatal rat. The T1 domain is also sufficient to cause preferential axonal localization of intracellular protein, which indicates that the domain probably does not work through compartment-specific endocytosis or compartment-specific vesicle docking. To determine whether the T1 domain mediates axonal trafficking of transport vesicles, we compared the trafficking of vesicles containing green fluorescent protein-labelled transferrin receptor with those containing the same protein fused with the T1 domain in living cortical neurons. Vesicles containing the wild-type transferrin receptor did not traffic to the axon, in accord with previously published results; however, those containing the transferrin receptor fused to T1 did traffic to the axon. These results are consistent with the T1 domain of Kv1.3 mediating axonal targeting by causing transport vesicles to traffic to axons and they represent the first evidence that such a mechanism might underlie axonal targeting.