Search

The two-pore-domain potassium channels TASK-1 (KCNK3) and TASK-3 (KCNK9) modulate the electrical activity of neurons and many other cell types. We expressed TASK-1, TASK-3 and related reporter constructs in Xenopus oocytes, mammalian cell lines and various yeast strains to study the mechanisms controlling their transport to the surface membrane and the role of 14-3-3 proteins. We measured potassium currents with the voltage-clamp technique and fused N- and C-terminal fragments of the channels to various reporter proteins to study changes in subcellular localisation and surface expression. Mutational analysis showed that binding of 14-3-3 proteins to the extreme C-terminus of TASK-1 and TASK-3 masks a tri-basic motif, KRR, which differs in several important aspects from canonical arginine-based (RxR) or lysine-based (KKxx) retention signals. Pulldown experiments with GST fusion proteins showed that the KRR motif in the C-terminus of TASK-3 channels was able to bind to COPI coatomer. Disabling the binding of 14-3-3, which exposes the KRR motif, caused localisation of the GFP-tagged channel protein mainly to the Golgi complex. TASK-1 and TASK-3 also possess a di-basic N-terminal retention signal, KR, whose function was found to be independent of the binding of 14-3-3. Suppression of channel surface expression with dominant-negative channel mutants revealed that interaction with 14-3-3 has no significant effect on the dimeric assembly of the channels. Our results give a comprehensive description of the mechanisms by which 14-3-3 proteins, together with N- and C-terminal sorting signals, control the intracellular traffic of TASK-1 and TASK-3.

The two-pore-domain potassium channels TASK-1, TASK-3 and TASK-5 possess a conserved C-terminal motif of five amino acids. Truncation of the C-terminus of TASK-1 strongly reduced the currents measured after heterologous expression in Xenopus oocytes or HEK293 cells and decreased surface membrane expression of GFP-tagged channel proteins. Two-hybrid analysis showed that the C-terminal domain of TASK-1, TASK-3 and TASK-5, but not TASK-4, interacts with isoforms of the adapter protein 14-3-3. A pentapeptide motif at the extreme C-terminus of TASK-1, RRx(S/T)x, was found to be sufficient for weak but significant interaction with 14-3-3, whereas the last 40 amino acids of TASK-1 were required for strong binding. Deletion of a single amino acid at the C-terminal end of TASK-1 or TASK-3 abolished binding of 14-3-3 and strongly reduced the macroscopic currents observed in Xenopus oocytes. TASK-1 mutants that failed to interact with 14-3-3 isoforms (V411*, S410A, S410D) also produced only very weak macroscopic currents. In contrast, the mutant TASK-1 S409A, which interacts with 14-3-3-like wild-type channels, displayed normal macroscopic currents. Co-injection of 14-3-3zeta cRNA increased TASK-1 current in Xenopus oocytes by about 70 %. After co-transfection in HEK293 cells, TASK-1 and 14-3-3zeta (but not TASK-1DeltaC5 and 14-3-3zeta) could be co-immunoprecipitated. Furthermore, TASK-1 and 14-3-3 could be co-immunoprecipitated in synaptic membrane extracts and postsynaptic density membranes. Our findings suggest that interaction of 14-3-3 with TASK-1 or TASK-3 may promote the trafficking of the channels to the surface membrane.