A triplet of residues, GYG, forms part of the consensus sequence of the pore domain and generates the selectivity of potassium channels. However, the two pore domains of tandem pore K+ channels lack sequence identity: P1 has GYG in all cases, (except TREK-1); P2 shows greater variation and has GFG in TASK-1. Do these two regions have similar effects on ionic selectivity? We have made substitutions in the triplet of residues in P1 and P2 of murine TASK-1. Channels were expressed in oocytes taken from Xenopus frogs that had been anaesthetised by immersion in 0.3 % w/v MS222 and killed humanely by destruction of the brain and spinal cord. Two-electrode voltage clamp was used to measure the shift in reversal potential when K+ in the external medium was replaced by Rb+ or Na+.
We made the mutations Y96 î{special} F, L, V, and M in P1, but only Y96F generated functional channels. Surprisingly, this mutation exhibited significantly altered selectivity compared to wild-type. PRb/PK was increased from 0.77 ± 0.02 (n = 6) to 0.99 ± 0.02 (n = 10; mean ± S.E.M.; P < 0.001, using ANOVA). Unlike wild-type, the channel was noticeably Na+-permeant, with an increase in PNa/PK from 0.02 ± 0.003 (n = 6) to 0.44 ± 0.03 (n = 10; P < 0.001). The equivalent residue in P2, F202, was more tolerant of substitution since the mutations F202 î{special} L, V, M, Y, and A all generated functional channels. Selectivity was reduced in every case; e.g. F202M gave PRb/PK = 0.92 ± 0.03 and PNa/PK = 0.65 ± 0.05 (n = 10). The permeability ratio for Rb+/K+ was significantly increased over wild-type (P < 0.01).
The Gly residues of the selectivity sequence are thought to be particularly important for selectivity (Heginbotham et al. 1994). We replaced each Gly in P1 and P2 in turn by Ala. Such mutant channels were non selective. In P1, G97A gave channels with PRb/PK = 0.89 ± 0.05 and PNa/PK = 0.71 ± 0.09 (n = 8). The equivalent mutation in P2, G203A, gave PRb/PK = 0.95 ± 0.05 and PNa/PK = 0.78 ± 0.12 (n = 7). Similar alterations in selectivity were seen with G95A and G201A.
Our results are similar to those obtained on mutation of the equivalent residues of voltage gated K+ channels (Heginbotham et al. 1994), though the effects of substitution of Y by F (or vice versa in P2) are greater. In contrast in inward rectifier K+ channels, mutations of the residues of the GYG triplet in two pore forming regions (out of the four in each channel) results in channels with unaltered selectivity (Dart et al. 1998; So et al. 2001).
We thank the BBSRC for support.
All procedures accord with current UK legislation