Functional consequences of leucine and tyrosine mutations in the dual pore motifs of the yeast K⁺ channel, Tok1p

Abstract

Tandem pore-loop potassium channels differ from the majority of K⁺ channels in that a single polypeptide chain carries two K⁺-specific segments (P) each sandwiched between two transmembrane helices (M) to form an MP₁M–MP₂M series. Two of these peptide molecules assemble to form one functional potassium channel, which is expected to have biaxial symmetry (commonly described as asymmetric) due to independent mutation in the two MPM units. The resulting intrinsic asymmetry is exaggerated in fungal 2P channels, especially in Tok1p of Saccharomyces, by the N-terminal presence of four more transmembrane helices. Functional implications of such structural asymmetry have been investigated via mutagenesis of residues (L290 in P₁ and Y424 in P₂) that are believed to provide the outermost ring of carbonyl oxygen atoms for coordination with potassium ions. Both complementary mutations (L290Y and Y424L) yield functional potassium channels having quasi-normal conductance when expressed in Saccharomyces itself, but the P₁ mutation (only) accelerates channel opening about threefold in response to depolarizing voltage shifts. The more pronounced effect at P₁ than at P₂ appears paradoxical in relation to evolution, because a comparison of fungal Tok1p sequences (from 28 ascomycetes) shows the filter sequence of P₂ (overwhelmingly TIGYGD) to be much stabler than that of P₁ (mostly TIGLGD). Profound functional asymmetry is revealed by the fact that combining mutations (L290Y + Y424L)—which inverts the order of residues from the wild-type channel—reduces the expressed channel conductance by a large factor (20-fold, cf. <twofold for the single mutants).

Appendix

Sequence alignments and dendrograms for the two MPM motifs in fungal 2P potassium channels are shown in Figs. 11 and 12, respectively. Sequences were obtained from the currently annotated fungal proteomes, and were first aligned pairwise with the Saccharomyces sequence, by means of the BLAST algorithm at National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/blast/bl2seq/wblast2.cgi). Final alignment was guided by reckoning on highly conserved residues at the ends of several segments shown (Fig. 11) and refined by means of the Clustal algorithm. The six residues (L67) at the end of MPM1 and six more at the beginning of MPM2 belong to the long cytoplasmic loop which connects MPM1 and MPM2 in each species. Sequences marked with (*), in the legend, appear truncated on the N-terminus, to contain fewer than eight transmembrane helices. The final alignments, consensus sequences, and dendrograms were calculated by means of Lasergene software (DNASTAR, Inc., Madison, WI, USA), scaled according to inferred base changes.

Fig. 11

MPM alignments. See text (above) for source and treatment of sequences. During the alignment process, probable missing residues (–) were replaced with “X” (any), which were formally treated as nonconserved. This was operationally equivalent to assigning an alignment bias to the segment ends. In the drawing, presumed functional segments of the molecule are artificially separated by vertical white bars. No homologue corresponding to Ac2 (Aspergillus clavatus (2)) was found in the genome–proteome database. Species key as follows: Mg = Magnaporthe grisea; Nf2 = Neosartorya fischeri (2)*; Gz = Gibberella zaea; An2 = Aspergillus niger (2); Nc = Neurospora crassa; At2 = Aspergillus terreus (2)*; Cglo = Chaetomium globosum; Ao2 = Aspergillus oryzae (2); Pn = Pheosphaeria nodorum; Af2 = Aspergillus fumigatus (2)*; Nf3 = Neosartorya fischeri (3); Yl1 = Yarrowia lipolytica (1); Ac3 = Aspergillus clavatus (3); Yl2 = Yarrowia lipolytica (2); Af3 = Aspergillus fumigatus (3); Ca = Candida albicans; Nf1 = Neosartorya fischeri (1); Cgla = Candida glabrata; An1 = Aspergillus niger (1); Le = Lodderomyces elongisporus; At1 = Aspergillus terreus (1); Kl = Kluyveromyces lactis; Ao1 = Aspergillus oryzae (1); Ps = Pichia stipitis; Af1 = Aspergillus fumigatus (1); Sc = Saccharomyces cerevisiae; Ac1 = Aspergillus clavatus (1); Pg = Pichia guilliermondii

Fig. 12

MPM dendrograms. See text (above) and legend to Fig. 11 for method details. Despite the apparent faster accumulation of mutations in MPM1 than in MPM2, the major clustering of species and isoforms, as denoted by equivalent colors on the two dendrograms, is similar between MPM1 and MPM2

Of the total of 32 sequences now available, the 28 shown are all from ascomycete fungi; the other four were from basidiomycete fungi (Cryptococcus, Ustilago, and Coprinopsis), but two of those were only weakly homologous. Sequence conservation, judged by identity with the consensus sequence for each MPM unit and shown by the shading in Fig. 11, varies from ~35% (Pichia guilliermondii) to ~75% (Neosartorya fischeri (3)); conservative substitutions (similarity) increase these numbers by ~15%.

The dendrograms (Fig. 12) make clear that the two channel-forming regions, MPM1 and MPM2, have evolved quite differently from each other, with MPM1 accumulating nearly twice as many coding mutations from the apparent stem as MPM2. This is globally consistent with the simple comparison of filter sequences (see “Introduction”).