Role of disulfide bonds in the structure and potassium channel blocking activity of ShK toxin

ShK toxin, a potassium channel blocker from the sea anemone Stichodactyla helianthus, is a 35 residue polypeptide cross-linked by three disulfide bridges: Cys3-Cys35, Cys12-Cys28, and Cys17-Cys32. To investigate the role of these disulfides in the structure and channel-blocking activity of ShK toxin, a series of analogues was synthesized by selective replacement of each pair of half-cystines with two cl-amino-butyrate (Abu) residues. The remaining two disulfide pairs were formed unambiguously using an orthogonal protecting group strategy of Cys(Trt) or Cys(Acm) at the appropriate position. The peptides were tested in vitro for their ability to block Kv1.1 and Kv1.3 potassium channels and their ability to displace [I-125]dendrotoxin binding to rat brain synaptosomal membranes. The monocyclic peptides showed no activity in these assays. Of the dicyclic peptides, [Abu12,28] ShK(3-35,17-32) (where the subscript indicates disulfide connectivities) had weak activity on Kv1.3 and Kv1.1. [Abu17,32] ShK(3-35,12-28) blocked Kv1,3 with low nanomolar potency, but was less effective (being comparable to [Abu12,28]ShK(3-35,17-32)) against Kv1.1, [Abu3,35]ShK(12-28-17-32), retained high picomolar affinity against both channels. Corroborating these results, [Abu3,35]ShK(12-28,17-32) had an IC50 ratio relative to native toxin of 18 in the displacement assay, whereas [Abu17,32] ShK(3-35,12-28) and [Abu12,28] ShK(3-35,17-32) had ratios of 69 and 390, respectively. Thus, the disulfide bond linking the N- and C-terminal regions is less important for activity than the internal disulfides. NMR analysis of the [Abu12,28] and [Abu17,32] analogues indicated that they had little residual structure, consistent with their significantly reduced activities. By contrast, [Abu3,35]ShK(12-28,17-32) had a moderately well-defined solution structure, with a mean pairwise root-mean-square deviation of 1.33 Angstrom over the backbone heavy atoms. This structure nevertheless showed significant differences from that of native ShK toxin. The possible interactions of this analogue with the channel and the distinction between native secondary and tertiary structure on one hand and global topology imposed by the disulfide bridges on the other are discussed.