The μ-conotoxin μ-KIIIA, from Conus kinoshitai, blocks mammalian neuronal voltage-gated sodium channels (VGSCs) and is a potent analgesic following systemic administration in mice. We have determined its solution structure using NMR spectroscopy. Key residues identified previously as being important for activity against VGSCs (Lys7, Trp8, Arg10, Asp11, His12 and Arg14) all reside on an α-helix with the exception of Arg14. To further probe structure-activity relationships of this toxin against VGSC subtypes, we have characterised the analogue μ-KIIIA[C1A,C9A], in which the Cys residues involved in one of the three disulfides in μ-KIIIA were replaced with Ala. Its structure is quite similar to that of μ-KIIIA, indicating that the Cys1-Cys9 disulfide bond could be removed without any significant distortion of the α-helix bearing the key residues. Consistent with this, μ-KIIIA[C1A,C9A] retained activity against VGSCs, with its rank order of potency being essentially the same as that of μ-KIIIA, namely, NaV1.2 > NaV1.4 > NaV1.7 ≥ NaV1.1 > NaV1.3 > NaV1.5. Kinetics of block were obtained for NaV1.2, NaV1.4 and NaV1.7, and in each case both kon and koff values of μ-KIIIA[C1A,C9A] were larger than those of μ-KIIIA. Our results show that the key residues for VGSC binding lie mostly on an α-helix and that the first disulfide bond can be removed without significantly affecting the structure of this helix, although the modification accelerates the on- and off-rates of the peptide against all tested VGSC subtypes. These findings lay the groundwork for the design of minimized peptides and helical mimetics as novel analgesics.

In the preparation of synthetic conotoxins containing multiple disulfide bonds, oxidative folding can produce numerous permutations of disulfide bond connectivities. Establishing the native disulfide connectivities thus presents a significant challenge when the venom-derived peptide is not available, as is increasingly the case when conotoxins are identified from cDNA sequences. Here, we investigate the disulfide connectivity of μ-conotoxin KIIIA, which was predicted originally to have a [C1-C9,C2-C15,C4-C16] disulfide pattern based on homology with closely-related μ-conotoxins. The two major isomers of synthetic μ-KIIIA formed during oxidative folding were purified and their disulfide connectivities mapped by direct mass spectrometric CID fragmentation of the disulfide-bonded polypeptides. Our results show that the major oxidative folding product adopts a [C1-C15,C2-C9,C4-C16] disulfide connectivity, while the minor product adopts a [C1-C16,C2-C9,C4-C15] connectivity. Both of these peptides were potent blockers of NaV1.2 (Kd 5 and 230 nM, respectively). The solution structure for μ-KIIIA based on NMR data was recalculated with the [C1-C15,C2-C9,C4-C16] disulfide pattern; its structure was very similar to the μ-KIIIA structure calculated with the incorrect [C1-C9,C2-C15,C4-C16] disulfide pattern, with an α-helix spanning residues 7–12. In addition, the major folding isomers of μ-KIIIB, an N-terminally extended isoform of μ-KIIIA identified from its cDNA sequence, were isolated. These folding products had the same disulfide connectivities as for μ-KIIIA, and both blocked NaV1.2 (Kd 470 and 26 nM, respectively). Our results establish that the preferred disulfide pattern of synthetic μ-KIIIA/μ-KIIIB folded in vitro is 1-5/2-4/3-6 but that other disulfide isomers are also potent sodium channel blockers. These findings raise questions about the disulfide pattern(s) of μ-KIIIA in the venom of Conus kinoshitai; indeed, the presence of multiple disulfide isomers in the venom could provide a means to further expand the snail's repertoire of active peptides.