µ-conotoxins are blockers of the voltage-gated sodium channels (VGSC)s. These toxins are suggested as drug leads for some disorders such as chronic pain. In this regard, selectivity of a toxin for a particular Nav1 channel is an essential requirement. µ-conotoxins show different affinity for various kinds of VGSCs, but they are not sufficiently selective. To improve selectivity, one needs to determine the binding modes for various Nav1-µ-conotoxin complexes and find the interactions that give rise to binding. To this end, we study and compare the interactions of three µ-conotoxins (GIIIA, PIIIA, KIIIA) with Nav1.4.
We use Bacterial Nav crystal structure as a template to create a model of Nav1.4 pore. We have first studied the binding of GIIIA to Nav1.4, which is a well-known conotoxin blocker. We create initial complexes of toxins and channel using HADDOCK. For each toxin we run blind and restrained docking to find all possible binding conformations. The complexes that reproduce the interactions of main residues identified in mutation experiments are further refined in molecular dynamics (MD) simulations lasting 20-30 ns. From the trajectory data, we determine the pair-residue interactions between the channel and toxin. Our results are in good agreement with several mutational data and rationalize some of the experimental observations. Further validation for the proposed Nav1.4-GIIIA complex is provided by binding free energy obtained from the potential of mean force of GIIIA. These data altogether show that our model of Nav1.4 is fairly accurate and provides a useful model for studying toxin interactions with VGSCs.
We use this model of Nav1.4 to study the interactions of KIIIA and PIIIA employing the same protocols used in GIIIA. We compare the interactions of three conotoxins (GIIIA, PIIIA, KIIIA) with Nav1.4 and study the differences between the binding modes. Our results show that the binding modes of the three toxins to Nav1.4 are different despite the high similarity of toxin sequences.
To get a better understanding of µ-conotoxins binding to Nav, we also study the binding mode of GIIIA and bacterial Nav channels. While the bacterial Nav channels are not of direct interest, they could still provide useful models for understanding the µ-conotoxin interactions with the mammalian Nav channels.
Our results show that in all complexes there is a Lys or Arg that is inserted to the pore but its partners are not the same. Also contributions of different channel domains are quite diverse. These data provide the first clues for designing selective peptides for blocking VGSCs.