Animals from cockroaches to humans, utilize electrochemical gradients to perform rapid, long-distance communication between cells like neurons and muscle for these functions, requiring the coordinated action of voltage-gated ion channels (VGICs). Such channels allow for the selective transport of ions down their individual electrochemical gradients and act en masse to control the voltage of the cell promoting further propagation of an electrical signal, or the triggering of other signaling processes. Two classes of VGIC, voltage-gated potassium (Kv) and voltage-gated sodium (Nav) channels are found throughout excitable tissues such as neurons, skeletal and cardiac muscle, and are particularly responsible for the outward and inward currents of action potentials, respectively. While these channel types share structural homology, the presence or absence of different domains and variations in their amino acid sequences lead to drastic differences in their function, and their concerted action is necessary for proper cell function. This variability is further compounded by differential tissue expression of Nav and Kv channel subtypes, leading to more specific physiological and pathophysiological roles. To better understand these channels and their roles in normal and disease states, we require targeted pharmacological tools and a knowledge of the structural underpinnings of their mode of action.The importance of VGICs can be further demonstrated by the ubiquity of toxins produced by organisms from dinoflagellates (saxitoxin) to venomous snakes (-bungarotoxin) that target ion channel function to alter the behavior of animal predators and prey alike. These toxins provide molecular tools for understanding the mechanisms of ion channel function and potential therapeutics for treating diseases either caused or affected by aberrant ion channel activity. Such toxins can be small molecules or peptide toxins and can act on a wide variety of VGICs or selectively target certain channel subtypes with high specificity. The diversity of channel sensitivities to such toxins is matched in the diversity of mechanisms by which these toxins act; they can bind to the conducting pore of the channel, bind to voltage sensing domains altering the kinetics of channel voltage sensitivity, and can even alter the properties of the cell membrane to affect channel function. I have studied the structural elements driving the modulation of two different ion channels by molecules from invertebrates and their derivatives. Two of the three invertebrate toxins, a small molecule toxin and a peptide toxin, were produced by marine snails and have the ability to alter the function of vertebrate VGICs either for defense or predation. The other is a peptide toxin found in the venom of a tarantula used both defensively and to immobilize prey. Such toxins selectively target voltage-gated ion channels to disrupt the normal motor function or produce pain in their victims. The two channels are involved in very different physiological roles: Kv1.4 is found in cardiac myocytes and is responsible for the transient outward current (Ito1) during the phase 1 repolarization of ventricular myocytes; Nav1.7 is found in the fibers of dorsal root ganglion neurons and known to be important in the generation of action potentials important in pain signaling. My Thesis is divided into 3 chapters that combine high-throughput and traditional electrophysiology experiments with computational modeling to study the molecular interactions of toxins and ion channels. While the relationship between structure and function is central to physiology in general, it is especially important in determining molecular mechanisms, potency, and selectivity of ion-channel modulators for specific targets. In Chapter 1 we examined how the structure of toxin derivatives affected their ability to partition and perturb membranes, and used high-throughput electrophysiology to determine how this translated to effects on channel function. In Chapters 2 and 3 we utilized predictive modelling of Nav1.7 channel structure to identify potential toxin binding modes and unseen states of channel domains. The recent explosion of structural data available for membrane proteins has aided in understanding of how channels work, and provided new avenues for modeling of channel-toxin interactions. Experimental validation of such models furthers our structural knowledge of these channels and their molecular mechanisms of modulation, and provides opportunities for designing novel molecular tools for study.

The first chapter of my thesis documents the study of a small molecule inhibitor of Kv1.4— 6-Bromo-2-mercaptotryptamine (BrMT) and its derivatives that alter the kinetics of channel activation. In this study we developed a method for distinguishing between the effects of membrane perturbation and direct interaction with ion channels in understanding the mechanism of ion channel toxins. As a part of a multi-assay structure-activity study, I performed an electrophysiology assay to determine the sensitivity of Kv1.4 channel currents in exogenously expressing Chinese Hamster Ovary (CHO) cells to toxin derivatives. Derivatives were synthesized to create greater stability than the native toxin and alter its potential to interact with the cell membrane. These modifications consisted of multiple substituted sidechains and linkages that affected the hydrophobicity and geometry allowing for a structure-activity relationship on these measures to be determined. Seeking to identify the contribution of these substitutions to their efficacy against Kv1.4 I conducted concentration response experiments examining both the efficacy of these variants and changes in the apparent kinetics of channel activation in response to depolarizing voltage steps with toxin present. I compared the results of my assay with those of measures of membrane partitioning and perturbation to assess the role of these factors in affecting channel function in the presence of toxin. While certain chemical groups affected membrane partitioning and perturbation, the observed effects on the channels could not be explained by the effect on membrane perturbation alone, suggesting direct interactions with the channel were influenced by toxin structure. These findings can help the future study of the mechanisms of ion channel-toxin interactions and enable the design of novel therapeutics from toxin scaffolds.

My second chapter details the search for an accurate model of the interaction of the small cyclic peptide conotoxin, KIIIA, with the voltage-gate sodium channel Nav1.7 to guide the future design of novel inhibitors of the channel. Lacking crystallographic or cryo-EM data on the structure of Nav1.7, we used homology modeling to identify contacts between the channel and toxin residues. I performed an electrophysiological assay to determine the association and dissociation kinetics of toxin variants with key residues. These variants were chosen in concert with corresponding mutations in the channel to remove contacts identified from previous studies and our modeling efforts. Double-mutant cycle analysis, a comparison of the toxin affinities from single mutant (toxin or channel) and double mutant (toxin and channel) conditions, allowed us to identify energetic contributions of the interactions between basic residues on the toxin with acidic residues on the outer channel pore. The orientation of the toxin in our computational model was further corroborated by molecular dynamics simulations. With this combined approach we were able to independently produce a well-supported model of the toxin binding to the outer pore of the channel that comports with recently published structure of the Nav1.2-KIIIA complex. Our model provides insight into potential mechanism of channel block and specificity, while providing a scaffold for future design of novel channel inhibitors for the treatment and study of Nav1.7-associated disorders. I assisted in the study design, prepared test solutions, performed the electrophysiology experiments and data analysis and writing and editing all sections of the manuscript.

My third chapter describes my efforts to characterize multiple states of the Domain IV voltage sensor (DIV-VSD) of the voltage-gated sodium channel Nav1.7, and its potential interaction with the peptide tarantula toxin SGTx-1, and the spider toxin ProTx-II. The DIV-VSD of Nav channels is known to be important for the channel fast-inactivation leading to rapid reduction in Na+ conduction following depolarizing voltages. The crucial role of this voltage-sensing domain has made it a target of researchers aiming to control Na+ conductance and is a known binding site of a class of scorpion toxins and new small-molecule inhibitors. The discovery that the tarantula toxin SGTx-1 could inhibit Nav1.2-DIV-VSD movement provides a potential smaller toxin for redesign of a Nav1.7-DIV-VSD-selective tool. I performed electrophysiological experiments that revealed a slowing of channel kinetics that could be explained by the toxin stabilizing a resting state of the Domain IV VSD as had been suggested for Nav1.2. I modeled active and resting states of the channel based on published structures of homologous ion channels and performed in silico docking of the toxin to identify potential points of contact between the channel and toxin. Candidate residues based on my docking results comport with findings of mutagenesis studies on toxin binding to the potassium channel Kv2.1. I further examined the interaction of the toxin ProTx-II with an active state of the Nav1.7-DIV-VSD guided by recently published density map data. I performed all structural modeling, electrophysiological experiments, data analysis, and writing in this study.