1 INTRODUCTION
Hyperexcitability is at the core of a rather diverse set of disorders affecting the heart, skeletal muscles, and the nervous system. Out-of-control electric activity is involved in the development of several types of epilepsies, chronic pain syndromes, neuromuscular disorders, cardiac arrhythmias, and even psychiatric disorders (Rogawski & Löscher, 2004). These conditions can originate from genetic conditions (mutations altering the operation of channels themselves, or proteins involved in their modulation), or can be due to damages caused by mechanical injury, inflammation or ischemia.
To suppress hyperexcitability, sodium channels are the primary target because they are responsible for the fast onset of action potentials, as well as their all-or-none and self-regenerating nature. A small fraction (0.5 to 5%) of sodium channels evades rapid inactivation and produces a current component named persistent or late sodium current. This component plays a crucial role in the initiation of action potentials, and its enlargement has been observed in several hyperexcitability-related pathologies (Cannon, 2018; Lampert et al., 2006; Makielski, 2016; Meisler, 2019; Stafstrom, 2007; Tang et al., 2015; van Zundert et al., 2012).
In order to design therapeutically useful sodium channel inhibitors, one encounters the seemingly impossible task of having to prevent pathological hyperexcitability, while maintaining the normal physiological activity of nerves and muscles. Interestingly, there are compounds, which are able to carry out this feat – at least to a certain extent. These compounds include antiarrhythmic, antiepileptic, antispastic, etc. compounds. The trick that enables them to do so is state-dependence, i.e., their preference for certain conformational states of the channel protein. Most sodium channel inhibitors prefer inactivated state to resting state, they bind to it more rapidly, and/or dissociate from it more slowly. The fact that an inhibitor has a higher affinity to inactivated state, means that drug-bound inactivated channels form an energetically favorable complex. This implies both slower dissociation of the ligand from this conformation and slower recovery from inactivated to resting state. The latter effect – called modulation of channel gating – is an inseparable element of state-dependent inhibition, as described by the modulated receptor hypothesis (Hille, 1977). High affinity to inactivated state ensures delayed dissociation, while modulation ensures delayed conformational transition to the low affinity resting state, thus restraining both possible pathways to recovery (dissociation followed by recovery, or recovery followed by dissociation).
Pathological states induced by injury, inflammation, ischemia, tumor, or epilepsy alter the electrical characteristics of excitable cells, which may include a depolarized membrane potential due to energy failure, increased leakage currents, left-shifted voltage sensitivity of sodium channels, and increased persistent component of the sodium current (Fischer et al., 2017; Hammarström & Gage, 2002; Ma et al., 2006; Morris & Joos, 2016; Q. Zhang et al., 2019; Zheng et al., 2012). These changes make sodium channels more likely to be in open and inactivated conformations, therefore state-selectivity alone is enough for preferential inhibition of pathological tissue. On a first impression, one could suppose that the stronger the state-preference is, the better the drug will be.
However, the temporal aspect must also be considered. Action potentials are fired repetitively, and pathological behavior of neurons is often manifested as high-frequency firing. The extent of state-dependence is, therefore, not the only crucial aspect, equally important is the onset/offset dynamics of state-dependent binding. Its significance is obvious in the case of Class 1 antiarrhythmics, where subclasses a, b, and c differ in their association/dissociation kinetics, but the same is true for the much higher firing rates in central and peripheral neurons. For selective inhibition of cells firing at pathologically high frequency, the ideal drug should work as a low pass filter, with a steep frequency response, to be able to distinguish pathological and physiological rates of firing. This, however, presents a theoretical limit for state-dependent binding, because fast dissociation precludes high affinity. One would want both fast binding/unbinding dynamics, and high state dependence. However, high state-dependence requires high affinity to inactivated state, and high affinity means slow dissociation, which means that binding/unbinding dynamics cannot be fast.
Intriguingly, riluzole seems to be able to elude this limitation. Here we examine how this is possible.
We first describe the peculiar pattern of inhibition, we observed during and after riluzole perfusion. We use a voltage-clamp protocol where block and modulation of the channels are monitored in parallel. Paradoxically, two distinct recovery processes seem to coexist, with rates differing by more than two orders of magnitude. We presume that this peculiarity may be the key to being able to feature fast kinetics and high affinity at the same time. A fast recovery allows channels to regain their ability to conduct ions within ~10 ms. In spite of their ability to conduct, however, channels remain modulated by the drug for a much longer time: it requires ~2 s for the channels to recover from the modulatory influence. This mechanism enables riluzole to function as a low pass filter with an exceptionally steep frequency response, which is probably a key element of its distinctive therapeutic efficacy.
We set out to identify the physical processes that underlie fast and slow recovery processes with the help of conformation-selective photolabeling-coupled electrophysiology, and in silico docking experiments.