**1. Introduction**

Quaternary ammonium salts (QAs) have long been used as channel blockers in the characterization of K+ channels [1,2]. Initial studies were performed on the giant squid axon and *D. melanogaster* Shaker channels [3–9], and it has already been suggested that QAs bind to the conduction pore, hence impeding the K+ current. Nonetheless, the length of the alkyl chains and the hydrophobicity of the different QAs determine differences in their molecular mechanisms of blockade. Thus, while the shorter-chain QAs seem to block the channel by simply obstructing the conduction pathway, the longer-chain, more hydrophobic derivatives are also believed to induce a slow channel inactivation process [6].

The elucidation of molecular details on the interaction between QAs and K+ channels have advanced significantly with the resolution of the X-ray structures of QAs bound

153

**Citation:** Giudici, A.M.; Díaz-García, C.; Renart, M.L.; Coutinho, A.; Prieto, M.; González-Ros, J.M.; Poveda, J.A. Tetraoctylammonium, a Long Chain Quaternary Ammonium Blocker, Promotes a Noncollapsed, Resting-Like Inactivated State in KcsA. *Int. J. Mol. Sci.* **2021**, *22*, 490. https://doi.org/10.3390/ijms22020490

Received: 10 December 2020 Accepted: 1 January 2021 Published: 6 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

to KcsA, a prokaryotic K+ channel from *Streptomyces lividans*. This channel is a homotetrameric membrane protein, where each monomer includes two α-helical transmembrane segments (TM1 and TM2) and N- and C-terminal cytoplasmic ends (Scheme 1). The four Cterminal ends are arranged as a helical bundle, with a conformation that is sensitive to pH, acting as a channel gate (inner gate) [10]. On the other hand, the channel pore includes an aqueous cavity, a short tilted helix, and a selectivity filter (SF) with the sequence TVGYG, homologous to that of the eukaryotic K+ channels [11], that constitutes a second channel gate (outer gate). The backbone carbonyls of the SF residues conform to four K+ binding sites (sites S1–S4, from the extracellular to the intracellular side) [11,12], which can adopt different conformations at high or low K+ concentrations [12–22]. The conformation at a low K+ concentration shows no ions at the center of the SF (sites S2 and S3), thus adopting a "collapsed" structure that impedes ion flow through it. K+ binds to S1 and S4 in this conformation, with an average occupancy of just one ion distributed between those two sites. However, at high K+ concentrations, a conformational change is induced by a second K+ entering the filter, with a final average occupancy of two K+ ions per channel, either at the S1–S3 or the S2–S4 positions, thus enabling ion conduction [12,13,17,22]. The nonpermeant Na+, on the other hand, does not induce such a conformational change and shows an average occupancy of one ion per channel at the S1 and S4 sites [23].

In terms of functional activity, KcsA undergoes a cycle that includes four main different states, which reveals the concerted action of the two channel gates. At neutral pH and in the presence of K+, the channel is in a closed/conductive resting state, whereby the cytoplasmic inner gate impedes ion flow, while the SF (the outer gate) displays a conductive form. At acidic pH, the inner gate opens, allowing ion flow in this open/conductive state and making KcsA a proton-activated channel [24–26]. However, this is not a stable state, and the outer gate evolves to a conformation reminiscent of C-type inactivation in eukaryotic K+ channels [27–29], which impedes ion flow in this open/inactivated form [27,28,30]. Such an inactivation process is modulated by a network of interactions that includes the so-called inactivation triad, i.e., residues E71, D80, and W67 from each subunit [17]. The cycle is completed when the pH returns to neutrality, which closes the inner gate, causing the transient closed/inactivated state to evolve to the initial closed/conductive resting state [31].

The crystallographic studies on KcsA complexed to different QAs reveal that the QA binding site for the hydrophilic ammonium head group is located in the internal waterfilled cavity of the channel, directly underneath the innermost cation binding site (S4) of the SF (see Scheme 1). The QAs are further stabilized in the cavity through the insertion of their alkyl chains of varying lengths into the hydrophobic channel wall so that this hydrophobic component becomes an important source of binding stability. Thus, hydrophobic compounds such as TBA+ (tetrabutylammonium), THA+ (tetrahexylammonium) and TOA+ (tetrabutylammonium) bind to KcsA with very high affinity (nM range) [32,33]. In the particular case of TOA+–KcsA complexes, the X-ray structure also reveals that the channel SF is in a collapsed conformation at pH 7.0 and high K+ concentrations, i.e., the S2 and S3 K+ binding sites are absent, similar to that previously seen in the collapsed X-ray structure of the channel alone in the presence of very low K+ concentrations. This led those authors to conclude that such a collapsed structure is an inactivated state induced by the binding of TOA+. However, those X-ray studies used an L90C KcsA mutant without the C-terminal domain, in which the additional presence of a Fab fragment bound to the extracellular channel loop is suspected to restrict the conformational plasticity of the SF [17,34]. Indeed, previous studies from our laboratory on several inactivated models of KcsA, some of which were also predicted as collapsed from X-ray studies, have shown that the stack of K+ binding sites in the inactivated filters remains accessible to cations, as in the resting channel. Therefore, rather than being collapsed, the inactivated SF seems "resting-like" [35]. In apparent agreemen<sup>t</sup> with our observations, other authors found only modest conformational changes in the G77 residue during inactivation compared to the resting state [36], further supporting the tenet that the SF remains "resting-like" upon inactivation, with all

four K+ binding sites accessible to cations. Our goal in this paper is to contribute to the elucidation of this controversy on the conformation of the SF in the inactivated channel state by studying in detail the effects of TOA+ binding to the KcsA channel, which, up to now, was believed to result in an inactivated channel with a collapsed SF [33].

**Scheme 1.** Transmembrane portion of KcsA. Crystallographic structure of the C-terminal truncated KcsA in the closed state with TOA+ (green balls and sticks, with a larger ball representing the quaternary nitrogen) bound at the cavity (PDB: 2W0F). In the top view (panel (**A**)), two of the four subunits appear faintly drawn to facilitate the observation. The four W67 residues are depicted as blue sticks. In the side view (panel (**B**)), only two of the four monomers have been drawn as a solid red ribbon for the sake of clarity. Each monomer consists of two transmembrane helices (TM1 and TM2) connected by the P-loop region, a short, tilted pore helix, and the selectivity filter (SF). The thick grey lines indicate the membrane limits. Both views illustrate how the TOA+ alkyl chains traverse the TM2 helix. Panel (**C**) represents the tetraoctylammonium ion in perspective.
