Voltage-gated sodium (Na_V) channels are large transmembrane proteins responsible for the initiation and propagation of action potentials in excitable cells (Ahern et al., 2016). After voltage-dependent opening, Na_V channels undergo a rapid spontaneous inactivation process that terminates Na⁺ permeation (Armstrong et al., 1973; Ahern et al., 2016). Fast inactivation is the main hallmark of eukaryotic Na_V channel function that allows cells to repolarize and Na_V channels to become ready for reactivation (Ulbricht 2005; Ahern et al., 2016). Eukaryotic Na_V channels count 9 isoforms (Na_V1.1–Na_V1.9), encoded by 9 distinct genes (SCN1–5A and SCN8–11A) with almost 50% of homology in the amino acid sequences (Goldin et al., 2000; Catterall et al., 2005). These channels consist of the heteromeric assembly of an α-subunit which forms the channel pore and provides its function with two auxiliary β-subunits. The α-subunit of Na_V channels contains four homologous, nonidentical domains, each consisting of six transmembrane segments (S1–S6) (Ren et al., 2001; Catterall 2014). Each domain is organized into two parts, the voltage-sensing domain (VSD) from S1 to S4 and the pore module (PM) between S5 and S6 (Catterall et al., 2017). The S4 segments known as voltage sensors modules are positively charged due to four to eight arginine/lysine residues flanked by two hydrophobic residues (Ahern et al., 2016; Catterall et al., 2017). This positively charged motif serves as a gating charge and moves outward upon depolarization to initiate the channel activation.

Rapid inactivation of Na_V channels results from the occlusion of the pore by a cytosolic inactivation motif that consists of a cluster of hydrophobic residues, isoleucine, phenylalanine, and methionine (IFM) in the intracellular loop connecting domain III (DIII) and DIV (Armstrong et al., 1973; Goldin 2003). In fact, during the activation process, the movement of the voltage sensors, particularly the DIV voltage sensor (VSD4), exposes a hydrophobic site between the S3 and S4 segments. The binding of the IFM motif to this site leads to a physical blockage of ion movement in the cell (Goldin 2003; Catterall 2014). Because VSD1–3 respond more rapidly to membrane depolarization than VSD4 (Chanda and Bezanilla, 2002), inactivation generally occurs after channel activation. Therefore, the inactivation process is intimately linked to channel activation following the movement of VSD4. Indeed, mutation of the IFM motif can completely abolish fast inactivation in Na_V channels (Ahern et al., 2016; Catterall et al., 2017).

Several venomous animals including scorpions have developed an arsenal of toxins that target and disrupt Na_V inactivation to immobilize prey or predators (Bosmans and Tytgat, 2007; Hanck and Sheets, 2007). AahII from Androctonus australis Hector is the most toxic polypeptide responsible for the noxious effects of the venom with an LD50 < 3 ng at an intracerebroventricular administration in a 20 g mouse (Devaux et al., 2004). At the pharmacological level, AahII is an α-scorpion toxin that targets site 3 on Na_V channels and slows the inactivation to sustain sodium influx (Martin et al., 1987; Catterall et al., 2007; Clairfeuille et al., 2019). At the structural level, AahII is a 64-amino-acid peptide stabilized by four disulfide bonds to form a compact β1–α1–β2–β3 scaffold that can highly interact with multiple mammalian Na_V channel subtypes (Housset et al., 1994).

To affect the fast inactivation mechanism, AahII interacts with VSD4 by trapping it in a deactivated state. AahII does not disturb channel activation because DI–III voltage sensors can ensure the opening of the channel even if VSD4 remains deactivated. In the absence of AahII, the S4 helix of VSD4 moves outward to unlatch the intracellular fast inactivation gating machinery, as described before (Clairfeuille et al., 2019).

Another well-known α-scorpion toxin with a similar effect is LqhIII from Leiurus quinquestriatus hebraeus (Jiang et al., 2021). It was recently reported that LqhIII anchors on top of VSD4 and traps the gating charges of the S4 segment in a unique intermediate-activated state stabilized by four ion pairs. This conformation weakens the binding of the fast inactivation gate and favors the opening of the activation gate (Jiang et al., 2021).

In order to counteract human envenoming caused by scorpion stings, several toxin-specific antivenoms were developed using different approaches. The standard immunotherapy method consists of using purified polyclonal antibody F (ab′)2 fragments prepared from equine hyperimmune sera (Chippaux and Goyffon, 1998; Bouaziz et al., 2008). However, the use of these antibody fragments of ≈100 kDa is only moderately effective due to their polyclonal nature and may cause dangerous adverse effects such as anaphylactic shocks (Pepin-Covatta et al., 1996).

Another method based on the use of murine monoclonal antibodies was later developed to neutralize the effect of the AahII toxin. This study led to the development of the murine 4C1 antibody (Bahraoui et al., 1988), subsequently used to develop an AahII-specific scFv (single-chain variable fragment) (Mousli et al., 1999). Similarly, other studies have allowed the development of an scFv against the AahI that belongs to a distinct antigenic and structural group of Androctonus australis Hector scorpion toxins (Devaux et al., 2001). Some years later, a bispecific scFv construct against both AahI and AahII toxins was obtained by engineering techniques able to protect mice against the whole Androctonus australis Hector venom (Juste et al., 2007).

However, all these constructs mentioned above have the major problem of synchronization of kinetic diffusion which is due to the huge difference between the molecular weight of antibodies (MW of approx. 150 kDa), F (ab′)2 (MW of approx. 100 kDa), and their target toxins (MW of approx. 7 kDa). Likewise, the ScFv fragment was not lastingly effective due to a VH–VL unstable complex interaction. Moreover, the neutralizing capacity remains moderate, and their use as a human therapeutic might still generate an undesirable human anti-mouse antibody response (HAMA).

More recently, we have developed an antivenom using another type of toxin binders based on the variable domains of the dromedary heavy-chain antibodies (HCAbs, heavy-chain antibody) naturally lacking light chains and CH1 domains of heavy chains, named VHHs. The specified VHH variable domain encodes an antibody fragment, also named Nb (nanobody), that represents the smallest, intact, natural antigen-binding fragment (Hmila et al., 2008), which has been used to develop an anti-Androctonus australis Hector venom that recognizes specifically the AahI′ toxin from Aah scorpion venom. This anti-AahI′ can neutralize 3 LD₅₀ when tested in vivo using s.c. injection in Swiss mice.

Later, by the same approach, an Nb that neutralizes the most toxic compound of Aah venom, AahII toxin, was developed. The so-called Nb10 is an anti-AahII that targets a unique epitope on AahII and neutralizes 7 LD₅₀ when tested in vivo using Swiss mice (Abderrazek et al., 2009). In silico studies demonstrated whether Nb10 binds the active site of AahII toxin (Ksouri et al., 2018).

These two Nbs are characterized by their small size (MW of approx. 15 kDa), good stability, high level of expression in prokaryotic systems, high solubility, and suitable specificity. At the preclinical level, the performance of a bispecific NbF12-10 toxin-specific Nb format (including the anti-AahII Nb10) was demonstrated in an envenoming simulated animal model (Hmila et al., 2010; Hmila et al., 2012).

While Nb10 constitutes the best candidate that neutralizes AahII toxic effects, the precise functional and structural mechanism by which Nb10 neutralizes AahII and interacts with the AahII–Na_V channel complex remains unknown. Here, we studied at the electrophysiological and structural levels the interactions of a purified AahII toxin and Nb10 nanobody with the Na_V1.5 channel stably expressed in a HEK cell model. Our data reveal the mode of action and structural basis of AahII and the α-scorpion toxins in general, on the Na_V1.5 channel, and proposes Nb10 as a promising antivenom candidate against Androctonus australis Hector scorpion stings.

The pharmacology of scorpion α-toxins is a great gateway to studying the structure and function of ion channels in order to increase our knowledge of their properties. In this report, we studied and characterized the α-scorpion toxin AahII mode of interaction with the Na_V1.5 channel and the modulation of the complex interaction by Nb10, a highly neutralizing anti-AahII nanobody (Abderrazek et al., 2009). For this accomplishment, we used a functional Na_V1.5 channel alpha subunit vector of an adult isoform transfected in HEK293.

Significantly, the AahII toxin slows the fast inactivation of the Na_V1.5 channel by a factor of 4 at 75 nM while current amplitude was not affected, demonstrating its nanomolar ranging effect. Our results agree with a recently reported study showing that AahII toxin slows the Na_V1.7 fast inactivation mechanism when used at 300 nM (EC₅₀ = 51.7 ± 1.5 nM) (Clairfeuille et al., 2019). Their investigations of the cryo-EM-based structural AahII-Na_V1.7 channel interaction showed that α-scorpion toxin AahII binds to two different sites on a nonfunctional chimera in both DI and DIV. However, they did not demonstrate whether AahII bound to either of these sites was functionally active in the chimera (Clairfeuille et al., 2019). In this paper and using a wash perfusion protocol to follow the kinetics of Na_V inactivation, we demonstrated that AahII slows the time inactivation of Na_V1.5 at a low concentration (75 nM) and that the binding of AahII is extracellular.

In the past, pharmacological studies retained a single neurotoxin receptor (site 3) per Na_V channel located in VSD4 at which related gating modifier toxins, such as scorpion α-toxins and sea anemone toxins, bind (Rogers et al., 1996; Gordon et al., 2007). To address this question, we performed a Na_V1.5 site 3 driven docking with AahII and generated 4,500 complexes from which the best orientation showed the involvement of the AahII C-terminal (Arg62 and His64) in the interaction with the DIV-S4 (1,575–1,615 residues) of Na_V1.5. Our data are in accordance with a recent similar study that focuses on LqhIII α-toxin from the deathstalker scorpion Leiurus quinquestriatus hebraeus and the Na_V1.5 channel, showing that LqhIII binds at the extracellular end of the aqueous cleft formed by S1–S2 and S3–S4 helical hairpins in VSD4 through its β2β3 loop and its C-terminal domain (Jiang et al., 2021). Other structure–activity relationship studies show that AahII interacts with Na_V channels through the β2–β3 loop from Gln37 to Pro60 (Clairfeuille et al., 2019) and its C-terminal segment especially involving Arg62 and His64 which are critical for potent modulation of Na_V channels by targeting neurotoxin receptor site 3 (Wang et al., 2003; Benkhadir et al., 2004; Clairfeuille et al., 2019). Indeed, the residues Phe15 and Trp38 and Asn44 from AahII establishing the β2–β3 loop have been described to recognize the VSD4 of the Na_V channels and play an important role in bioactivity of the α-toxin (Kharrat et al., 1989; Gur et al., 2011; Clairfeuille et al., 2019).

It is well admitted that scorpion α-toxins bind to the neurotoxin receptor in a voltage-dependent manner, with a high-affinity binding in the resting state (Clairfeuille et al., 2019; Jiang et al., 2021). Likely, depolarization of the channel reduces toxin affinity and causes its dissociation (Clairfeuille et al., 2019; Jiang et al., 2021). The binding of α-toxins prevents the outward movement of the gating charges of the DIV-S4 segment. In standard conditions, this normal outward movement of the DIV-S4 segment is followed by the unbending of the elbow formed by the DIV S4–S5 loop and the opening of the receptor site for binding of the IFM motif to guarantee the fast inactivation. The binding of AahII α-toxin would counteract this succession of conformational events that lead to fast inactivation.

We have previously demonstrated that Nbs are better tools to interact with toxins and neutralize their effects because of their small size, good stability, high level of expression in a prokaryotic systems, high solubility, and suitable specificity.

Following the development of a highly AahII-specific Nbs, we obtained Nb10 which was the “best-in-class” nanobody able to neutralize the most toxic scorpion α-toxin. This encouraged us to study and characterize the capacity of Nb10 to modulate the interaction of AahII with Na_V1.5 using a whole-cell patch clamp. Nb10 was produced in a prokaryotic system and purified from E. coli bacterial periplasm that meets the required quality control standards (Abderrazek et al., 2009), which allowed us to have enough purified Nb10 quantity that specifically recognizes and binds AahII. Nb10 is a 14 kDa protein able to neutralize 7 LD₅₀ AahII toxicity when injected in mice (association rate constant: K_on = 1.14 × 10⁶ M⁻¹.S⁻¹; dissociation rate constants: K_off = 5.69 × 10^–4 s⁻¹; equilibrium dissociation constant: K_D = 0.49 nM) (Abderrazek et al., 2009).

By an electrophysiological approach, we demonstrated that Nb10 can fully abolish the AahII effect on fast inactivation of the Na_V1.5 channel, with a 1:1 ratio, while current amplitude was not affected. Indeed, perfusion kinetics monitoring of (AahII + Nb10) shows that Nb10 can abolish the AahII-binding action, strongly suggesting that Nb10 shifted or at least altered the AahII-binding site of interaction with VSD4, which releases the DIV-S4 segment to adopt its normal movement and initiates the rapid inactivation of the Na_V1.5 channel.

Recent in silico 3D modeling of the interaction of AahII with Nb10 revealed the potential contribution of the Trp38 residue from AahII with Met103, Arg108, Tyr105, and Ala111 of Nb10 (Ksouri et al., 2018). Indeed, the segment from Ala39 to Ala45 in AahII has been identified as the main region responsible for antigenic reactivity (Devaux et al., 1993).

The in-silico prediction of the AahII–Nav1.5 alpha-subunit interaction showed that AahII binds at the extracellular end of the aqueous cleft formed by the S1–S2 and S3–S4 helical hairpins via its loop (from Trp38 to Asn44) and its C-terminal (Arg62 and His64). Our findings are comparable to the Cryo-EM experimentally solved results which revealed the structure of LqhIII bound to Na_V1.5 at the 3.3 Å resolution, showing that the LqhIII scorpion toxin anchors on top of VSD4, wedged between the S1–S2 and S3–S4 linkers, and through the β2–β3 loop and the C-terminal of the LqhIII α-Scorpion toxins (Jiang et al., 2021). This confirms that α-Scorpion toxins bind to neurotoxin receptor site 3 in a voltage-dependent manner, with high-affinity binding to the resting state (Catterall 1977; Catterall 1979).

Despite the topological difference of the implicated amino acids in the interaction with the two distinct scorpion toxins (AahII and LqhIII), the positions of the bound toxin with the S1–S2 and S3–S4 helical hairpins are remarkably similar. Therefore, we can trust the structural outputs based on a very accurate computational approach. Herein, 3D modeling of the best AahII-Na_V1.5 α-subunit complex showed that AahII Ala39 and Arg62 residues play a crucial role in establishing a stable interaction, through H-bounds, with Na_V1.5 Gln1615 and Lys1616 residues.

These results indicated that Nb10 recognizes a similar or identical surface that directly interacts with the DIV-S4 segment of the Na_V1.5 channel, involving two major residues at positions 39 and 62 in AahII. Indeed, the C-terminal AahII His64 and Arg62 are similarly involved in the Nb10–AahII and Na_V1.5–AahII interaction complexes and participate in the AahII active site (Ksouri et al., 2018).

Furthermore, to address the need for a highly physiological characterization of the interaction between AahII and Nb10 on cells, we performed invasion assays using MDA-MB-231 cells that express a neonatal isoform of Na_V1.5. This Na_V1.5 channel is involved in both migration and invasion processes of the MDA-MB-231 triple-negative breast cancer cell line and promotes metastasis (Roger et al., 2003; Kamarulzaman et al., 2017; Zhang et al., 2018; Luo et al., 2020). Interestingly, as expected AahII increases the invasion of MDA-MB-231 cells by 67.16 ± 9%; the equimolar range of Nb10 can completely suppress this effect (1:1 M ratio). Therefore, both AahII and Nb10 may be useful for biomedical models assessing the involvement of Na_V channel subtypes in the invasion process capability of several tumor cell lines as well as drug delivery into tissues.

Altogether, our results suggest that Nb10 perfectly fits both AahII toxin-binding surfaces on the Na_V channel (β2–β3 loop and C-terminal domain) and dynamically blocks AahII–Na_V1.5 interactions. The two amino acid positions in AahII are crucial for both interactions with Na_V1.5 and Nb10.