Sodium channel mechanosensitivity: pay a-tension to voltage sensor movement

Organisms must respond appropriately to relevant mechanical stimuli and be structurally and functionally resistive to mechanical perturbations. In animals, excitable cell types detect mechanical stimuli and orchestrate the coordinated movement of contractile organ systems. Hearing, balance, touch and mechanical pain depend on specialized sensory cells, whereas the heart, the gut and skeletal muscles, for example, all contain networks of neurons and muscle cells that generate and respond to movement.

Central to the physiology of mechanosensory cells are effects of mechanical stimuli on membrane ion channels, including the voltage-gated sodium channels (VGSCs) that are fundamental to action potential generation. Membrane stretch was known to increase the magnitude and kinetics of voltage-activated Na⁺ currents (Ou et al. 2003; Morris & Juranka, 2007) but recently in The Journal of Physiology, Beyder et al. (2010) presented an elegant study of how membrane stretch affects the gating of the VGSC Na_V1.5 – a phenomenon of particular interest given the channel's selective expression in the heart and gut.

Ion channels implicated in mechanosensation are either mechanically gated or ‘mechanomodulated’. The primary gating stimulus of mechanomodulated channels is well defined but their sensitivity to it is changed by membrane tension. For voltage-gated channels, a shift in sensitivity to membrane potential means cellular stretch or compression alters membrane excitability.

With the exception of the bacterial mechanogated channel MScL, knowledge of the protein domains that mediate mechanogating or modulation is limited. By expressing Na_V1.5 in a mammalian cell line and testing which gating transitions were affected by membrane stretch, the present study by Beyder et al. demonstrated that modulation of S4 voltage-sensor segments are primarily responsible for Na_V1.5 mechanosensitivity.

For each VGSC gating transition, specific parts of the protein move relative to one another and the cell membrane to change the channel's conformation. Channel opening from the closed, resting state (C) is a two-step process involving a voltage-dependent transition to a closed-activated state (CA), by outward movement of the S4 voltage-sensor segments, the so-called ‘sliding helix’, and, then, a voltage-independent switch to the open configuration (O), by S4 interacting with pore-forming segments S5 and S6. Once open, most channels rapidly inactivate, by movement of a single intracellular domain, although inactivation can also occur prior to channel opening from the closed-activated state (Catterall, 2010).

Using cell-attached patch recordings, Beyder et al. confirmed that membrane stretch (applied through the pipette) increased the peak voltage-activated Na⁺ current by accelerating current activation and shifting voltage sensitivity towards hyperpolarized potentials. These voltage shifts are pronounced and if achieved in situ would substantially move spike threshold towards resting potential.

Beyder et al. went on to show that stretch accelerated both current activation and inactivation. However, when control currents were temporally scaled to stretch currents, a single constant gave good overlap, suggesting that speeding of a single process affects the two processes, most simply activation, as this occurs first. To distinguish if stretch accelerated C→CA or CA→O, cells were held at potentials (normally) insufficient to open channels but enough to favour C→CA. If stretch increased the rate of C→CA then more channels would enter CA and be prone to closed-state inactivation. If stretch favoured CA→O, then closed state inactivation would be equal. They observed the former, suggesting that the membrane tension increased the outward movements of S4.

Beyder et al. also found that membrane stretch slowed recovery from inactivation, suggesting that the inactivated state is stabilized by stretch. This may also be due to effects on S4, as its movement favours binding of the inactivation gate to its receptor.

Given the extensive interaction of S4 domains with the lipid bilayer, this important demonstration that voltage-sensor movement is the locus of mechanosensitivity in VGSCs provides support for lipid–protein interactions influencing eukaryotic channel mechanogating. Determining if such mechanisms play a general role in mechanosensory channels will be of great interest.

It is now unequivocal that VGSCs are mechanosensitive. However, establishing functional significance has proven difficult. A central problem is that given sufficient force all biological processes are ‘mechanosensitive’, i.e. at some threshold, determined by its weakest component, a system's function will change. Although many ion channels are affected by tensions considerably less than the membrane rupture point, it remains contentious if such levels are reached in situ. Hence, case by case, it must be determined if ion channel mechanosensitivity functions in sensory physiology, is a weak point that contributes to pathophysiology or, potentially, only occurs experimentally.

To demonstrate that mechanomodulation of VGSCs (or other channels) contributes to sensory physiology, it must be shown that they are exposed to sufficient membrane tension in vivo and that such modulation shapes natural responses. Unfortunately, no drugs selectively affect VGSC mechanomodulation, but Beyder et al.'s data could potentially help in the design of mutant VGSCs with altered mechanosensitivity.

Mechanosensitivity is not restricted to Na_V1.5, as Na_V1.4 (expressed in skeletal muscle) and Na_v1.6 are also mechanosensitive, whilst Na_V1.8 and 1.9 are found selectively in sensory neurons activated by painful pressure. Comparison of S4 sequences across these channels for evidence of selection for mechanosensitivity would now be of interest. Alternatively, susceptibility to membrane tension in multimodal channel proteins maybe a tolerable evolutionary byproduct because they are rarely exposed to such forces. However, such susceptibility could be a factor in pathological processes involving aberrant mechanical forces, such as cardiac arrhythmias or traumatic injuries (again if the generation of sufficient membrane tension occurs in situ). Interestingly, stretching axons containing Na_v1.6 generated a sodium flux that adversely affected neuronal survival (Wang et al. 2009). Furthermore, two mutations in Na_V1.5 associated with cardiac arrhythmias alter its mechanosensitivity (Banderali et al. 2010) and VGSC blockers are used to treat this condition. If an unavoidable sensitivity to membrane stretch makes VGSCs an especially susceptible element to mechanical trauma, they may be important therapeutic targets for a number of conditions.