A misstep in the multistep process of fast inactivation

Voltage-gated sodium (Na_V) channels are activated by membrane depolarization, allowing an influx of sodium ions into the cell to further depolarize the membrane and drive the activation of additional Na_V channels, producing the upstroke of the action potential. To help break this positive feedback loop and permit membrane repolarization, Na_v channels undergo fast inactivation, in which movement of the domain IV voltage sensor causes the inactivation particle—an isoleucine–phenylalanine–methionine (IFM) sequence in the channel’s intracellular domain—to bind to the channel and inhibit further Na⁺ influx. In this issue of JGP, Liu and Bezanilla reveal that the Na_v channel “CW” mutant abolishes fast inactivation without inhibiting inactivation particle binding, highlighting the importance of the downstream conformational changes that couple IFM binding to pore closure (1).

The inactivation particle was originally proposed to inhibit Na⁺ flux directly by binding to and blocking the Na_v channel pore (2, 3). Later, however, cryoEM structures of fast-inactivated channels revealed that the IFM motif actually binds to a hydrophobic pocket far away from the pore (4, 5). Recently, Yichen Liu and Francisco Bezanilla at the University of Chicago found that the binding of IFM to this pocket triggers a conformational change that closes two hydrophobic rings at the bottom of the pore (6).

Fast inactivation is abolished if the phenylalanine residue in the IFM motif is mutated to glutamine. But the resulting IQM mutant is poorly expressed at the cell surface; so, like many researchers interested in studying fast inactivation, Liu turned to a different inactivation-deficient mutant named CW, for the cysteine and tryptophan residues it substitutes in the pore region of domain I. While the IQM mutation is known to inhibit inactivation particle binding, the mechanism by which the CW mutation abolishes fast inactivation was unknown.

“When we started to make current recordings of CW mutant channels, we noticed that there were some subtle differences from the IQM mutant,” Liu says. For example, CW channels deactivated much more slowly than wild-type or IQM channels upon membrane hyperpolarization.

The channels also showed different responses to ATX-II, a sea anemone toxin that slows fast inactivation by stabilizing the domain IV voltage sensor in the resting state, thereby preventing IFM binding. Since IFM binding is already abolished by the IQM mutation, ATX-II has little effect on IQM channels. But application of the toxin to CW mutant channels doubled the peak ionic current and allowed the channel to open at lower voltages.

Taken together, these differences suggested that the CW mutation acts differently to IQM and that it might abolish fast inactivation without disrupting inactivation particle binding. Indeed, adding the IQM mutation to CW mutants significantly altered the channel’s deactivation kinetics and negated the effects of ATX-II treatment, indicating that inactivation particle binding is crucial to the CW mutant’s phenotype.

Liu and Bezanilla wondered whether, instead of inducing fast inactivation, IFM binding might push CW channels into an alternative open state with low conductance and slower deactivation kinetics. In this scenario, ATX-II treatment would double the peak ionic current by inhibiting IFM binding and preventing the transition to the low conductance state.

In support of this idea, the researchers found that longer depolarization times slowed the deactivation of CW channels even further, as the channels had more time to enter the alternative open state. Accordingly, the off-gating kinetics of CW channels were also sensitive to depolarization time, and voltage-clamp fluorometry experiments suggested that this is mainly because the alternative open state slows the return of the domain II voltage sensor to its resting state.

Although Na_vs can occupy multiple open states, the alternative open state that occurs in CW mutants does not appear to be accessible to either wild-type channels or the IQM mutant. “Researchers may therefore need to be careful about interpreting results obtained using the CW mutant,” Liu says.

Liu and Bezanilla speculate that the bulky tryptophan residue introduced by the CW mutation may interfere with the conformational changes induced by inactivation particle binding. Though this may make the CW mutant a poor model for normal fast inactivation, it highlights the importance of events downstream of the IFM motif. “That’s what we’re working on now,” says Bezanilla. “What is the connection between IFM binding and the movement of the two hydrophobic rings to actually block the pore and inactivate the channel?”