A Na+ channel mutation linked to hypokalemic periodic paralysis exposes a proton-selective gating pore

Abstract

The heritable muscle disorder hypokalemic periodic paralysis (HypoPP) is characterized by attacks of flaccid weakness, brought on by sustained sarcolemmal depolarization. HypoPP is genetically linked to missense mutations at charged residues in the S4 voltage-sensing segments of either CaV1.1 (the skeletal muscle L-type Ca(2+) channel) or NaV1.4 (the skeletal muscle voltage-gated Na(+) channel). Although these mutations alter the gating of both channels, these functional defects have proven insufficient to explain the sarcolemmal depolarization in affected muscle. Recent insight into the topology of the S4 voltage-sensing domain has aroused interest in an alternative pathomechanism, wherein HypoPP mutations might generate an aberrant ionic leak conductance by unblocking the putative aqueous crevice ("gating-pore") in which the S4 segment resides. We tested the rat isoform of NaV1.4 harboring the HypoPP mutation R663H (human R669H ortholog) at the outermost arginine of S4 in domain II for a gating-pore conductance. We found that the mutation R663H permits transmembrane permeation of protons, but not larger cations, similar to the conductance displayed by histidine substitution at Shaker K(+) channel S4 sites. These results are consistent with the notion that the outermost charged residue in the DIIS4 segment is simultaneously accessible to the cytoplasmic and extracellular spaces when the voltage sensor is positioned inwardly. The predicted magnitude of this proton leak in mature skeletal muscle is small relative to the resting K(+) and Cl(-) conductances, and is thus not likely to fully account for the aberrant sarcolemmal depolarization underlying the paralytic attacks. Rather, it is possible that a sustained proton leak may contribute to instability of V(REST) indirectly, for instance, by interfering with intracellular pH homeostasis.

Figure 1.

Gating charge displacement of rNaV1.4 WT and R663H channels. Oocytes were held at −100 mV, and gating charge displacement was determined by a series of 15-ms voltage commands between −130 and +40 mV in 5-mV increments, following a 15-ms prepulse to −130 mV. Linear leak and capacitance currents were subtracted by a P/−8 protocol from −130 mV. Bath solutions contained TEA⁺ as the predominant cation. Representative gating currents are displayed in A, for a mock-injected oocyte (0), and oocytes expressing WT or R663H channels (denoted at top). Command voltages eliciting the currents in each trace are displayed at the right. The voltage dependence of normalized Q_On charge displacement is shown in B, for WT channels (open circles, n = 8), and R663H mutants (filled circles, n = 7). Curves are fit with a Boltzmann function yielding the following values: WT, V_1/2 = −37.3 ± 2.4 mV, k = 13.3 ± 0.6 mV; and R663H, V_1/2 = −38.7 ± 1.9 mV, k = 11.0 ± 0.6 mV).

Figure 2.

R663H channels are associated with an aberrant inward current. In A, representative steady-state current responses to 300-ms voltage commands between −140 and +30 mV from a holding potential of −100 mV are shown, for mock-injected oocytes (0), or oocytes expressing WT or R663H channels (denoted at top; recordings are from the same oocytes whose gating charge movement is depicted in Fig. 1 A). The mean steady-state current during the last 100 ms of the command pulse is plotted versus voltage in B, for mock-injected (open squares, lying beneath WT points), WT- (open circles), and R663H-expressing oocytes (filled circles). The R663H-expressing oocyte exhibits a hyperpolarization-induced inward current not observed in the WT expressing oocyte. C depicts the R663H and WT-associated currents (same data as in B) after subtraction of the mean nonspecific leak from a pooled population of mock-injected oocytes.

Figure 3.

R663H channels are the origin of the aberrant inward current. In A, the nonspecific leak from a pooled population of mock-injected oocytes was subtracted from steady-state currents from both WT- and R663H-expressing oocytes. The leak-corrected current amplitudes elicited by the −140 mV command voltage are plotted against maximal gating charge displacement for individual oocytes expressing WT (open circles) or R663H channels (filled circles). Linear fits to the data are overlaid. The amplitude of the nonlinear, inward current scales with increased R663H channel expression (black line), whereas no aberrant inward current is evident in oocytes expressing comparable levels of WT channels (dotted line). In B, this scaling is used to normalize current amplitudes from different oocytes to corresponding Q_on,Max, to facilitate comparison across oocyte populations exhibiting different levels of Na⁺ channel expression. A prominent inward current is characteristic of R663H-expressing oocytes (filled circles), whereas there is little aberrant inward current in WT-expressing oocytes (open circles).

Figure 4.

The R663H-associated gating pore is impermeable to large cations, but permissive for protons. The selectivity of R663H-associated currents for different cations was assessed. In all experiments, leak subtraction was achieved as in Fig. 2 B, using nonspecific leak currents derived from pooled data from mock-injected oocytes exposed to the same ionic conditions as the experimental group. The current–voltage relationship of R663H-specific currents, normalized to maximal Q_On as in Fig. 3 B, is plotted in A. Normalized R663H gating-pore currents measured with TEA⁺ as the external cation (filled circles, same data as in Fig. 2 C for reference), is not significantly different from normalized current densities recorded in external K⁺ (open circles, n = 6), Na⁺ (open diamonds, n = 5), or NMDG⁺ (open squares, n = 5). The selectivity of the hyperpolarization-activated R663H conductance for protons was assessed in experiments depicted in B, by manipulating the proton driving force through changes in the transmembrane pH gradient. This was accomplished by buffering the intracellular pH to different values while the extracellular pH was kept constant at 7.4. The R663H current density exhibits the same directionality and amplitude as the densities in A when the pH gradient is symmetric (intracellular pH 7.4, open circles). When the proton driving force was increased at hyperpolarized voltages by buffering the intracellular pH to 9.0, the normalized R663H-associated inward currents increased (filled circles). No aberrant inward current was observed in WT-expressing oocytes when the intracellular pH was either 7.4 (open squares) or 9.0 (filled squares).

Figure 5.

Intracellular acidification promotes outward proton current via the R663H gating-pore. To elicit an outward current through the R663H gating-pore, E _H+ was shifted to ∼−140 mV by buffering the cytoplasmic pH to ∼5.0, while the external pH was maintained at 7.4. Representative raw current traces are shown in A, for oocytes expressing WT or R663H channels (denoted at top). Command voltages eliciting individual current responses are indicated (in mV) in the figures. Outward R663H gating pore currents are evident. After subtraction of the nonspecific leak derived from pooled data from mock-injected oocytes recorded under the same ionic conditions, and normalization to maximal Q_On, the corresponding current–voltage relationships for R663H (closed circles, n = 5) and residual WT (open circles, n = 4) currents are shown in B. In C, the normalized conductance–voltage relationship of the R663H gating-pore proton current (black circles, note inverted scale on the right) is compared with the R663H Q_On–voltage relationship derived from the same population of oocytes (gray circles). Both datasets are fit with Boltzmann functions yielding the following values: G _H+,(V), V_1/2 = −21.3 ± 3.9 mV, k = −8.0 ± 2.1 mV; Q_On,(V), V_1/2 = −26.1 ± 2.5 mV, k = 12.0 ± 1.3 mV (n = 5 for both sets).