ClC-1 mutations in myotonia congenita patients: insights into molecular gating mechanisms and genotype-phenotype correlation

Abstract

Key points: Loss-of-function mutations of the skeletal muscle ClC-1 channel cause myotonia congenita with variable phenotypes. Using patch clamp we show that F484L, located in the conducting pore, probably induces mild dominant myotonia by right-shifting the slow gating of ClC-1 channel, without exerting a dominant-negative effect on the wild-type (WT) subunit. Molecular dynamics simulations suggest that F484L affects the slow gate by increasing the frequency and the stability of H-bond formation between E232 in helix F and Y578 in helix R. Three other myotonic ClC-1 mutations are shown to produce distinct effects on channel function: L198P shifts the slow gate to positive potentials, V640G reduces channel activity, while L628P displays a WT-like behaviour (electrophysiology data only). Our results provide novel insight into the molecular mechanisms underlying normal and altered ClC-1 function.

Abstract: Myotonia congenita is an inherited disease caused by loss-of-function mutations of the skeletal muscle ClC-1 chloride channel, characterized by impaired muscle relaxation after contraction and stiffness. In the present study, we provided an in-depth characterization of F484L, a mutation previously identified in dominant myotonia, in order to define the genotype-phenotype correlation, and to elucidate the contribution of this pore residue to the mechanisms of ClC-1 gating. Patch-clamp recordings showed that F484L reduced chloride currents at every tested potential and dramatically right-shifted the voltage dependence of slow gating, thus contributing to the mild clinical phenotype of affected heterozygote carriers. Unlike dominant mutations located at the dimer interface, no dominant-negative effect was observed when F484L mutant subunits were co-expressed with wild type. Molecular dynamics simulations further revealed that F484L affected the slow gate by increasing the frequency and stability of the H-bond formation between the pore residue E232 and the R helix residue Y578. In addition, using patch-clamp electrophysiology, we characterized three other myotonic ClC-1 mutations. We proved that the dominant L198P mutation in the channel pore also right-shifted the voltage dependence of slow gating, recapitulating mild myotonia. The recessive V640G mutant drastically reduced channel function, which probably accounts for myotonia. In contrast, the recessive L628P mutant produced currents very similar to wild type, suggesting that the occurrence of the compound truncating mutation (Q812X) or other muscle-specific mechanisms accounted for the severe symptoms observed in this family. Our results provide novel insight into the molecular mechanisms underlying normal and altered ClC-1 function.

Figure 1

Three dimensional representation of hClC-1 channel The representation is modelled upon the structure of CmClC showing the localization of the dominant and recessive MC mutations.

Figure 2

Functional characteristics of WT and F484L hClC-1 channels in high intracellular chloride A, representative chloride currents recorded from tsA cells transfected with hClC-1 WT and F484L variants. Cells were held at 0 mV and 400 ms voltage pulses were applied from −200 to +200 mV in 10 mV intervals every 3 s. Mutant channels displayed different kinetics and reduced amplitude compared to WT. B, the instantaneous currents were measured at the beginning of test voltage pulses, normalized with respect to cell capacitance (pA pF^–1), and reported as a function of voltage. The F484L mutant does not show the strong inward rectification typical of WT channels. C, steady-state currents were measured at the end of test voltage pulses and reported as mean current density ± SEM in function of voltage. The F484L mutant generated reduced current densities with respect to WT. D, the voltage dependence of activation was determined by plotting the apparent open probability (P_o), calculated from tail currents measured at −105 mV, as a function of test voltage pulses. The relationships obtained from averaged data were fitted with a Boltzmann equation, and fit parameters are reported in Table2. The mutant channels displayed a positively shifted voltage dependence of activation. Each point is the mean ± SEM from 6 to 14 cells.

Figure 3

Functional characteristics of WT and F484L hClC-1 channels in low intracellular chloride A, chloride currents were recorded in tsA cells transfected with WT and F484L hClC-1 variants. Cells were held at −95 mV and 400 ms voltage pulses were applied from −180 to +180 mV in 10 mV intervals every 3 s. Mutant chloride currents displayed slower kinetics of activation compared to WT. B, the steady-state current density–voltage relationships were drawn as in Fig. 2C. C, the voltage dependence of activation, determined as in Fig. 2D, was fitted with a Boltzmann function. Fit parameters are reported in Table2. The F484L mutant channels displayed a positively shifted voltage dependence. Each point is the mean ± SEM from 6 to 14 cells.

Figure 4

Apparent open probabilities for fast and slow gating of WT and F484L hClC-1 channels Apparent open probabilities for slow (A) and fast gating (B) in mutant F484L compared to WT. Open probability for slow gating was obtained as described in the Methods section. Open probability of the fast gates was calculated for a given test voltage by dividing the relevant total P_o by its corresponding P_o,slow. The substantial positive shifts in overall P_o for F484L shown in Fig. 2 was mainly due to a large positive shift in P_o,slow. For each channel, n = 6–8 cells. Fit parameters are reported in Table3.

Figure 5

Functional characteristics of heteromeric WT+F484L hClC-1 channels in high intracellular chloride A, representative current traces elicited from tsA cells co-transfected with equal amount of WT and F484L cDNAs in high intracellular chloride. B, the instantaneous currents in high intracellular chloride were measured as in Fig. 2B, for WT, F484L and WT+F484L channels. C, steady-state currents in high intracellular chloride were measured as described in Fig. 2C. D, the voltage dependence of activations, determined as in Fig. 2D, were fitted with a double Boltzmann function. Fit parameters are reported in Table4. The heteromeric channels displayed a positively shifted voltage dependence compared to WT. Each point is the mean ± SEM from 7 cells.

Figure 6

Zoom of the CmClC-based hClC-1 homology model used in the present study Important residues are shown as sticks and the H-bond between E232 and Y578 is shown as dashed line.

Figure 7

Analysis of hydrogen bond interactions occurring in WT and F484L dimers A and B, time-dependent evolutions of the distance between the oxygen donor of Y578 side-chain and the nearest oxygen acceptor of E232 ^side-chain, for WT and F484L channels. The results computed for both the two monomers (A and B) are presented. C and D, selected frames showing the different H-bond interactions established in WT and F484L in the EY domain. Important residues are shown as sticks while the H-bond interactions are depicted by a dashed line.

Figure 8

Functional characteristics of L198P, V640G and L628P hClC-1 channels in high intracellular chloride A, representative whole-cell current traces elicited from tsA cells transfected with the dominant L198P and recessive variants V640G or L628P. Cells were held at 0 mV and 400 ms voltage pulses were applied from −200 to +200 mV in 10 mV intervals every 3 s. Mutant channels displayed similar kinetics compared to WT. B, the instantaneous currents were measured as in Fig. 2B for WT, L198P, V640G and L628P. C, steady-state currents were measured as described in Fig. 2C. L198P and V640G generated reduced current densities with respect to WT. D, the voltage dependence of activation, determined as in Fig. 2D, was fitted with a Boltzmann function. Fit parameters are reported in Table2. The activation curve for V640G and L628P was similar to the WT curve while that of L198P was severely right-shifted. Each point is the mean ± SEM from 8–9 cells.

Figure 9

Functional characteristics of L198P and L628P hClC-1 channels in low intracellular chloride A, chloride currents were recorded in tsA cells transfected with L198P and L628P variant. Cells were held at −95 mV and 400 ms voltage pulses were applied from −180 to +180 mV in 10 mV intervals every 3 s. B, the steady-state current density–voltage relationships were drawn as in Fig. 2C for WT, L198P and L628P. C, the voltage dependence of activation, determined as in Fig. 2D, was fitted with a Boltzmann function. Fit parameters are reported in Table2. Each point is the mean ± SEM from 8 cells.

Figure 10

Open probabilities for fast and slow gating of WT, L198P, V640G and L628P hClC-1 channels, and functional characteristics of heteromeric WT+L198P hClC-1 channels Apparent open probabilities for slow (A) and fast (B) gating in mutants L198P, V640G and L628P compared to WT. Open probability for common gating (P_o,slow) was obtained as described in the Methods section. Open probability of the fast gates (P_o,fast) was calculated for a given test voltage by dividing the relevant P_o by its corresponding P_o,slow. For each channel, n = 4–6 cells. Fit parameters are reported in Table3. C, representative current traces elicited from tsA cells co-transfected with equal amounts of WT and L198P cDNAs in high intracellular chloride. D, the instantaneous currents in high intracellular chloride were measured as in Fig. 8B, for WT, L198P and WT+L198P channels. E, steady-state currents in high intracellular chloride were measured as described in Fig. 8C. F, the voltage dependence of activations, determined as in Fig. 8D, were fitted with a double Boltzmann function. Fit parameters are reported in Table4. The heteromeric channels displayed a positively shifted voltage dependence compared to WT. Each point is the mean ± SEM from 6 cells.