Nonfunctional NaV1.1 familial hemiplegic migraine mutant transformed into gain of function by partial rescue of folding defects

Abstract

Familial hemiplegic migraine (FHM) is a rare subtype of migraine with aura. Mutations causing FHM type 3 have been identified in SCN1A, the gene encoding the Nav1.1 Na(+) channel, which is also a major target of epileptogenic mutations and is particularly important for the excitability of GABAergic neurons. However, functional studies of NaV1.1 FHM mutations have generated controversial results. In particular, it has been shown that the NaV1.1-L1649Q mutant is nonfunctional when expressed in a human cell line because of impaired plasma membrane expression, similarly to NaV1.1 mutants that cause severe epilepsy, but we have observed gain-of-function effects for other NaV1.1 FHM mutants. Here we show that NaV1.1-L1649Q is nonfunctional because of folding defects that are rescuable by incubation at lower temperatures or coexpression of interacting proteins, and that a partial rescue is sufficient for inducing an overall gain of function because of the modifications in gating properties. Strikingly, when expressed in neurons, the mutant was partially rescued and was a constitutive gain of function. A computational model showed that 35% rescue can be sufficient for inducing gain of function. Interestingly, previously described folding-defective epileptogenic NaV1.1 mutants show loss of function also when rescued. Our results are consistent with gain of function as the functional effect of NaV1.1 FHM mutations and hyperexcitability of GABAergic neurons as the pathomechanism of FHM type 3.

Keywords: Dravet syndrome; ankyrin; calmodulin; generalized epilepsy with febrile seizures plus; spreading depression.

Fig. 1.

hNa_V1.1-L1649Q is a rescuable folding-defective mutant. (A) (Left) Mean maximum current density in tsA-201 cells transfected with WT, L1649Q, or mock transfected and maintained at 37 °C (269 ± 52 pA/pF, n = 7; 15.1 ± 1.5 pA/pF, n = 11, P < 0.01; 5.9 ± 3.0 pA/pF, n = 5, P < 0.01) or incubated at 30 °C (336 ± 124 pA/pF, n = 17; 155 ± 26 pA/pF, n = 25; P < 0.01; 6.5 ± 2.7 pA/pF, n = 6, P < 0.01). (Right) Fold increase in current density with incubation at 30 °C (P < 0.01 for L1649Q). (B) Representative whole-cell Na⁺ currents recorded in cells incubated at 30 °C applying depolarizing steps from −65 to +90 mV in 5-mV increments, from a holding potential of −100 mV. Calibration: 2 nA, 2 ms. (C) Normalized current–voltage plots for WT and L1649Q. (D) Mean current density in tsA-201 cells maintained at 37 °C and transfected with L1649Q (15.1 ± 1.5 pA/pF, n = 11); L1649Q and β1 (16.5 ± 6.1 pA/pF, n = 15); L1649Q and β2 (11.8 ± 4.9 pA/pF, n = 6); L1649Q, β1 and β2 (12.1 ± 6.0 pA/pF, n = 5); L1649Q and ankyrin G (66 ± 16 pA/pF, n = 8; P < 0.01); L1649Q and calmodulin (37.4 ± 7.5 pA/pF, n = 19; P < 0.05); and L1649Q, β1, and calmodulin (36.7 ± 6.9 pA/pF, n = 7; P < 0.05). Data presented as mean ± SEM. *P < 0.05, **P < 0.01.

Fig. 2.

Functional properties of hNa_V1.1-L1649Q rescued with incubation at 30 °C in tsA-201 cells: activation and fast inactivation. (A) Average normalized current for WT (n = 15) and L1649Q (n = 21) elicited with steps to −10 mV from a holding potential of −100 mV (error bars for selected data points). (Scale bar: 0.5 ms.) (B) Times of half-activation at the indicated potentials (P < 0.01 for all of the data). (C) Voltage dependence of the time constant (τ) of current decay (exponentials fits at the indicated potentials); P < 0.01 for all potentials. (D) Mean voltage dependence of activation and fast inactivation; lines are mean Boltzmann fits; mean parameters: activation (voltage of half activation, Va, and slope, Ka), WT, Va = −28.9 ± 0.1 mV, Ka = 6.6 ± 0.1 mV (n = 17), L1649Q, Va = −25.2 ± 0.1 mV, P < 0.05; Ka = 7.3 ± 0.1 mV, P < 0.05 (n = 25), inactivation (voltage of half inactivation Vh, and slope, Kh), WT V_h = −63.8 ± 0.2 mV, K_h = 6.8 ± 0.1 mV (n = 17), L1649Q V_h = −43.3 ± 0.4 mV, P < 0.01; K_h = 8.7 ± 0.4 mV, P < 0.01 (n = 23). (Left Inset) τ for recovery from a 150-ms inactivating pulse at the indicated potentials: −110 mV, τ_REC-WT = 2.1 ± 0.1 ms (n = 5), τ_REC-L1649Q = 1.0 ± 0.1 ms (n = 6); −100 mV, τ_REC-WT = 2.8 ± 0.2 ms (n = 5), τ_REC-L1649Q = 1.3 ± 0.1 ms (n = 5); −90 mV, τ_REC-WT = 3.98 ± 0.04 ms (n = 7), τ_REC-L1649Q = 1.45 ± 0.14 ms (n = 8); −80 mV τ_REC-WT = 6.6 ± 1.6 ms (n = 5), τ_REC-L1649Q = 2.5 ± 0.2 ms (n = 5); −70 mV, τ_REC-WT =13.1 ± 3.5 ms (n = 4), τ_REC-L1649Q = 4.6 ± 0.5 ms (n = 5); −60 mV, τ_REC-L1649Q = 9.3 ± 1.2 ms (n = 5); P < 0.01 or <0.05 for all of the potentials. (Right Inset) τ of development of fast inactivation at the indicated potentials: −60 mV τ_DEV-WT = 38.0 ± 4.1 ms (n = 8); −50 mV, τ_DEV-WT = 14.0 ± 2.0 ms (n = 8), τ_DEV-L1649Q = 13.8 ± 2.0 ms (n = 9); −40 mV, τ_DEV-WT = 4.7 ± 0.7 ms (n = 11), τ_DEV-L1649Q = 9.9 ± 0.5 ms (n = 16); −30 mV, τ_DEVWT = 0.19 ± 0.17 ms (n = 3), τ_DEV-L1649Q = 2.35 ± 0.46 ms (n = 5); P < 0.01 or <0.05 for all of the potentials. (E) Same average normalized currents as in A, shown enlarged and for a duration of 70 ms. (Left Inset) Mean current–voltage plots for I_NaP after 5 min (I_NaP-max WT 2.2 ± 0.3%, L1649Q, 8.3 ± 0.8%; P < 0.01). (Right Inset) Mean current–voltage plots for I_NaP after 15 min (I_NaP-max WT 1.1 ± 0.1%, L1649Q, 6.9 ± 0.7%; P < 0.01); dash-dot and solid lines are the calculated window currents for L1649Q and WT respectively. Data presented as mean ± SEM.

Fig. 3.

Functional properties of hNa_V1.1-L1649Q expressed in neocortical neurons. (A) Representative whole-cell Na⁺ current traces recorded in the presence of TTX 1 μM with steps from −60 mV to +20 mV in 10-mV increments (holding potential −100 mV) for WT-F383S and L1649Q-F383S. (Scale bars: 200 pA, 10 ms.) (Insets) First 11 ms of the traces. (B) (Upper) Average normalized current for WT-F383S (solid, n = 9) and L1649Q-F383S (dash-dot, n = 10) elicited with steps to −10 mV (holding potential of −100 mV); error bars are the SEM of selected data points. (Scale bar: 1 ms.) (Lower Left) Times of half-activation at the indicated potentials (P < 0.01 for all potentials). (Lower Right) Voltage dependence of τ decay obtained from fits of exponentials to the decay of the current traces at the indicated potentials (P < 0.01 for all of the potentials). (C) Current density–voltage plots for WT-F383S and L1649Q-F383S. (D) Mean voltage dependence of activation and fast inactivation; lines are mean Boltzmann fits: mean parameters, WT-F383S-activation (V_a = −21.0 ± 0.5 mV, K_a = 6.6 ± 0.5 mV, n = 9); L1649Q-F383S-activation (V_a = −22.9 ± 0.4 mV; K_a = 7.4 ± 0.6 mV; n = 10); WT-F383S-inactivation (V_h = −54.2 ± 0.5 mV, K_h = 5.5 ± 0.4 mV, n = 9); L1649Q-F383S-inactivation (V_h = −34.5 ± 0.8 mV, P < 0.01; K_h = 7.1 ± 0.6 mV, P < 0.01; baseline 0.16 ± 0.04, P < 0.01; n = 10). (E) (Upper) Same traces as in B Upper shown enlarged and for a duration of 150 ms. (Lower) Current–voltage plots for I_NaP recorded after 5 min of whole-cell configuration: I_NaP-max WT-F383S 4.9 ± 0.8% (n = 9), L1649Q-F383S, 20.5 ± 2.5% (n = 10, P < 0.01); calculated window currents: dash-dot (L1649Q) and solid lines (WT). (F) Mean whole-cell TTX-resistant action-Na⁺ currents recorded using an AP discharge as voltage stimulus (Top) and normalized to the maximal current of the I–V plot of each cell, WT-F383S, n = 7; L1649Q-F383S, n = 8. (Scale bar: 20 ms.) First current: WT-F383S (0.77 ± 0.08), L1649Q-F383S (0.93 ± 0.03; P < 0.01); second: WT-F383S (0.23 ± 0.03), L1649Q-F383S (0.79 ± 0.04; P < 0.01); 20th: WT-F383S (0.24 ± 0.05), L1649Q-F383S (0.77 ± 0.05; P < 0.01). Data presented as mean ± SEM.

Fig. 4.

Effect of hNa_V1.1-L1649Q on the firing of transfected neurons. (A) Representative firing traces for WT (Left) or L1649Q (Right) during injections of 400-ms-long depolarizing current steps (holding potential −65 mV): 50 pA for the top trace with 10-pA increments (Left) and 30 pA for the top trace with 10-pA increments (Right). Dotted lines indicate the 0-mV level for each trace. Calibration: 40 mV, 100 ms. (B) Input–output plot of number of APs vs. injected depolarizing current (P < 0.05 or P < 0.01 for all of the data besides the first three points). (C) Input–output plot of number of APs elicited in 500 ms vs. injected depolarizing current in a computational model of a simplified neuron; traces correspond to different amount of WT and L1649Q current, indicated in the plot as % of maximal conductance (mS/cm²); the control WT condition is WT 0.2 mS/cm² (200%) and L1649Q 0 mS/cm² (0%); the hypothetical heterozygous condition with L1649Q fully rescued is WT 0.1 mS/cm² (100%) and L1649Q 0.1 mS/cm2 (100%). (D) Plot of the maximal number of APs generated by the model (displayed as a percent of variation in comparison with the control WT condition) vs. the amount of L1649Q conductance (displayed as a percent of variation in comparison with the hypothetical heterozygous condition with L1649Q fully rescued). (E) Plot of the rheobase obtained with the model (displayed as a percent of variation in comparison with the control WT condition) vs. the amount of L1649Q conductance (displayed as in D).