Functional Effects of Epilepsy Associated KCNT1 Mutations Suggest Pathogenesis via Aberrant Inhibitory Neuronal Activity

Abstract

KCNT1 (K⁺ channel subfamily T member 1) is a sodium-activated potassium channel highly expressed in the nervous system which regulates neuronal excitability by contributing to the resting membrane potential and hyperpolarisation following a train of action potentials. Gain of function mutations in the KCNT1 gene are the cause of neurological disorders associated with different forms of epilepsy. To gain insights into the underlying pathobiology we investigated the functional effects of 9 recently published KCNT1 mutations, 4 previously studied KCNT1 mutations, and one previously unpublished KCNT1 variant of unknown significance. We analysed the properties of KCNT1 potassium currents and attempted to find a correlation between the changes in KCNT1 characteristics due to the mutations and severity of the neurological disorder they cause. KCNT1 mutations identified in patients with epilepsy were introduced into the full length human KCNT1 cDNA using quick-change site-directed mutagenesis protocol. Electrophysiological properties of different KCNT1 constructs were investigated using a heterologous expression system (HEK293T cells) and patch clamping. All mutations studied, except T314A, increased the amplitude of KCNT1 currents, and some mutations shifted the voltage dependence of KCNT1 open probability, increasing the proportion of channels open at the resting membrane potential. The T314A mutation did not affect KCNT1 current amplitude but abolished its voltage dependence. We observed a positive correlation between the severity of the neurological disorder and the KCNT1 channel open probability at resting membrane potential. This suggests that gain of function KCNT1 mutations cause epilepsy by increasing resting potassium conductance and suppressing the activity of inhibitory neurons. A reduction in action potential firing in inhibitory neurons due to excessively high resting potassium conductance leads to disinhibition of neural circuits, hyperexcitability and seizures.

Keywords: K+ channels; channelopathies; epilepsy; gain-of-function mutations; patch clamping.

Figure 1

Current-voltage plots of WT and mutant KCNT1 channels expressed in HEK293T cells. (A) I-V plots recorded in response to 100 ms ramps ranging from −120 to 120 mV in a cell transfected with WT KCNT1 and an untransfected cell. (B). Representative I-V plots of KCNT1 channels carrying GoF mutations. (C). The same data as on panel B, shown at a different scale. (D). WT KCNT1 currents recorded in inside-out patch in the presence of 10 mM or 60 mM intracellular Na⁺.

Figure 2

The effect of GoF mutations on KCNT1 current amplitude. The average amplitudes of WT and GoF mutant KCNT1 currents, measured at +10 mV using I-V plots similar to those shown in Figure 1. Brown-Forsythe and Welch’s ANOVA tests with multiple comparisons produced significant difference between the amplitude of WT and each mutant current, except T314A and N449S. Calculated individual p values were as follows: G228S—0.0009; T314A—0.6136; R398Q—0.0022; N449S—0.0535; L781V—0.0026; E893K—0.0224; M896V—0.0001; F932L—0.0001; R928C—0.0004; S937G—<0.0001; L942F—0.0049; R961A—0.0024; A965T—<0.0001. The dots on the graph represent current amplitudes in individual cells; the asterisks denote the level of the significance (*—<0.05; **—<0.01; ***—<0.001; ****—<0.0001); n.s.—not significant. (Note: The amplitudes of the largest currents are underestimated due to a residual uncompensated series resistance. See Section 4).

Figure 3

The effect of GoF mutations of the kinetics of KCNT1 currents. WT and Mutant KCNT1 currents were recorded in response to the voltage protocol shown in the lower right-hand corner.

Figure 4

The effect of GoF mutations on the relative open probability of KCNT1 channels. Apparent P_o curves of WT and mutant KCNT1 channels were obtained from tail currents normalised to the instantaneous tail current amplitude corresponding to a voltage step to 100 mV using recordings similar to those shown in Figure 3. Each data set was fitted with the Boltzmann equation (Section 4, Equation (1)).

Figure 5

T314A and M896V GoF mutations do not alter KCNT1 pore permeability to Na⁺. (A). WT KCNT1. (B). T314A KCNT1. (C). M896V KCNT1. The I-V plots were recorded in response to 100 ms voltage ramps from −120 mV to 120 mV in the control bath solution (4 mM K⁺) and in a solution with 126 mM NaCl replaced with 126 mM KCl. (D). The shifts in the membrane potential caused by replacing NaCl with KCl were calculated using I-V plots similar to those shown in panels (A–C) and plotted against the KCl concentration changes. The solid curves are the fits of GHK equation (Equation (2), Section 4) to the experimental data.

Figure 6

Dependence of WT and mutant KCNT1 currents on intracellular Na⁺ concentration. (A). WT KCNT1. (B). T314A KCNT1. (C). M896V KCNT1. The I-V plots (i) were recorded in response to 100 ms voltage ramps from −120 mV to 120 mV in the bath solution with 130 mM K⁺ and using either control pipette solution or pipette solution with 130 mM K⁺ replaced with 130 mM Na⁺. The amplitudes of the inward K⁺ currents recorded at −100 mV using low (10 mM) and high (130 mM) intracellular Na⁺ are compared on the panels below (ii). *—<0.05; ****—<0.0001.

Figure 7

Membrane currents recorded in inside-out patches from the cells expressing T314A or WT KCNT1. (A). T314A KCNT1 current traces recorded in response to 3 s voltage steps ranging from −60 mV to 60 mV (i) or a voltage ramp from −120 mV to 120 mV (ii). (B). WT KCNT1 current trace recorded in response to 3 s voltage step to −40 mV (i) or a voltage ramp from −120 mV to 120 mV (ii). All traces were recorded using symmetrical K⁺ (130 mM) and Na⁺ (20 mM) concentrations in the bath and the pipette solutions.

Figure 8

Distribution of the studied mutations on the 3D model of the KCNT1 channel. Mutations analysed in this paper were mapped onto the 3D structure of the chicken KCNT1 homolog, Slo2_2 in the open conformation deposited in the rcsb.org Protein Data Bank (PDB, http://doi.org/10.2210/pdb5U70/pdb, accessed on 3 November 2022) [20,21], with colours corresponding to the functional domains they are located in (see Section 4). The projection of a single KCNT1 subunit shown in panel (B) is obtained by anticlockwise rotation of the projection shown in panel (A) by 90 degrees around the vertical axis.