Genetic activation of BK currents in vivo generates bidirectional effects on neuronal excitability

Abstract

Large-conductance calcium-activated potassium channels (BK) are potent negative regulators of excitability in neurons and muscle, and increasing BK current is a novel therapeutic strategy for neuro- and cardioprotection, disorders of smooth muscle hyperactivity, and several psychiatric diseases. However, in some neurons, enhanced BK current is linked with seizures and paradoxical increases in excitability, potentially complicating the clinical use of agonists. The mechanisms that switch BK influence from inhibitory to excitatory are not well defined. Here we investigate this dichotomy using a gain-of-function subunit (BK(R207Q)) to enhance BK currents. Heterologous expression of BK(R207Q) generated currents that activated at physiologically relevant voltages in lower intracellular Ca(2+), activated faster, and deactivated slower than wild-type currents. We then used BK(R207Q) expression to broadly augment endogenous BK currents in vivo, generating a transgenic mouse from a circadian clock-controlled Period1 gene fragment (Tg-BK(R207Q)). The specific impact on excitability was assessed in neurons of the suprachiasmatic nucleus (SCN) in the hypothalamus, a cell type where BK currents regulate spontaneous firing under distinct day and night conditions that are defined by different complements of ionic currents. In the SCN, Tg-BK(R207Q) expression converted the endogenous BK current to fast-activating, while maintaining similar current-voltage properties between day and night. Alteration of BK currents in Tg-BK(R207Q) SCN neurons increased firing at night but decreased firing during the day, demonstrating that BK currents generate bidirectional effects on neuronal firing under distinct conditions.

Fig. 1.

BK^R207Q subunits generate gain-of-function currents. (A) BK α pore-forming subunit with R207Q mutation (red circle). (B) Voltage of half-maximal activation (V_1/2) as a function of [Ca²⁺]_i. Currents were recorded from inside-out patches from HEK293 cells transfected with cDNA encoding the BK^WT subunit, BK^R207Q, or both, in 5 mM K⁺_out/140 K⁺_in at the indicated Ca²⁺_i. V_1/2 values were derived from Boltzmann fits of the conductance–voltage (G–V) relationships of the macroscopic currents. BK^R207Q currents could be obtained only in 0 Ca²⁺_i (Table S1). (C) BK^R207Q and BK^WT/R207Q currents are fast activating relative to BK^WT currents. Activation kinetics (τ_act) were determined by single exponential fit of the current rise. Data points are mean ± SEM with n in parentheses in C.

Fig. 2.

Tg-BK^R207Q mice exhibit increased BK channel expression across tissues. (A) Tg-BK^R207Q transgenic construct containing the 6.8-kb mouse Period1 (mPer1) regulatory sequence driving expression of the BK^R207Q cDNA. βg pA, β-globin polyadenylation sequence. (B) Normalized total BK expression from Western blots of brain (4 μg) and smooth muscle (10 μg) using an antibody that recognizes both endogenous BK^WT and BK^R207Q subunits. BK levels in hemizygous (light gray) and homozygous (dark gray) Tg-BK^R207Q tissues were normalized to levels in WT (black) tissues. Hip, hippocampus; Ctx, cortex; Cb, cerebellum; H/T/M, hypothalamus/thalamus/midbrain; Ao, aorta; Bl, bladder. Data points are mean ± SEM from three animals.

Fig. 3.

Tg-BK^R207Q expression in the SCN generates similar fast-activating BK currents under two membrane conditions, night and day. (A) Schematic illustrating the expected temporal pattern of endogenous BK^WT expression (black line) (29) and Per1-driven BK^R207Q expression (dotted line) (33). (B) Western blot analysis showing BK expression in SCN, normalized to DM1α (Fig. S3A). Protein was isolated at time points indicated by arrows in A (day, 4–8 h after lights on; night, 4–8 h after lights off). Total BK expression was high at night and low during the day in WT SCNs. In contrast, total BK expression was high during both night and day in Tg-BK^R207Q SCNs, which express both endogenous BK^WT and Per1-driven BK^R207Q. (C) BK current density at +30 mV from voltage-clamp recordings of SCN neurons. BK currents were isolated with 10 μM paxilline (Fig. S4). (D) BK current at each voltage normalized to the maximal current (I/I_max) for WT and Tg-BK^R207Q neurons. The V_1/2 values did not differ (P = 0.13, ANOVA). (E) I/I_max plots for the largest versus smallest daytime BK currents isolated from Tg-BK^R207Q neurons (P = 0.43, ANOVA). (F) Activation of BK currents in Tg-BK^R207Q neurons was faster than BK currents in WT neurons for both night and day currents. τ_act was determined from single exponential fit of the current rise at +30 mV. (Inset) Average nighttime BK current rising phase for WT (black) and Tg-BK^R207Q (gray) neurons. [Scale bar: x axis, 5 ms; y axis (WT), 600 pA; Tg-BK^R207Q, 400 pA.] All data points are mean ± SEM, with n in parentheses. P values are indicated for pairwise comparisons (t tests) in B and C.

Fig. 4.

BK currents from Tg-BK^R207Q neurons induce bidirectional effects on action potential frequency. (A) Spontaneous action potential activity in the SCN at night. Histogram shows the percent of total events at each instantaneous frequency; the arrow denotes a second high-frequency peak absent from WT neuronal recordings. WT (median, 6 Hz), Tg-BK^R207Q (median₁, 4 Hz; median₂ 14 Hz), P = 0.01, Kolmogorov–Smirnov test. (B) Representative spontaneous action potential waveforms from nighttime recordings. (C) Spontaneous action potential activity during the day; arrow denotes a higher-frequency WT peak that was absent from Tg-BK^R207Q neuronal recordings. WT (median₁, 7.5 Hz; median₂, 32 Hz), Tg-BK^R207Q (median, 8 Hz), P = 0.002, Kolmogorov–Smirnov test. (D) Representative spontaneous action potential waveforms from day. n indicated in parentheses.