Differential contribution of Ca2+ sources to day and night BK current activation in the circadian clock

Abstract

Large conductance K⁺ (BK) channels are expressed widely in neurons, where their activation is regulated by membrane depolarization and intracellular Ca²⁺ (Ca²⁺_i). To enable this regulation, BK channels functionally couple to both voltage-gated Ca²⁺ channels (VGCCs) and channels mediating Ca²⁺ release from intracellular stores. However, the relationship between BK channels and their specific Ca²⁺ source for particular patterns of excitability is not well understood. In neurons within the suprachiasmatic nucleus (SCN)-the brain's circadian clock-BK current, VGCC current, and Ca²⁺_i are diurnally regulated, but paradoxically, BK current is greatest at night when VGCC current and Ca²⁺_i are reduced. Here, to determine whether diurnal regulation of Ca²⁺ is relevant for BK channel activation, we combine pharmacology with day and night patch-clamp recordings in acute slices of SCN. We find that activation of BK current depends primarily on three types of channels but that the relative contribution changes between day and night. BK current can be abrogated with nimodipine during the day but not at night, establishing that L-type Ca²⁺ channels (LTCCs) are the primary daytime Ca²⁺ source for BK activation. In contrast, dantrolene causes a significant decrease in BK current at night, suggesting that nighttime BK activation is driven by ryanodine receptor (RyR)-mediated Ca²⁺_i release. The N- and P/Q-type Ca²⁺ channel blocker ω-conotoxin MVIIC causes a smaller reduction of BK current that does not differ between day and night. Finally, inhibition of LTCCs, but not RyRs, eliminates BK inactivation, but the BK β2 subunit was not required for activation of BK current by LTCCs. These data reveal a dynamic coupling strategy between BK channels and their Ca²⁺ sources in the SCN, contributing to diurnal regulation of SCN excitability.

Figure 1.

BK current activation requires Ca²⁺ influx in SCN neurons. Macroscopic voltage-activated currents were recorded in whole-cell voltage-clamp mode from WT SCN neurons during the day (4–8 h) or night (17–21 h). The pipette (internal) solution contained 0.9 mM EGTA, which allows stable recordings in the presence of Ca²⁺ influx from endogenous channels (Jackson et al., 2004; Fakler and Adelman, 2008; Montgomery et al., 2013; Whitt et al., 2016). From a holding potential of −90 mV, currents were elicited from 150-ms voltage steps in 20-mV increments. BK currents were isolated by subtracting currents elicited in 10 µM Pax from baseline. (A) Representative daytime and nighttime BK currents. The daytime current is BK_i (τ_inact < 110 ms), and the nighttime current is noninactivating BK_s (τ_inact > 110 ms). (B) BK current versus voltage relationship in day and night SCN neurons and under Ca²⁺ channel inhibition. BK current levels were quantified from peak values. The Ca²⁺ cocktail contained 10 µM Nim, 10 µM Dan, and 3 µM MVIIC. Data are mean ± SE. *, BK current value at 90 mV: One-way ANOVA with Bonferroni post hoc (day and night versus cocktail: P = 10⁻¹⁵ and 10⁻¹⁵, respectively). Day, n = 24 neurons; night, n = 25; day (Ca²⁺ cocktail), n = 10; night (Ca²⁺ cocktail), n = 6.

Figure 2.

VGCCs in SCN neurons. (A) Representative daytime and nighttime Ca²⁺ currents from WT neurons, elicited with the same voltage protocol (in 10-mV increments) used to record BK currents in Fig. 1. (B and C) Peak Ca²⁺ current versus voltage relationships before application of 10 µM Nim (total currrent) and the Nim-sensitive current. (D) Day-versus-night relationship for the total Ca²⁺ current. (E) Day-versus-night relationship for the L-type (Nim-sensitive) Ca²⁺ current. *, One-way ANOVA with Bonferroni post hoc (P = 0.04). Day, n = 13 neurons; night, n = 9. All data are mean ± SE.

Figure 3.

Effect of inhibition of Ca²⁺ channels on BK currents during the day. (A) Representative daytime peak BK currents recorded from WT SCN neurons in control conditions or with 10 µM Nim, 10 µM Dan, 5 µM TG, or 3 µM MVIIC. Voltage protocol was applied as in Fig. 1. Traces truncated at 50 ms. (B) Peak daytime BK current at 90 mV was reduced in Nim, Dan, and MVIIC compared with control. *, One-way ANOVA with Bonferroni post hoc (P = 10⁻¹³, 0.04, and 0.04, respectively). At 30 mV, only Nim significantly reduced the BK current (P = 10⁻⁶). (C) The percentage of neurons with a BK current was decreased with Nim. *, P = 10⁻⁶, Fisher’s exact test. (B and C) Control, n = 24 neurons; Nim, n = 14; Dan, n = 17; TG, n = 10; and MVIIC, n = 10. (D) Peak current-versus-voltage relationship for only those cells exhibiting a BK current in the presence of each inhibitor. *, One-way ANOVA with Bonferroni post hoc (control vs. Nim at 90 mV: P = 10⁻⁸). At 30 mV, only Nim significantly reduced the BK current (P = 0.006). (E) Voltage of half-maximal activation values (V_1/2) from fits of I/I_max relationship of only those cells exhibiting a BK current. No significant differences were obtained in any conditions versus control (one-way ANOVA, P = 0.08). (D and E) Control, n = 19 neurons; Nim, n = 7; Dan, n = 17; TG, n = 10; and MVIIC, n = 4. All data are mean ± SE.

Figure 4.

Effect of inhibition of Ca²⁺ channels on BK currents at night. (A) Representative nighttime BK currents in control, Nim, Dan, TG, and MVIIC as in Fig. 3. (B) Peak BK current at 90 mV was reduced with Dan and TG compared with control. *, One-way ANOVA with Bonferroni post hoc (control vs. Dan at 90 mV: P = 10⁻⁴; TG: P = 10⁻⁴). At 30 mV, Dan and TG both significantly reduced the BK current (P = 10⁻⁵ and 10⁻⁴, respectively), but Nim did not. (C) The percentage of neurons with a BK current was decreased with Dan and TG versus control. *, Fisher’s exact test, Dan: P = 0.03; TG: P = 0.02. There was no difference with the other inhibitors versus control. (B and C) Control, n = 25 neurons; Nim, n = 17; Dan, n = 18; TG, n = 10; and MVIIC, n = 9. (D) Peak current-versus-voltage relationship for only those cells exhibiting a BK current in the presence of each inhibitor. *, One-way ANOVA with Bonferroni post hoc (control vs. Dan at 90 mV: P = 10⁻⁴; TG: P = 10⁻³). At 30 mV, Dan and TG both significantly reduced the BK current (P = 10⁻⁵ and 10⁻⁴, respectively), but Nim did not. (E) V_1/2 values. No significant differences were obtained in any conditions versus control (one-way ANOVA, P = 0.22). Control, n = 25 neurons; Nim, n = 17; Dan, n = 13; TG, n = 7; and MVIIC, n = 7. All data are mean ± SE.

Figure 5.

Effect of agonist activation of Ca²⁺ channels on BK currents. (A) Representative BK currents recorded in control, 5 µM BayK, and 100 nM Ryan. Voltage protocol same as in Fig. 1. (B) Peak current density versus voltage from daytime WT SCN neurons. BayK and Ryan each produced an increase BK current magnitude. *, One-way ANOVA with Bonferroni post hoc test at 90 mV: control (n = 24) versus BayK, P = 0.002 (n = 8) and Ryan, P = 0.017 (n = 7). At 30 mV, BayK and Ryan both significantly increased the BK current (P = 0.001 and 0.001, respectively). (C) V_1/2 values from I/I_max relationships. BayK and Ryan produce larger currents via a shift in the voltage dependence of activation to more hyperpolarized membrane potentials. *, One-way ANOVA with Bonferroni post hoc test: control versus BayK, P = 10⁻³ and Ryan, P = 10⁻³. (D) Nighttime peak current density versus voltage. Only Ryan produced larger BK currents. *, One-way ANOVA with Bonferroni post hoc test at 90 mV: control versus BayK, P = 0.99 (n = 17) and Ryan, P = 0.05 (n = 8). At 30 mV, only Ryan significantly increased the BK current (P = 0.03). (E) V_1/2 values, with only Ryan producing larger BK current via a shift to more hyperpolarized membrane potentials. *, One-way ANOVA with Bonferroni post hoc test: Control versus BayK, P = 0.99 and Ryan, P = 0.03. All data are mean ± SE.

Figure 6.

Effect of inhibition and activation of Ca²⁺ channels on BK current inactivation. BK currents from daytime WT SCN neurons were categorized as BK_i or BK_s as described previously (Whitt et al., 2016). (A) Percentage of cells with inactivating BK currents in control (n = 24), Nim (n = 14), Dan (n = 17), TG (n = 10), and MVIIC (n = 10). Nim, but not the other inhibitors, abolished BK_i currents during the day. *, P < 0.05, Fisher’s exact test: control versus Nim, P = 10⁻⁴. (B) Percentage of cells with BK_i and BK_s currents from day and night WT SCN neurons in BayK and Ryan. BayK increased the number of BK_i neurons compared with control during day and night, but Ryan had no effect. *, P = 0.04 (day) and 0.039 (night), control versus BayK, Fisher's exact test. Control (day, n = 24; night, n = 25), BayK (day, n = 8; night, n = 17), and Ryan (day, n = 7; night, n = 8).

Figure 7.

Effect of inhibition of Ca²⁺ channels on BK currents from β2 KO SCN neurons. (A–C) BK currents recorded from β2 KO SCN neurons during the day. (A) Representative BK current traces in control (n = 22), 10 µM Nim (n = 18), and 10 µM Dan (n = 13). (B) The percentage of cells with a detectable BK current was reduced with Nim and Dan. *, Fisher's exact test (control vs. Nim, P = 0.01 and Dan, P = 0.04). (C) Peak daytime BK current density versus voltage from only those cells with a BK current in β2 KO control (n = 22), Nim (n = 13), Dan (n = 10), or a cocktail of 10 µM Nim, 10 µM Dan, and 3 µM MVIIC (n = 4). In β2 KO neurons, both Nim and Dan decreased BK current magnitude compared with control. Unlike with WT neurons (Fig. 3 D), the Ca²⁺ cocktail did not completely inhibit all BK current. *, One-way ANOVA with Bonferroni post hoc test at 90 mV: control versus Nim, P = 10⁻³; Dan, P = 10⁻³; and Ca²⁺ cocktail, P = 10⁻⁴). At 30 mV, only Nim significantly decreased the BK current (P = 0.03). (D–F) BK currents recorded from β2 KO SCN neurons during the night. (D) Representative BK current traces. (E) The percentage of cells with a detectable BK current was decreased with Dan at night. *, Fisher's exact test (control vs. Dan, P = 10⁻⁴). β2 KO Control, n = 17; Dan, n = 8; Nim, n = 12; and Ca²⁺ cocktail, n = 5. (F) Peak daytime BK current density versus voltage from only those cells with a BK current in control (n = 17), Nim (n = 12), Dan (n = 5), or cocktail (n = 5). Dan, but not Nim, produced a large decrement in BK current at night, similar to WT neurons (Fig. 4 D). *, One-way ANOVA with Bonferroni post hoc test at 90 mV: control vs. Dan, P = 10⁻⁴ or Ca²⁺ cocktail, P = 10⁻⁴. At 30 mV, Dan and the Ca²⁺ cocktail, but not Nim, significantly decreased the BK current (P = 0.001 and 0.002, respectively). All data are mean ± SE.

Figure 8.

Effect of inhibition and activation of Ca²⁺ channels on firing rates in SCN neurons. (A and B) Spontaneous action potential activity from representative daytime (A) and nighttime (B) WT SCN neurons in 10 µM Nim, 10 µM Dan, 5 µM BayK, or 100 nM Ryan. Dotted line denotes −50 mV. (C) During the day, both Nim (n = 8) and Ryan (n = 7) decreased the firing rate compared with the control (n = 8). Dan and BayK had no effect on daytime firing. However, at night, Dan increased firing (n = 7) compared with the control (n = 13), whereas Nim (n = 8), BayK (n = 7), and Ryan (n = 8) had no effect. *, One-way ANOVA with Bonferroni post hoc: control versus Nim (day), P = 0.02; Ryan (day), ‡, P = 0.04; or control versus Dan (night), P = 10⁻³. All data are mean ± SE.