Circadian regulation of dentate gyrus excitability mediated by G-protein signaling

Abstract

The central circadian regulator within the suprachiasmatic nucleus transmits time of day information by a diurnal spiking rhythm driven by molecular clock genes controlling membrane excitability. Most brain regions, including the hippocampus, harbor similar intrinsic circadian transcriptional machinery, but whether these molecular programs generate oscillations of membrane properties is unclear. Here, we show that intrinsic excitability of mouse dentate granule neurons exhibits a 24-h oscillation that controls spiking probability. Diurnal changes in excitability are mediated by antiphase G-protein regulation of potassium and sodium currents that reduce excitability during the Light phase. Disruption of the circadian transcriptional machinery by conditional deletion of Bmal1 enhances excitability selectively during the Light phase by removing G-protein regulation. These results reveal that circadian transcriptional machinery regulates intrinsic excitability by coordinated regulation of ion channels by G-protein signaling, highlighting a potential novel mechanism of cell-autonomous oscillations.

Keywords: Bmal1; CP: Neuroscience; G-protein; GIRK; NALCN; circadian rhythms; dentate gyrus; excitability; granule cell; intrinsic properties.

Figure 1.. Reduced recruitment of GCs during the Light phase compared to the Dark phase

(A) Current-clamp recordings (I = 0) in response to stimulation of the perforant path were compared in two recording windows: ZT 8–11 (Light) and ZT 14–17 (Dark). (B) Left, example EPSPs at increasing stimulus intensities. Right, the percentage of GCs recruited to spike was reduced during the Light phase. (C) Gabazine (GBZ; 10 μM) increased spiking during both the Light (left) and Dark (middle) phases. Right, GBZ, a smaller percentage of GCs were recruited during the Light. (B and C) χ² tests, *p < 0.05; **p < 0.01; ***p < 0.001. (D) Depolarization by subthreshold EPSPs was reduced during the Light (in GBZ). Inset, EPSPs evoked at 100 μA. (E) EPSP half-widths were reduced during the Light in control (left) and GBZ (right). (D and E) Multiple unpaired t test comparison by the Holm-Sidak method. *p < 0.05; **p < 0.01; ***p < 0.001. (B–E) n = 19, 19. Symbols represent mean ± SEM.

Figure 2.. Reduced intrinsic excitability during the Light phase

(A) Left, examples of current injections (top) used to measure GC intrinsic properties (bottom). Right, number of spikes elicited by current steps revealed reduced spiking during the Light phase. n = 90 (Light); 69 (Dark). Unpaired t test to compare area under the curve (AUC) t = 2.9 **p < 0.01. (B) Violin plots show RMP, IR, TC, and rheobase during the Light (blue) and Dark (gray) phases. Mann-Whitney U test, U = 1,141 (RMP), U = 1,997 (IR), U = 1,845 (TC), and unpaired t test, t = 5.1 (rheobase). ****p < 0.0001. n = 90 (Light); 69 (Dark). Dot plots show intrinsic properties averaged in 1-h bins fit with Cosinor function across 24 h. Lights on at ZT0 and off at ZT12. n = 9–39 cells per bin. (C) Overlay of Cosinor fits for intrinsic properties, normalized in amplitude (same color code as in B). Blue and gray bars denote the Light and Dark recording windows, respectively. Arrows point to trough and peak of excitability based on acrophase for rheobase. (D) Violin plots show rheobase recorded at Subjective Day (Circadian Time or CT 8–11, where CT 12 refers to activity onset) and Subjective Night (CT 14–17) from mice housed in constant dark. Unpaired t test, t = 5 ****p < 0.0001. n = 27, 27. Bars on violin plots represent quartile and median values. Symbols represent mean ± SEM.

Figure 3.. Diurnal regulation of excitability requires intracellular GTP

(A) Left, example EPSPs at increasing stimulus intensities (as in Figure 1) in the absence of Na-GTP (GTP⁻) in the recording pipette. Right, the percentage of GCs recruited to spike by perforant path stimulation was similar during the Light and Dark phases; χ² tests p > 0.05 at all stimulus intensities. (B) Left, example voltage traces in response to somatic current injections of −10 pA (red), rheobase (blue), and 150 pA (black). Right, the number of spikes elicited by current steps was similar during the Light and Dark phases. Unpaired t test to compare area under the curve (AUC) t = 0.78; p = 0.44. (C) Intrinsic properties in the absence of GTP were similar between Light and Dark phases for RMP (left; Mann-Whitney U test, U = 212; p = 0.24), IR (middle; unpaired t test, t = 0.94; p = 0.35), and rheobase (right; unpaired t test, t = 0.5; p = 0.61). (A–C) n = 19 Light and 28 Dark. Bars on violin plots represent quartile and median values. Symbols represent mean ± SEM.

Figure 4.. Diurnal regulation of G-protein gated K⁺ and Na⁺ conductances

(A) Left, example voltage traces in response to current injections (−10 pA, rheobase and +150 pA) in CGP55845 (10 μM; red) or ML297 (10 μM; green) during Light (top) or Dark phases (bottom). Dotted line indicates −70 mV. Right, the number of action potentials elicited by current injections in the same conditions. Controls for application of CGP and ML297 (a paired analysis) were pooled. In Light, n = 11 for CGP, n = 12 for ML297. In Dark, n = 10 for CGP, n = 17 for ML297. One-way ANOVA to compare area under the curve (AUC) F_(2,43) = 46.9 and F_(2,51) = 65.2 for Light and Dark, respectively, followed by Tukey’s multiple comparisons test. **p < 0.01; ****p < 0.0001. (B) Violin plots showing effects of CGP55845 (red) or ML297 (green) during the Light and Dark phases. For RMP in Light, CGP paired t test t = 5.6; ML297 t = 7.1. In Dark, CGP t = 0.2; ML297 t = 10.3. For IR in Light, CGP paired t test t = 6.1; ML297 t = 8.9. In Dark, CGP t = 1.7; ML297 t = 8.7. For rheobase in Light, CGP paired t test t = 5.5; ML297 t = 8.1. In Dark, CGP t = 0.0; ML297 t = 9.7. n.s. p > 0.05; ***p < 0.001; ****p < 0.0001. (C) Left, example voltage traces in response to current injections (−10 pA, rheobase and +150 pA) in L-703,606 (10 μM; dark blue) during Light (top) or Dark phases (bottom). Dotted line indicates −70 mV. Middle, violin plots showing effects of L-703,606 during the Light and Dark phases. For RMP in Light, L-703,606 paired t test t = 0.5; p = 0.57, in Dark t = 4.8. For IR in Light, paired t test t = 1.6; p = 0.13, in Dark t = 4. **p < 0.01; n = 9, 9. Right, the number of action potentials elicited by current injections. Unpaired t test to compare area under the curve (AUC) t = 1.6; p = 0.11 and t = 0.8; p = 0.41 for Light and Dark, respectively. (D) Left, examples of isolated Na⁺ leak currents at −70 mV during the Light (blue) or Dark (gray) phases. Dashed line indicates the current blocked by NMDG in each condition. Right, violin plots show larger NMDG-sensitive Na⁺ currents during the Dark phase and in presence of GDPβS during Light phase. Unpaired t test in 2 mM Ca²⁺, t = 5.6; in 0.1 mM Ca²⁺, t = 4.5; GDPβS unpaired t test against Light 2 mM Ca²⁺, t = 5.9. ***p < 0.001; ****p < 0.0001. n = 9, 7, 9, 13, and 4 for Light 2 mM Ca²⁺, Dark 2 mM Ca²⁺, Light 0.1 mM Ca²⁺, Dark 0.1 mM Ca²⁺, and Light GDPβS, respectively. Bars on violin plots represent quartile and median values. Symbols represent mean ± SEM.

Figure 5.. GTP-dependent diurnal regulation of spike bursts

(A) Example voltage traces in response to rheobase current injection during the Light phase (left), Dark phase (middle), and Light phase using a GTP⁻ internal (right). Insets show enlargement of the action potentials (APs). Sectors represent the fraction of GCs with single or multiple APs in each condition (χ² tests, *p < 0.05; ***p < 0.001). (B) Examples of APs evoked in response to perforant path stimulation at 200 μA in GBZ. Sectors represent the percentage of GCs with each spike pattern, as well as non-spiking cells. χ² tests, *p < 0.05; **p < 0.01. No. of cells is indicated.

Figure 6.. Diurnal regulation of synaptic recruitment is disrupted by Bmal1 cKO

(A) Low-magnification confocal images from Pomc-Cre:tdT (control) and Pomc-Cre:tdT:Bmal1^lox/lox (cKO) mice showing BMAL1 immunolabeling. (B) Left, high-magnification images from control or cKO mice. Dashed lines show regions of interest (ROIs) where fluorescence was measured in arbitrary units (a.u.). *Highlights a tdT⁺ cell without Bmal1 deletion (see discussion). Right, examples of fluorescence measures from ROIs in the indicated channels. (C) R values show a negative correlation for tdT and Bmal1 only in cKOs. Filled symbols represent examples in (B). Unpaired t test t = 5.4. ****p < 0.0001. (D) BMAL1 levels in tdT⁻ and tdT⁺ neighboring cells in control and cKO mice. Paired t test t = 0.5; p = 0.58 and t = 5.8 for control and cKO, respectively. **p < 0.01. n = 18, 16 ROIs from three control and six cKO mice. (E) Schematic of recording from neighboring WT (tdT⁻) and cKO (tdT⁺) GCs. A greater fraction of cKO GCs compared with WT GCs spiked in response to perforant path stimulation in the Light phase (left, n = 14 pairs of GCs) but not in the Dark phase (right, n = 15 pairs of GCs). (F) Bmal1 cKO GCs exhibited burst firing in response to perforant path stimulation during Light phase. Sectors represent non-spiking (gray), single spikes (blue), and burst firing patterns (red). n = 14 pairs of GCs. (E and F) χ² tests, *p < 0.05; **p < 0.01. Symbols represent mean ± SEM.

Figure 7.. Diurnal regulation of intrinsic excitability is disrupted by Bmal1 Cko

(A) Left, examples of voltage traces in response to current injections (of −10 pA, rheobase, and 150 pA) for tdT⁻ (WT) and tdT⁺ (cKO) GCs. Dashed line marks avg RMP during the Dark phase. Right, number of APs elicited by current injections in the same conditions, blue symbols denote Light phase and black denotes Dark phase. Welch’s ANOVA to compare area under the curve (AUC) F_(3,28) = 3.44 followed by Dunnett’s multiple comparisons test. *p < 0.05. (B) Bmal1 cKO GCs exhibited burst firing during the Light phase. Sectors represent differences in discharge profile. χ² tests, ***p < 0.001. n = 14 pairs of WT and cKO GCs. (C) Intrinsic properties of WT and cKO GCs during the Light (blue) and Dark (black) phases. One-way ANOVA for RMP (F_(3,54) = 16.24), IR (F_(3,54) = 3.06), and rheobase (F_(3,54) = 4.52) followed by Tukey’s multiple comparison test. *p < 0.05; **p < 0.01; ****p < 0.0001. (A–C) Sample size n = 14 pairs in the Light phase, n = 15 in the Dark phase. (D) Top, example voltage traces before and after application of GCP55845 (10 μM; CGP) in a cKO GC in the Light phase. Dashed line marks avg RMP during the Dark phase. Bottom, comparison of the additional number of action potentials elicited by CGP (CGP – control) in Bmal1 WT and Bmal1 cKO GCs in the Light phase. Unpaired t test to compare area under the curve (AUC) t = 5.1; ****p < 0.0001. (E) Violin plots of intrinsic properties from Bmal1 cKO GCs show that CGP had no effect during the Light phase. Paired t test for RMP (t = 1.9), IR (t = 2), and rheobase (t = 0.8); p > 0.05. n = 12. (F) Violin plot showing a robust NMDG-sensitive current in Bmal1 cKO GCs in the Light phase (contrast with Figure 4C). Unpaired t test t = 3.7. **p < 0.01. n = 5. Bars on violin plots represent quartile and median values. Symbols represent mean ± SEM.