Thalamic Cav3.1 T-type Ca2+ channel plays a crucial role in stabilizing sleep

Abstract

It has long been suspected that sensory signal transmission is inhibited in the mammalian brain during sleep. We hypothesized that Cav3.1 T-type Ca2+ channel currents inhibit thalamic sensory transmission to promote sleep. We found that T-type Ca2+ channel activation caused prolonged inhibition (>9 s) of action-potential firing in thalamic projection neurons of WT but not Cav3.1 knockout mice. Inhibition occurred with synaptic transmission blocked and required an increase of intracellular Ca2+. Furthermore, focal deletion of the gene encoding Cav3.1 from the rostral-midline thalamus by using Cre/loxP recombination led to frequent and prolonged arousal, which fragmented and reduced sleep. Interestingly, sleep was not disturbed when Cav3.1 was deleted from cortical pyramidal neurons. These findings support the hypothesis that thalamic T-type Ca2+ channels are required to block transmission of arousal signals through the thalamus and to stabilize sleep.

Fig. 1.

Generation of K128-Cre transgenic mice that delete loxP-flanked sequence in rostral–midline thalamic projection neurons. (A) Schematic diagram of the procedure for creating a transgenic construct in which nuclear-targeted Cre recombinase gene with polyadenylation sequence (nCre-pA) is placed under control of the K_v3.2 promoter in a 120-kb genomic BAC DNA fragment. (B–D) K128-Cre transgene deletes loxP-flanked sequence to label neuron processes by alkaline phosphatase staining in 1-week-old (anterior ventral thalamic nucleus, representative example, n = 15 cells) (B) and 1-month-old (anterior ventral thalamic nucleus, C; presubicular cortex, D) K128-Cre/+; Z/AP/+ double-transgenic male mouse. (E–G) K128-Cre transgene deletes loxP-flanked sequence to label neuron nuclei blue by X-gal staining in a 4-month-old male K128-Cre/+; Rosa26/+, double-transgenic male mouse. Coronal brain sections were counterstained with nuclear fast red.

Fig. 2.

Generation of mice carrying a loxP-flanked Ca_v3.1 gene (fCa_v3.1/fCa_v3.1) for postnatal, region-restricted deletion of Ca_v3.1 T-type Ca²⁺ channel. (A) Schematic representation of the procedure for inserting loxP sequences around exons 9–12 of the gene encoding T-type Ca²⁺ channel, Ca_v3.1 (fCa_v3.1; see Supporting Materials and Methods, which is published as supporting information on the PNAS web site, for details). (B and C) Distribution of Ca_v3.1 mRNA in brains of 4-month-old WT (B) and global Ca_v3.1 deleted (C) male mice. Parasagittal brain sections are shown. (D–F) Distribution of Ca_v3.1 mRNA in brains of 4-month-old WT (D), thalamic (E), or cortical (F) Ca_v3.1 KO male mice. The genotypes are fCa_v3.1/fCa_v3.1; fCa_v3.1/fCa_v3.1, K128-Cre/+; and fCa_v3.1/fCa_v3.1, CW2-Cre/+, respectively. Coronal brain sections are shown. DNA flanked by LoxP and FRT sequences is deleted by Cre and Flp recombinase, respectively. DT-A and PGK-Neo are negative and positive selection markers, respectively. See Materials and Methods for more details.

Fig. 3.

Ca_v3.1 T-type calcium channel activation inhibits Na⁺ action-potential firing. A hyperpolarizing current injection (1 s) caused prolonged inhibition of Na⁺ action potentials in neurons of WT (A–C)(n = 13), but not global Ca_v3.1 KO (B and C)(n = 16), mice. Cells were obtained from lateral dorsal thalamic nucleus, and they displayed a multipolar dendritic arbor. (C) TEA (4 mM) blocks inhibition in WT neurons (P < 0.002, repeated-measures ANOVA, n = 5 each). (D) Depolarizing current injections, in hyperpolarized neurons (–80 mV, left traces), evoked a Ca²⁺ potential and brief Na⁺ action-potential burst in WT (n = 12/12) but no Ca²⁺ potential (n = 0/15) and persistent Na⁺ action-potential firing in KO (n = 15/15). Ca²⁺ potential activation and Na⁺ action-potential inhibition are missing with current injections to neurons held at –55 to –60 mV (right traces). (E) The ratio of Na⁺ action-potential firing frequencies evoked by current injections from –80 vs. –55 to –60 mV revealed inhibition in WT (n = 7), but not KO (n = 7), neurons. Inhibition was blocked in WT neurons perfused with intracellular BAPTA (15 mM, n = 9) or extracellular forskolin (10 μM, n = 6). Current pulses are shown below the traces. *, P < 0.02; **, P < 0.01, unpaired t test.

Fig. 4.

Global and thalamic Ca_v3.1 deletion disrupts sleep. (A) Vigilance-state hypnograms, during the 12-h light (rest) period, revealed NR sleep interrupted by frequent brief wake bouts (W) in global and thalamic, but not cortical Ca_v3.1, KO mice. REM was unchanged. (B) Global and thalamic Ca_v3.1 KO mice displayed more brief and numerous wake and NR bouts. Bar graphs of bout number (Left) and duration (Right) for wake, REM, and NR are shown for WT (floxed), global, cortical, and thalamic Ca_v3.1 KO mice. Data are from the 12-h light period. *, P < 0.02, unpaired t test. (C) Vigilance-state hypnograms during the 12-h dark (active) period revealed delayed sleep onset after lights out in global and thalamic Ca_v3.1 KO. (D) NR sleep time over the 24-h light/dark period showed decreased NR sleep during the 12-h dark (hatched bar) period in global and thalamic Ca_v3.1 KO mice. Black bar indicates P < 0.02 (repeated-measures ANOVA; n = 8 mice each genotype).