Glutamatergic Neurons in the Cerebellar Lateral Nucleus Contribute to Motor Deficits Induced by Chronic Sleep Disturbance

Abstract

Background/Objectives: The cerebellum is essential for motor coordination and has recently been implicated in sleep-related disorders. However, the neural mechanisms linking sleep disruption to motor dysfunction remain poorly understood. This study aimed to elucidate the roles of the deep cerebellar nuclei (DCN), particularly the lateral nucleus, in motor dysfunction induced by chronic sleep disruption (CSD). Methods: Using a validated mouse model of CSD with periodic sleep fragmentation induced by an orbital shaker during the light phase, we assessed neuronal activation via c-Fos immunostaining and performed chemogenetic manipulation of glutamatergic neurons within the lateral nucleus. Behavioral performance was evaluated using open-field and rotarod tests. Results: CSD selectively increased c-Fos expression in the lateral nucleus, with no significant changes observed in other DCN subregions. Chemogenetic activation or ablation of glutamatergic neurons in the lateral nucleus decreased locomotor activity in the open-field test and shortened latency to fall in the rotarod task. Conversely, chemogenetic inhibition of these neurons attenuated CSD-induced impairments, restoring locomotor performance toward control levels. Conclusions: Our findings provide direct experimental evidence that glutamatergic neurons in the lateral nucleus play a crucial role in mediating CSD-induced motor dysfunction. These results highlight the cerebellar contribution to the interplay between sleep and motor control and identify a potential target for therapeutic intervention in sleep-related motor disorders.

Keywords: cerebellum; chronic sleep disruption; deep cerebellar nuclei; glutamatergic neurons; lateral nucleus; motor dysfunction.

Figure 1

Chronic sleep disruption alters sleep–wake states in mice. (A) Schematic representation of the chronic sleep disruption (CSD) protocol. Mice were subjected to a 7-day CSD protocol using an automated orbital shaker operating at 110 rpm in cycles of 20 s on and 100 s off, from 07:00 to 19:00. EEG/EMG recordings were performed in C57BL/6J mice housed in video-monitored recording chambers for a 24 h period after 7 consecutive days of sleep disturbance, aiming to evaluate changes in sleep architecture. (B–D) Hourly distribution of time spent in NREM sleep (B), REM sleep (C), and wakefulness (D) across the 24 h cycle in control (Ctrl, n = 5) and CSD (n = 6) groups. Statistical analysis was performed using repeated-measures ANOVA with multiple comparisons corrected by the false discovery rate (FDR): NREM, F(1,9) = 59.67, p < 0.001; REM, F(1,9) = 41.23, p = 0.001; Wake, F(1,9) = 84.36, p < 0.0001. Data are presented as mean ± SEM. * p < 0.05, ** p < 0.01. (E) Total time spent in NREM sleep, REM sleep, and wakefulness during a 24 h recording period in control and CSD groups (Ctrl, n = 5; CSD, n = 6). Group differences were assessed using unpaired two-tailed Student’s t-tests: NREM sleep, t(9) = 7.72, p = 2.9 × 10⁻⁵; REM sleep, t(9) = 6.42, p = 1.22 × 10⁻⁴; Wake, t(9) = 9.19, p = 7.0 × 10⁻⁶. Data are presented as mean ± SEM. ** p < 0.01.

Figure 2

Chronic sleep disruption impairs locomotor activity and motor coordination. (A) Schematic diagram of the open field test (OFT) used to assess locomotor behavior. (B) Representative movement trajectories of control mice (left) and CSD mice (right) during the 30 min OFT. (C–E) Quantification of locomotor parameters during the 30 min OFT. Compared with controls, CSD mice showed a significant reduction in total distance traveled (C), average velocity (D), whereas no significant difference was observed in percentage of active time (E). Data are presented as mean ± SEM (n = 9 per group). Unpaired two-tailed t-tests: C: t(16) = 2.229, p = 0.0405; D: t(16) = 2.228, p = 0.041; E: t(16) = 0.797, p = 0.437. * p < 0.05. (F) Schematic representation of the rotarod apparatus used to assess motor coordination and learning. (G) Rotarod performance across three consecutive testing days (three trials per day). On Day 1, CSD mice exhibited a significantly shorter latency to fall compared to controls. Performance gradually improved across days in both groups, and no significant differences were detected on Days 2 and 3. Data are presented as mean ± SEM (n = 9 per group). Two-way repeated-measures ANOVA followed by Sidak’s post hoc test, F(1,16) = 6.516, p = 0.0213. * p < 0.05.

Figure 3

Chronic sleep disruption alters the pattern of c-Fos expression in the deep cerebellar nuclei. (A,B) Representative immunofluorescence images showing c-Fos expression (red) in the cerebellum (CB) and deep cerebellar nuclei (DCN) of control (Ctrl, (A)) and chronic sleep disruption (CSD, (B)) mice. Dotted boxes indicate the regions of interest corresponding to the lateral (Lat), interposed anterior (IntA), and medial (Med) nuclei, which are shown at higher magnification in (a1–a3) and (b1–b3). DAPI (blue) was used for nuclear counterstaining. Scale bars: 1 mm in CB overview images; 500 μm in DCN overview images; 250 μm in high-magnification panels. (C) Quantification of c-Fos-positive cell numbers across DCN subregions, including the lateral (Lat), lateral parvicellular (LatPC), interposed anterior (IntA), interposed dorsolateral (IntDL), interposed posterior (IntP), medial (Med), and medial dorsolateral (MedDL) nuclei. Data are presented as mean ± SEM (n = 5 per group). Statistical comparisons were performed using unpaired two-tailed t-tests: Lat, t(8) = 3.06, p = 0.015; LatPC, t(8) = 0.48, p = 0.64; IntA, t(8) = 4.21, p = 0.003; IntDL, t(8) = 0.77, p = 0.459; IntP, t(8) = 0.31, p = 0.765; Med, t(8) = 1.36, p = 0.212; MedDL, t(8) = 0.22, p = 0.832. * p < 0.05, ** p < 0.01.

Figure 4

Chemogenetic activation of Lat^Vglut2+ neurons reduces locomotor activity and motor coordination. (A) Schematic of bilateral stereotaxic injection of rAAV-DIO-hM3Dq-mCherry into the Lat of Vglut2-IRES-Cre mice. (B) Representative immunofluorescence images showing expression of hM3D(Gq)-mCherry (red) and c-Fos (green) in the Lat of Vglut2-IRES-Cre mice. Compared with mCherry controls (left), hM3Dq-expressing mice (right) exhibited robust c-Fos expression following CNO administration, confirming chemogenetic activation of Lat^Vglut2+ neurons. White dashed lines outline the boundaries of the lateral nucleus. Scale bar: 200 μm. (C) Representative movement trajectories from mCherry or hM3Dq mice during the 30 min OFT. (D–F) Quantification of locomotor parameters during the 30 min OFT. Compared with controls, hM3Dq mice showed a significant reduction in total distance traveled (D), average velocity (E), and active time (F). Data are presented as mean ± SEM (n = 9 per group). Unpaired two-tailed t-tests: (D) t(16) = 3.960, p = 0.0011; (E) t(16) = 3.960, p = 0.0011; (F) t(16) = 4.029, p = 0.0010. ** p < 0.01. (G) Rotarod performance over three consecutive days (three trials per day). On Day 1, hM3Dq-expressing mice exhibited a significantly shorter latency to fall compared to controls. Performance progressively improved across subsequent days in both groups. Data are presented as mean ± SEM (n = 9 per group). Two-way repeated-measures ANOVA with Sidak’s post hoc test, F(1,16) = 14.72, p = 0.0015. ** p < 0.01.

Figure 5

Selective ablation of Lat^Vglut2+ neurons impairs locomotor performance and motor coordination. (A) Schematic of bilateral stereotaxic injection of rAAV-EF1α-FLEX-taCasp3 into the Lat of Vglut2-IRES-Cre mice to induce Cre-dependent neuronal ablation. (B) Representative immunofluorescence images showing NeuN (green) and DAPI (blue) staining in the Lat of control and taCasp3-treated mice, confirming neuronal loss following taCasp3 expression. White dashed lines outline the boundaries of the lateral nucleus. Scale bars: 500 μm. (C) Representative movement trajectories of control and taCasp3-treated mice in the 30 min OFT. (D–F) Quantification of locomotor activity during the 30 min OFT. Compared to controls, taCasp3-treated mice exhibited significantly reduced total distance traveled (D) and average velocity (E), whereas the percentage of active time (F) was not significantly different. Data are presented as mean ± SEM (n = 10 per group). Unpaired two-tailed t-tests: (D) t(18) = 2.580, p = 0.019; (E) t(18) = 2.592, p = 0.018; (F) t(18) = 1.584, p = 0.131. * p < 0.05. (G) Rotarod performance across three consecutive days (three trials per day). On Day 1, taCasp3-treated mice exhibited significantly shorter latency to fall compared to controls. Motor performance gradually improved over subsequent trials in both groups. Data are presented as mean ± SEM (n = 11 per group). Two-way repeated-measures ANOVA with Sidak’s post hoc test, F(1,20) = 10.07, p = 0.0048. ** p < 0.01.

Figure 6

Selective ablation of Lat^Vglut2+ neurons does not affect anxiety-like behavior or spatial working memory. (A) Schematic of the elevated plus maze (EPM) used to assess anxiety-like behavior. (B–G) Quantification of EPM parameters, including total arm entries (B), percentage of open arm entries (C), number of open arm entries (D), number of closed arm entries (E), time spent in open arms (F), and time spent in closed arms (G). No significant differences were observed between control and taCasp3-treated mice, indicating that ablation of Lat^Vglut2+ neurons did not alter anxiety-like behavior. Data are presented as mean ± SEM (n = 10 per group). Unpaired two-tailed t-tests: (B) t(18) = 0.886, p = 0.387; (C) t(18) = 0.754, p = 0.461; (D) t(18) = 0.417, p = 0.681; (E) t(18) = 0.9284, p = 0.366; (F) t(18) = 0.170, p = 0.867; (G) t(18) = 0.440, p = 0.665. (H) Schematic of the Y-maze used to evaluate spatial working memory. (I,J) Quantification of spontaneous alternation percentage (I) and total arm entries (J) during the Y-maze test. taCasp3-treated mice exhibited significantly fewer total arm entries, while spontaneous alternation was unaffected. Data are presented as mean ± SEM (n = 12 per group). Unpaired two-tailed t-tests: (I): t(22) = 0.089, p = 0.930; (J): t(22) = 3.694, p = 0.001. ** p < 0.01.

Figure 7

Chemogenetic inhibition of Lat^Vglut2+ neurons attenuates CSD-induced motor deficits. (A) Schematic representation of bilateral stereotaxic injection of rAAV-DIO-hM4Di-mCherry into the Lat of Vglut2-IRES-Cre mice. (B) Representative immunofluorescence images showing mCherry (red) and c-Fos (green) expression in control (mCherry) and hM4Di-expressing mice. White dashed lines delineate the boundaries of the Lat. Quantification of c-Fos-positive cells revealed a significant reduction in neuronal activity in the hM4Di group (red circles) compared with the mCherry control group (white circles). Data are presented as mean ± SEM (n = 5 per group). Unpaired two-tailed t-tests: t (4) = 4.079, p = 0.004. ** p < 0.01. Scale bar: 200 μm. (C) Representative movement trajectories of mice from mCherry-Ctrl, mCherry-CSD, and hM4Di-CSD groups during the 30 min OFT. (D–F) Quantification of locomotor activity during the 30 min OFT. Compared with mCherry-CSD mice, hM4Di-expressing CSD mice exhibited significantly increased total distance traveled (D), average velocity (E), while the percentage of active time showed a non-significant increasing trend (F). Data are presented as mean ± SEM (n = 5 per group), One-way ANOVA with Tukey’s post hoc test. * p < 0.05. (G) Rotarod performance across three consecutive days (three trials per day). On Day 1, hM4Di-expressing CSD mice displayed longer latency to fall compared with mCherry-CSD control. No significant group differences were detected on Days 2–3. Data are presented as mean ± SEM (n = 5 per group). Two-way repeated-measures ANOVA with Tukey’s post hoc test, * p < 0.05.