Nigrostriatal Dopamine Acting on Globus Pallidus Regulates Sleep

Abstract

Lesions of the globus pallidus externa (GPe) produce a profound sleep loss (∼45%) in rats, suggesting that GPe neurons promote sleep. As GPe neuronal activity is enhanced by dopamine (DA) from the substantia nigra pars compacta (SNc), we hypothesized that SNc DA via the GPe promotes sleep. To test this hypothesis, we selectively destroyed the DA afferents to the caudoputamen (CPu) using 6-hydroxydopamine and examined changes in sleep-wake profiles in rats. Rats with 80-90% loss of SNc neurons displayed a significant 33.7% increase in wakefulness (or sleep reduction). This increase significantly correlated with the extent of SNc DA neuron loss. Furthermore, these animals exhibited sleep-wake fragmentation and reduced diurnal variability of sleep. We then optogenetic-stimulated SNc DA terminals in the CPu and found that 20-Hz stimulation from 9 to 10 PM increased total sleep by 69% with high electroencephalograph (EEG) delta power. We finally directly optogenetic-stimulated GPe neurons and found that 20-Hz stimulation of the GPe from 9 to 10 PM increased total sleep by 66% and significantly increased EEG delta power. These findings elucidate a novel circuit for DA control of sleep and the mechanisms of abnormal sleep in BG disorders such as Parkinson's disease and Huntington's disease.

Keywords: Parkinson's disease; dopamine; globus pallidus; sleep; substantia nigra pars compacta.

Figure 1.

Selective loss of SNc DA neurons and projections to the CPu. TH-immunostained sections at the level of either SNc (A–C) or CPu (D–F) from a control rat (A and E), from a rat with partial lesion (B), from a rat with near complete loss of DA neurons in SNc and DA terminals in the CPu (C and F) and from a rat with unilateral 6-OHDA injections (D). The loss of DA cells and their projections were seen in SNc (B and C) and CPu (F) but did not extend to DA neurons in the VTA (B and C) and their terminals in the nucleus of accumbens (NAc) (D, left hemisphere and F).

Figure 2.

Loss of DA terminals in CPu leads to sleep alterations (A and B) Changes in daily amounts of wake, NREM, REM sleep, and hourly percentages of wake across 24 h in animals with complete SNc lesions (CL; >80% DA cell loss). (C and D) Changes in daily amounts of wake, NREM, REM sleep, and hourly percentages of wake across 24 h in animals with partial SNc lesions (PL; 50–80% DA cell loss). Both sets of animals showed a significant increase in wake and decrease in NREM sleep. While REM sleep was decreased in CL animals, it was unaltered in PL animals. When compared with the control animals, the CL rats displayed attenuation diurnal variation of wake due to higher increase in wakefulness during daytime (B) but there was no change in the diurnal pattern of wake in the PL animals as the wake increase was evenly distributed throughout the day (D). *P < 0.05, **P < 0.01.

Figure 3.

Correlation of SNc dopamine neuron loss with total wakefulness. The increase in total wake amounts/24 h negatively correlated with the number of remaining DA neurons in the SNc by 6-OHDA lesions in both CL and PL groups.

Figure 4.

Optogenetic stimulating DA fibers in the CPu promotes sleep. (A) The expression of AAV–ChR2–mCherry and the location of the plastic optic fiber tip were confirmed after each experiment. mCherry immunostaining shows that the ChR2 proteins were expressed in cells mostly confined to the SNc but not SNr (a), and the mCherry-positive fibers were obviously seen in the CPu (b). The arrow in (b) shows the cannula tip in the CPu. (B) Total amount of wake, NREM and REM sleep in control and 20-Hz photostimulation during 9 PM. (C and D) The hourly amount of NREM sleep of baseline and 20-Hz stimulation group at dark or light phase, respectively. The blue columns indicate the photostimulation period. (E) Wake, NREM, and REM sleep stage changing of baseline, 5-, 10-, and 20-Hz photostimulation of DA fibers in the CPu during 9 PM. (F, G, and H) Sleep-wake state transitions, number and mean duration of wake, NREM, and REM bouts in baseline and 20-Hz photostimulation during 9 PM. Stimulation does not significantly change state transitions and sleep bouts, but mean duration of NREM sleep are significantly increased. (I) EEG power densities of NREM sleep of baseline and photostimulation at 5, 10, or 20 Hz. Power spectrum at 1–3 range of 20-Hz stimulation is increased over the baseline but the increase is not significant (p > 0.05). *P < 0.05, **P < 0.01.

Figure 5.

GPe stimulation increase sleep. (A) “a and b” sections were immunostained with mCherry to verify the expression and injection site of the AAV. “a” shows the location of AAV–ChR2–mCherry in the bilateral GPe. “b” higher magnification of the square region indicated in “a.” The arrows indicate the cannula tip above the GPe; “c and d” sections were immunostained with c-Fos to estimate the ranges of optogenetic stimulations. The arrow in “c” indicates the cannula tip above the GPe. Dotted line indicated the boundary of c-Fos expression. “d,” higher magnification of the square region indicated in “a.” (B) Total amount of wake, NREM, and REM sleep in control and 20-Hz photostimulation during 9 PM. (C) The hourly amount of NREM sleep of baseline and 20-Hz stimulation group at dark or light phase, respectively. The blue columns indicate the photostimulation period. (D) Wake, NREM, and REM sleep stage changes of baseline, 5-, 10-, and 20-Hz photostimulation in the GPe during 9 PM. (E, F, and G): Sleep-wake state transitions, number, and mean duration of wake, NREM, and REM bouts in control and 20-Hz photostimulation during 9 PM. GPe stimulation induces more transitions from NREM sleep into REM sleep and from REM sleep to wakefulness, that is, more number of REM sleep episodes occurred, while average duration of NREM sleep is significantly increase, on the contrary, the mean duration of wake bouts is decreased. (H) EEG power densities of NREM sleep of control and photostimulation at 5, 10, or 20 Hz. There was no essential difference in EEG power density during NREM sleep between 5-, 10-Hz photostimulation, and baseline. Twenty-Hertz photostimulation significantly increases slow delta power. The horizontal bar indicates where there is a statistical difference (P < 0.05). *P < 0.05, **P < 0.01.

Figure 6.

Putative neural circuitry for SNc DA regulation of sleep. SNc DA acting on D₂ receptors on striatopallidal neurons lead to disinhibition of GPe neurons, which inhibits the cerebral cortex via their direct and/or indirect projections. Arrowheads represent the excitatory inputs and circles represent the inhibitory inputs. The light lines indicate the primary proposed pathways by which SNc dopamine and basal ganglia regulate sleep, and dark lines represent the submissive pathways that regulate sleep by SNc dopamine and basal ganglia.