Effect of dopaminergic substances on sleep/wakefulness in saline- and MPTP-treated mice

Authors


Philippe Derambure, EA 2683, Laboratoire de Physiologie, Faculté de médecine Henri Warembourg, Pôle recherche, 1 place de Verdun, F-59045 Lille cedex, France. Tel.: 33-320-62-68-56; fax: 33-320-62-69-93; e-mail: p-derambure@chru-lille.fr

Summary

Sleep/wakefulness (S/W) disorders are frequent in Parkinson’s disease (PD). The underlying causes have yet to be elucidated but dopaminergic neurodegenerative lesions seem to contribute to appearance of the disorders and anti-Parkinsonian medication is known to accentuate S/W problems. Hence, we reasoned that studying the acute effect of dopaminergic compounds on S/W in an animal model of PD might improve our knowledge of S/W regulation in the context of partial dopaminergic depletion. To this end, we tested the effect of levodopa (l-dopa), pergolide (a mixed D2/D1 agonist) and lisuride (a D2 agonist) on S/W recordings in MPTP-treated mice, in comparison with controls. Our results showed that dopaminergic compounds modify S/W amounts in both control and MPTP mice. Wakefulness amounts are greater in MPTP mice after l-dopa (50 mg kg−1) and lisuride (1 mg kg−1) injections compared with control mice. Moreover, the paradoxical sleep latency was significantly longer in MPTP mice after high-dose l-dopa administration. Our observations suggest that the actions of both l-dopa and lisuride on S/W differ slightly in MPTP mice relative to controls. Hence, MPTP-induced partial DA depletion may modulate the effect of dopaminergic compounds on S/W regulation.

Introduction

Insomnia, paradoxical sleep (PS) behaviour disorders, excessive daytime sleepiness and ‘sleep attacks’ are frequent complaints in Parkinson’s disease (PD) (Arnulf, 2005; Frucht et al., 1999; Lees et al., 1988; Monaca et al., 2006). Together with the prime motor symptoms, these sleep-related disorders are the main causes of disability in PD and have a substantial impact on a patient’s quality of life. The underlying causes of the sleep disorders in PD are still poorly known but may be due to the pathophysiology of the disease itself and/or the side effects of dopaminergic medication.

The dopaminergic system has long been considered as an ‘awakening’ system but an increasing body of evidence suggests that dopamine (DA) is also involved in sleep regulation (Monti and Monti, 2007; Rye, 2004; Rye and Jankovic, 2002). In particular, a number of pharmacological studies have shown that DA reuptake inhibitors, levodopa and D1, D2 & D3 receptor agonists all modify the amounts of slow wave sleep (SWS) and PS (Galarraga et al., 1986; Lagos et al., 1998; Mayers and Baldwin, 2005; Monti et al., 1988; Nishino et al., 1998; Python et al., 1996; Trampus et al., 1993) without necessarily changing wakefulness (Crochet and Sakai, 2003; Gillin et al., 1973). Moreover, numerous clinical observations have described levodopa and DA-receptor agonists (pergolide, ropinirole, lisuride, etc.) as being responsible for excessive daytime sleepiness and ‘sleep attacks’ (Arnold, 2000; Arnulf, 2005; Ferreira et al., 2000, 2001; Garcia-Borreguero et al., 2003; Monaca et al., 2006; Schapira, 2000).

Elsewhere, alterations in dopaminergic function seem to modify sleep/wakefulness (S/W). In an animal model study, Dzirasa et al. (2006) recently demonstrated the role of DA in sleep–wake states by using hyperdopaminergic mice and acutely DA-depleted mice. Both types of animal showed PS disturbances which could be reversed by administering dopaminergic compounds. In human, central dopaminergic synaptic transmission is known to be altered in several neurological and psychiatric disorders (such as PD, schizophrenia and attention deficit hyperactivity disorder) and patients suffering from these conditions frequently display sleep disorders: sleep-onset PS, PS behaviour disorders, disturbed sleep architecture, excessive daytime sleepiness, etc. (Arnulf, 2005; Arnulf et al., 2000; Maggini et al., 1986; Monaca et al., 2006; O’Brien et al., 2003; Rye et al., 2000). Similarly, animal models of PD (characterized by striatal dopaminergic depletion) can also suffer from S/W disorders – particularly those concerning PS (Almirall et al., 1999; Daley et al., 1999; Decker et al., 2000; Monaca et al., 2004; Pungor et al., 1990). Our previous analysis of the S/W structure in MPTP-treated mice (presenting a 30% reduction in the number of nigral DA neurons) revealed a modified sleep architecture, with (i) a longer mean duration of wakefulness (W) and PS episodes and (ii) a greater total amount of PS over the nycthemeral period, compared with saline-treated control mice (Monaca et al., 2004). Although these disorders do not correspond directly to those seen in Parkinsonian patients, our data suggested that S/W regulation can indeed be modified by partial dopaminergic depletion. These results were obtained in the absence of any dopaminergic treatment and strongly suggested that neurodegenerative lesions play a role in the induction of S/W disturbances.

The aim of the present study was to evaluate the effects of acute injections of dopaminergic compounds on S/W recordings in MPTP mice (compared with control mice) and thus to determine the responsiveness of the dopaminergic system in these animals. To this end, we compared the effect of acute injection of levodopa, pergolide (a mixed D1/D2 agonist) and lisuride (a D2 agonist) on S/W recordings in MPTP-treated mice versus control mice.

Materials and methods

Animals and treatment

We used male C57BL/6 mice (at 2–3 months of age). All procedures involving animals and their care were performed in compliance with our institutional guidelines, which in turn comply with current national and international laws and guidelines (notably the Governmental Decree no. 87–848 issued on 19 October 1987, by the French Ministry of Agriculture and Forestry’s Veterinary Service for Animal Health and Protection).

The mice were injected intraperitoneally (i.p.) once a day for 5 days with 0.1 mL of a saline solution containing 0 (control mice) or 25 mg kg−1 (MPTP mice) of 1-methyl-4-phenyl-1,2,3,6-tetrahydropiridine (MPTP) (Sigma Aldrich, St Louis, MO, USA).

Surgery

Ten days after the last injection of MPTP, mice were implanted (under sodium pentobarbital anaesthesia, 75 mg kg−1, i.p.) with a standard set of electrodes (enamelled nickel–chrome wire, 150 μm in diameter) for polygraphic sleep monitoring, as described previously (Boutrel et al., 1999; Monaca et al., 2003; Popa et al., 2005). Briefly, electroencephalography electrodes were inserted through the skull into the dura over the right cortex (2 mm lateral and 4 mm posterior to the bregma) and cerebellum (at the midline, 2 mm posterior to lambda). Electro-oculography electrodes were also positioned subcutaneously on each side of the left orbit, and electromyography electrodes were inserted into the neck muscles. All electrodes were anchored to the skull with Superbond (Limoge-Lendais et al., 1994) and acrylic cement and were then soldered to a miniconnector (also embedded in cement). After surgery, animals were housed in individual cages (20 × 20 × 30 cm) and maintained under standard laboratory conditions: a 12 h light/dark cycle (lights on at 07:00 hours), an ambient temperature of 22 ± 2 °C and food and water available ad libitum. The animals were allowed 7–10 days to recover and habituate to the recording conditions.

Pharmacological treatments

Pergolide, lisuride and levodopa were purchased from Sigma (St Louis, MO, USA). Pergolide and lisuride were each dissolved in 0.1 mL of saline +0.05% of ethanol and injected i.p. at doses of 0.5, 1, 2.5 and 5 mg kg−1 and 1, 2.5 and 5 mg kg−1, respectively. Levodopa (l-dopa) was dissolved in 0.1 mL of saline +12 mg kg−1 of benserazide (a peripherally acting inhibitor of dopa-decarboxylase) and injected at doses of 10, 30 and 50 mg kg−1 i.p. Each mouse received a single drug at all doses, and injections of different doses were randomized. To obtain baseline data (i.e. 0 mg kg−1 of the substance), mice were injected with saline + ethanol or saline + benserazide, as appropriate. A washout period of 2–4 days was allowed between two consecutive treatments.

Recording of S/W parameters after injection of saline alone was performed in each mouse (2 days before administering other injections) and compared with saline + ethanol or saline + benserazide to verify the lack of effect of these additives on vigilance states. Our results indicated that injection of saline + ethanol or saline + benserazide had no effect on vigilance states (compared with injection of saline alone) in both control and MPTP mice.

Polygraphic recording

Polygraphic sleep monitoring was initiated at 10:00 hours (just after dopaminergic compound injection) and continued for 8 h thereafter. These time conditions are commonly used to analyse the early-onset effects of pharmacological treatments on sleep in mice (Monaca et al., 2003; Popa et al., 2005).

Tissue preparation and tyrosine hydroxylase immunohistochemistry

Mice were killed after the last sleep recording. They were deeply anaesthetized with sodium pentobarbital (180 mg kg−1, i.p.) and, following a heparinized saline flush, transcardially perfused with 4% paraformaldehyde in 0.1 m phosphate buffer (pH 7.4). The brains were removed and, after a post-fixation process, were included in paraffin.

Coronal sections of 6-μm thick were prepared from the striatum and the substantia nigra, using a microtome (Leica, Nussloch, Germany).

The sections were incubated successively with rabbit polyclonal anti-tyrosine hydroxylase antibody (1 : 1000; Chemicon International, Temecula, CA, USA), goat biotinylated polyclonal anti-rabbit antibody (1 : 500, Vectastain Elite ABC Kit; Vector Laboratories, Burlingame, CA, USA) and horseradish-peroxidase-conjugated avidin complex (Vectastain Elite ABC Kit; Vector Laboratories). Sections were then exposed to diaminobenzidine for detection.

Data analysis and statistics

Polygraphic recordings were divided into 5-s epochs and scored visually as wakefulness (W), SWS and PS in accordance with standard criteria (Lena et al., 2005; Tobler et al., 1997) and using Deltamed® software (Deltamed SA, Paris, France). The total amount of each vigilance state and the numbers and mean durations of each type of episode were calculated per 2, 4 and 8 h of recording.

For a given treatment, each animal was referred to its own baseline. The values for each group were expressed as means in min ± SEM. To allow for the small number of animals, our statistical analysis used Conover’s free distribution method, a non-parametric analysis of variance based on ranks (Conover and Iman, 1982). These analyses were performed using two factors: GROUP (controls, MPTP mice) and DOSE (Levodopa 10 and 30 mg kg−1, Pergolide 2.5 and 5 mg kg−1, Lisuride 2.5 and 5 mg kg−1). First, main effects were tested. If a DOSE effect was found to be significant at the 0.05 level, we then performed a contrast analysis. Finally, if an interaction was observed, the Mann–Whitney test was applied.

Some of our data could not have been interpreted because of the presence of artefacts on the EEG or environmental disturbance during the day of recording. These data corresponded to the doses: Levodopa 50 mg kg−1, Pergolide 0.5 and 1 mg kg−1 and Lisuride 1 mg kg−1. As a consequence, the effects induced by these doses were tested separately using Wilcoxon and Mann–Whitney tests.

Statistical analysis was performed using spss software v.13 (SPSS Inc., Chicago, IL, USA).

The number of tyrosine hydroxylase (TH)-positive neurons in each mouse (in three, pre-determined sections of the substantia nigra pars compacta) was counted under a light microscope at a magnification of ×400. This value was considered to be representative of the number of dopaminergic nigral cells in each animal. Striatal TH staining was evaluated (as an optical density) in MPTP mice and compared with the value from control mice.

A Mann–Whitney test was used to compare the groups in terms of the number of TH-positive neurons and the TH staining intensity.

Results

Results for control mice

l-dopa had no effect on W and SWS. A significant reduction in the amount of PS was observed for l-dopa at the dose of 10 and 30 mg kg−1, compared with baseline during the first 4 h of recording (14.9 ± 2.4 min and 17.4 ± 2.7 min compared with 21.2 ± 3.0 min, P < 0.05). These effects were accounted by a lower number of PS episodes for both doses (Table 2). An increase in PS latency was observed for the dose of 50 mg kg−1 (Table 3).

Table 2.   Effect of various doses of levodopa, pergolide and lisuride on the number of paradoxical sleep (PS) episodes in control and MPTP mice during the 8-h recording period. Data are expressed as means ± SEM (for n animals) for the four 2-h periods and the whole 8 h of recording
DrugsDose (mg kg−1)n 0–2 h 2–4 h 4–6 h 6–8 h 0–8 h
  1. *Indicates a significant (P < 0.05) difference from baseline in each group of mice.

Control mice
l-dopa01010.9 ± 2.015.7 ± 1.716.9 ± 1.120.0 ± 2.063.5 ± 5.1
1069.5 ± 2.210.8 ± 2.3*18.0 ± 3.316.5 ± 3.854.8 ± 9.5
3065.8 ± 1.1*14.3 ± 3.219.5 ± 2.818.5 ± 1.758.2 ± 6.2
5065.8 ± 1.215.7 ± 2.717.5 ± 2.417.7 ± 3.256.7 ± 4.0
Pergolide087.5 ± 1.513.7 ± 1.314.3 ± 1.313.4 ± 2.048.9 ± 4.3
0.568.2 ± 1.710.7 ± 0.814.2 ± 1.210.7 ± 1.343.7 ± 4.3
161.8 ± 0.6*11.5 ± 2.411.0 ± 1.312.3 ± 2.636.7 ± 5.2
2.562.0 ± 1.0*8.3 ± 1.79.8 ± 1.611.3 ± 2.031.5 ± 2.8
563.8 ± 1.48.3 ± 2.312.7 ± 1.217.8 ± 3.342.2 ± 5.2
Lisuride077.8 ± 1.313.1 ± 1.711.6 ± 1.611.4 ± 2.443.9 ± 5.8
1610.0 ± 1.714.5 ± 1.511.7 ± 1.710.5 ± 1.746.7 ± 4.8
2.5612.7 ± 1.011.3 ± 2.48.3 ± 1.712.3 ± 2.544.7 ± 6.1
569.8 ± 1.011.0 ± 0.913.3 ± 2.513.7 ± 1.747.8 ± 4.6
MPTP mice
l-dopa0119.9 ± 1.017.4 ± 1.717.9 ± 1.619.4 ± 1.764.6 ± 4.8
1088.5 ± 2.117.9 ± 3.820.7 ± 2.819.0 ± 1.666.1 ± 9.1
3086.9 ± 1.113.7 ± 1.818.2 ± 3.123.5 ± 3.562.4 ± 6.4
5071.6 ± 0.5*14.6 ± 3.223.6 ± 4.021.0 ± 2.560.7 ± 8.7
Pergolide077.0 ± 1.612.0 ± 1.813.9 ± 1.811.7 ± 1.144.6 ± 6.0
0.567.2 ± 1.39.5 ± 0.710.0 ± 2.07.7 ± 0.834.3 ± 3.2
162.0 ± 0.7*11.3 ± 2.313.7 ± 1.010.7 ± 1.437.7 ± 3.9
2.573.0 ± 1.09.8 ± 1.510.7 ± 1.012.5 ± 0.936.0 ± 2.9
573.6 ± 1.28.9 ± 2.111.0 ± 1.514.3 ± 3.737.7 ± 6.1
Lisuride089.4 ± 1.711.5 ± 2.012.0 ± 1.211.1 ± 2.644.0 ± 6.6
166.8 ± 3.16.3 ± 1.9*9.2 ± 2.78.2 ± 2.730.5 ± 9.2*
2.5610.3 ± 1.59.7 ± 1.79.2 ± 1.3*10.0 ± 1.939.2 ± 5.6
568.7 ± 1.110.0 ± 1.112.0 ± 1.113.0 ± 2.643.7 ± 5.4
Table 3.   Effect of various doses of levodopa, pergolide and lisuride on slow-wave sleep (SWS) and paradoxical sleep (PS) latencies in control and MPTP mice. Data are expressed as means ± SEM (for n animals)
DrugDose (mg kg−1)Control miceMPTP mice
nSWSPSnSWSPS
  1. *Indicates a significant (P < 0.05) difference from baseline in each group of mice.

  2. Indicates a significant (P < 0.05) difference between both groups of mice.

l-dopa01010.0 ± 1.245.4 ± 5.91113.5 ± 2.348.4 ± 2.9
10613.0 ± 4.142.6 ± 6.689.6 ± 2.045.4 ± 5.4
30612.2 ± 3.047.4 ± 11.1812.2 ± 3.854.7 ± 4.5
50611.8 ± 1.767.1 ± 7.4*717.3 ± 3.3104.7 ± 10.7*
Pergolide0814.0 ± 4.057.0 ± 13.0712.5 ± 2.749.5 ± 10.4
0.564.3 ± 1.034.1 ± 4.466.3 ± 2.0*37.6 ± 5.7
1661.1 ± 9.7*109.3 ± 16.6*652.1 ± 14.1*98.4 ± 14.0*
2.5655.1 ± 6.9*127.6 ± 24.7*748.9 ± 6.9*109.3 ± 16.2*
5661.9 ± 24.4*115.7 ± 30.7*766.8 ± 20.6*98.6 ± 23.1*
Lisuride0713.4 ± 2.351.9 ± 3.5811.4 ± 2.041.9 ± 4.0
1612.3 ± 2.238.0 ± 6.6610.2 ± 2.0112.1 ± 38.2
2.569.4 ± 2.132.7 ± 3.6611.0 ± 2.538.9 ± 7.1
5613.3 ± 2.048.6 ± 2.8618.5 ± 2.349.7 ± 3.8

For the dose of 1, 2.5 and 5 mg kg−1 of pergolide, during the 0–2 h period, a significant dose effect was observed for W and SWS: these doses increased W and decreased SWS amounts (Fig. 2). This effect on W in also observed during the whole 8 h of recording (Table 1). Effect on SWS was accounted by a significant modification in SWS latency (Table 3). For the dose of 1, 2.5 and 5 mg kg−1, a significant decrease in PS amounts was induced during the 0–2 h period (Fig. 2), and was accounted by an increase in PS latency (Table 3) and a number of PS episodes reduction (Table 2). For both higher doses, this effect was lasting during the four first hours and the whole 8 h of recording (Table 1).

Figure 2.

 Effect of pergolide on the amounts of wakefulness (W), slow-wave sleep (SWS) and paradoxical sleep (PS) in control (left) and MPTP mice (right) during the 8 h post-injection. Data are expressed as means ± SEM (n = 6 to 8 animals) for each of the four 2-h periods following injection of pergolide at various doses (0, 0.5, 1, 2.5 and 5 mg kg−1). *Indicates a significant (P < 0.05) difference from baseline in each group of mice.

Table 1.   Effect of various doses of levodopa, pergolide and lisuride on amounts of wakefulness (W), slow-wave sleep (SWS) and paradoxical sleep (PS) in control and MPTP mice throughout the 8-h recording period. Data are expressed as means ± SEM (for n animals)
DrugDose (mg kg−1)Control miceMPTP mice
nWSWSPSnWSWSPS
  1. *Indicates a significant (P < 0.05) difference from baseline in each group of mice.

  2. Indicates a significant (P < 0.05) difference between both groups of mice.

l-dopa010174.3 ± 5.6 256.8 ± 4.748.9 ± 2.811182.8 ± 6.4243.2 ± 7.549.5 ± 2.8
106195.7 ± 11.6243.6 ± 11.440.9 ± 4.08195.1 ± 9.4240.4 ± 12.144.5 ± 5.2
306187.6 ± 14.2248.6 ± 13.843.8 ± 3.48184.4 ± 7.8247.8 ± 7.2547.8 ± 5.0
506169.1 ± 10.5265.1 ± 7.745.8 ± 3.07203.2 ± 9.6232.3 ± 10.144.5 ± 5.3
Pergolide08203.5 ± 15.3229.5 ± 12.247.0 ± 3.97187.4 ± 15.5244.3 ± 12.548.2 ± 4.8
0.56164.8 ± 14.9272.2 ± 12.642.9 ± 4.26162.9 ± 9.6278.9 ± 8.938.2 ± 3.1
16233.6 ± 16.6*208.6 ± 17.037.8 ± 4.26201.7 ± 14.1230.7 ± 12.047.6 ± 3.7
2.56224.1 ± 19.6220.1 ± 16.5 35.8 ± 5.0*7209.4 ± 9.7232.2 ± 7.838.4 ± 2.6*
56207.1 ± 20.9233.4 ± 18.039.5 ± 4.6*7223.7 ± 19.8218.2 ± 16.138.1 ± 5.6*
Lisuride07185.4 ± 8.9253.1 ± 5.641.4 ± 4.78178.4 ± 8.5260.9 ± 9.840.7 ± 5.6
16169.4 ± 10.4267.9 ± 9.242.6 ± 3.36195.5 ± 16.2256.5 ± 12.928.1 ± 7.2*
2.56174.5 ± 8.2262.9 ± 7.042.5 ± 5.66186.2 ± 14.6253.5 ± 12.640.3 ± 5.0
56161.9 ± 9.5267.5 ± 9.250.6 ± 3.36186.61 ± 17.1250.4 ± 13.243.0 ± 5.0

At doses of 2.5 and 5 mg kg−1, lisuride significantly reduced the amount of W and increased the amount of SWS during the 0–2 h post-injection period (Fig. 3). Lisuride had no effect on PS.

Figure 3.

 Effect of lisuride on the amounts of wakefulness (W), slow-wave sleep (SWS) and paradoxical sleep (PS) in control (left) and MPTP mice (right) during the 8 h post-injection. Data are expressed as means ± SEM (n = 6 to 8 animals) for each of the four 2-h periods following injection of pergolide at various doses (0, 1, 2.5 and 5 mg kg−1). *Indicates a significant (P < 0.05) difference from baseline in each group of mice. $Indicates a significant (P < 0.05) difference between both groups of mice.

Results for MPTP mice

With l-dopa, a significant dose effect was found for W and SWS amounts during the 0–2 h period. Post-hoc analysis showed a decrease in W amounts with concomittant increase in SWS amounts for the dose of 30 mg kg−1 compared with baseline (Fig. 1). For the dose of 50 mg kg−1, W amounts tend to increase and SWS amounts tend to decrease during the 0–2 h period (P = 0.063). A significant dose effect was found for PS amounts during the first 4 h. A reduction was observed for l-dopa at doses of 10 and 30 mg kg−1 compared with baseline (16.9 ± 3.1 and 15.0 ± 1.8 min compared with 19.3 ± 1.3 min, P < 0.05). At a dose of 50 mg kg−1, a reduction of PS amounts was found during the 0–2 h period (Fig. 1) and the four first hours (13.0 ± 2.5 min compared with 21.9 ± 1.7 min, P < 0.05) accounted by a lower number of episodes (Table 2). An increase in PS latency was observed for this dose (Table 3).

Figure 1.

 Effect of levodopa on the amounts of wakefulness (W), slow-wave sleep (SWS) and paradoxical sleep (PS) in control (left) and MPTP mice (right) during the 8 h post-injection. Data are expressed as means ± SEM (n = 6 to 11 animals) for each of the four 2-h periods following injection of l-dopa at various doses (0, 10, 30 and 50 mg kg−1 of l-dopa + benserazide). *Indicates a significant (P < 0.05) difference from baseline in each group of mice. $Indicates a significant (P < 0.05) difference between both groups of mice.

With pergolide, we observed a biphasic effect: during the 0–2 h period, the amount of W was significantly decreased by a low dose (0.5 mg kg−1) but was increased by high doses (1, 2.5 and 5 mg kg−1) (Fig. 2). An opposite effect was observed for SWS over the same period (Fig. 2). A SWS rebound was observed during the 2–4 h period at a dose of 1 and 2.5 mg kg−1 (Fig. 2). Effect on SWS was accounted by a significant modification in SWS latency (Table 3). For the dose of 1, 2.5 and 5 mg kg−1 a significant decrease in PS amounts was induced during the 0–2 h period (Fig. 2), and was accounted by an increase in PS latency (Table 3) and a number of PS episodes reduction (Table 2). For both higher doses, this effect was lasting during the four first hours and the whole 8 h of recording (Table 1).

In MPTP mice, at doses of 2.5 and 5 mg kg−1, lisuride significantly reduced the amount of W and increased the amount of SWS during the 0–2 h post-injection period (Fig. 3). The lower dose (1 mg kg−1) significantly reduced the PS during the 2–4 h period (Fig. 3), the four first hours and the whole 8 h of recording (Table 1). Effects on PS amounts were accounted by a PS episodes number reduction (Table 2).

Comparison between control and MPTP groups

After saline injection in the l-dopa groups of mice, W amounts are significantly higher and SWS lower in MPTP mice compared with control mice during the 0–2 h period.

At a dose of 50 mg kg−1 of l-dopa, a group effect was found during the 0–2 h period for all vigilance states. MPTP mice showed more W and less SWS /PS compared with controls (Fig. 1). This difference was also found for the four first hours for W and SWS and lasting the whole 8 h of recording for W amounts (Table 1). For this dose, the increase in PS latency was significantly higher in MPTP mice than in controls (Table 3).

The two groups of mice did not differ in terms of the effects of pergolide on vigilance state amounts.

At a 1 mg kg−1 of lisuride, a group effect was found. MPTP mice showed significantly more W and less SWS for the 2–4 h period of recording than controls (Fig. 3).

Histological analysis

All mice were killed following completion of the recordings, i.e. between 31 and 54 days after the last injection of MPTP. The mean value for TH-positive neurons in the substantia nigra pars compacta was significantly lower in MPTP mice than in controls (306.6 ± 13.2 verses 438.4 ± 19.9, respectively, n = 16 and 18, P < 0.05), corresponding to a neuronal loss in MPTP mice of 30.1%. We found no difference between the two groups in terms of the number of stained neurons in the ventrotegmental area.

The striatal staining intensity was significantly lower in MPTP mice than in controls (n = 16 and 18, respectively, P < 0.05). The reduction in TH staining in MPTP mice was calculated to be 51.3%.

Discussion

The present study showed that dopaminergic drugs modify sleep and wakefulness in both control and MPTP mice. The inter-group comparisons did not reveal major differences. However, some significant dissimilarities following injection of levodopa and lisuride indicate that the responsiveness of mice with a partial DA deficit is not exactly the same as in controls.

The effects of l-dopa, pergolide and lisuride on S/W parameters in mice have not previously been evaluated. In our control mice, l-dopa reduced the quantity of PS during the first 4 h. In the rat, it has been shown that l-dopa induces significant decreases in the length and number of episodes of both PS and SWS (Galarraga et al., 1986). Here, in the control group, pergolide increased W and decreased SWS and PS. These effects are consistent with those described by Monti et al. (1988) in the rat. Lisuride had a sedative effect in our control group by decreasing W and raising SWS, with no significant effect on PS. In comparison, lisuride reportedly induces a sleep-like state in the chick (Ferrari and Giuliani, 1993), although polysomnography recording was not performed in this latter study.

Our results show that l-dopa and dopaminergic agonists have an effect on vigilance states in control mice. The significant effects of low-dose pergolide on vigilance states and the drug’s high-dose trends seemed to confirm the biphasic effect induced by D2 and D2/D1 agonists and which has been described in several studies (Cianchetti, 1985; Lagos et al., 1998; Monti et al., 1989; Ongini et al., 1993; Python et al., 1996). It appears that the D2 receptors are probably responsible for the drug’s low-dose effect, i.e. sedation. In fact, DA levels are reduced by low doses of pergolide – an effect which can be reversed by a D2 antagonist (Fuller et al., 1982) – and it has been suggested that low doses of l-dopa promote sleep principally via D2-like receptors (Andreu et al., 1999; Ferreira et al., 2002). Moreover, the D2 agonist and partial D1 antagonist lisuride has a sedative effect. Conversely, high doses effect of l-dopa and pergolide- (which potential affect D2 and D1 receptors) induced wakefulness, confirming the general observation that high dopaminomimetic doses enhance W and suppress SWS and PS – probably via postsynaptic D1-like and/or D2-like receptors (Lagos et al., 1998; Monti et al., 1989; Ongini et al., 1993).

One particular relevant finding in the present study is that l-dopa, pergolide and lisuride can decrease the amount of PS, thus suggesting a potential role for DA in PS regulation. Of the many studies showing the wakefulness-promoting effect of dopaminergic treatments, Nishino et al. (1998) suggested that the effect on PS is probably a consequence of the reduction in SWS. However, this idea is contradicted by our observation (in control mice) of an l-dopa reduction in PS in the absence of a modification in SWS. Likewise, Crochet and Sakai (2003) have shown that a D3 receptor agonist induces an increase in PS in the absence of any effect on SWS. Thus, a dopaminergic substance can modify PS without necessarily modifying W and SWS, suggesting in turn that DA may participate in the mechanisms regulating PS. Furthermore, the results of the study by Crochet and Sakai (2003)– showing an increase in the amount of PS after D3 injection agonist – prompt us to think that the effect of l-dopa in control mice seen here (i.e. a reduction in PS) does not involve activation of the D3 receptor.

Hence, to better understand the effects of DA on S/W, the investigation of vigilance state regulation in the MPTP mouse (a partial DA depletion model) can provide useful information. In a previous study, we studied S/W cycles in this model; MPTP mice displayed a greater total amount of PS over the nycthemeral period, compared with saline-treated control mice (Monaca et al., 2004). In the present study, during S/W recording after vehicle injection, we did not observe as much PS in MPTP mice (compared with controls) as we had found previously (Monaca et al., 2004). Nevertheless, it is important to note that the present recordings were made over 8 h, whereas the data in our previous work was collected over 48 consecutive hours. Furthermore, the baseline recordings were made after injection of vehicle (which may be a source of stress), while our 2004 study involved spontaneous recording without external stress (i.e. neither injection nor human intervention). This PS difference cannot be explained by disparities in DA neuronal loss, as the latter value was similar in both studies (29.8% in our 2004 study and 30.1% in the present work).

The effects of dopaminergic drugs in MPTP mice can differ from those observed in control mice. Wakefulness amounts are greater in MPTP mice after l-dopa (50 mg kg−1) and lisuride (1 mg kg−1) injections compared with control mice. This result suggests that these two compounds prompt a different response in our model. It is noteworthy that the dopaminergic ability of l-dopa (the DA precursor) is dependent on the presence of surviving DA neurons. Thus, the greater effect of l-dopa (and probably also that of lisuride) may reflect hyper-responsiveness of the surviving DA neurons after intoxication by the MPTP treatment. Few studies have explored sleep regulation in DA-depleted animal models with. Rotenone-treated mice presented a decrease in PS and SWS, whereas injection of l-dopa induced an increase in PS (Garcia-Garcia et al., 2005). However, interpretation of the latter study’s results is complicated by the unexpectedly intense effects on sleep parameters induced by the vehicle used to dissolve the active compound. In another study, Daley et al. (1999) showed that MPTP primates experienced greater sleepiness and presented sleep-onset in PS. These sleep abnormalities were reversed by l-dopa but not by pergolide. The results of our study are generally consistent with this latter finding, as l-dopa (but not pergolide) differs in its effects (on PS in particular) in MPTP mice relative to control mice. This disparity may be due to the differential affinity of each compound for DA- (or non-DA) receptors, although further studies involving concomitant injection of antagonists and receptor autoradiography would probably be required to provide an explanation. In contrast to pergolide (a DA agonist), the effect of l-dopa on S/W is comparable with that of endogenous DA. Hence, the greater effect of l-dopa on S/W (and on PS in particular) in MPTP mice suggests modification of the participatory role of DA in S/W regulation.

One remaining question is how DA actually participates in sleep regulation. Dopamine’s effects on S/W regulation are probably mediated by its ability to modulate the function of other neurotransmission systems. In fact, a number of studies have shown that noradrenergic neurons in the locus coeruleus, serotoninergic neurons in the raphe nuclei and cholinergic neurons in the pedunculopontine nuclei and the hypothalamus receive dopaminergic projections from the ventrotegmental area and substantia nigra pars compacta (for a review, see Monti and Monti, 2007); all these structures play a determining role in sleep regulation (Fuller et al., 2006; Jones, 2005; Stenberg, 2007). In agreement with the hypothesis whereby noradrenalin could be involved, Crochet and Sakai (2003) demonstrated in the cat that the PS-reducing effect of DA (when injected into the caudal peri-LC-alpha) is mediated by alpha-adrenoreceptors (for which DA is a potent agonist) rather than by the DA receptors themselves.

Furthermore, it is particularly noteworthy that hypocretin has also been shown to modify S/W (Bourgin et al., 2000; Chemelli et al., 1999; Nishino and Kanbayashi, 2005). One cannot rule out the possibility that the DA system is capable of modulating the function of hypothalamic hypocretin neurons. In fact, this hypothesis was reinforced by a study by Bubser et al. (2005), who showed that D1 and D2 receptor agonists modulated neuronal hypocretin activity. The latter authors observed the presence of D2 receptors on hypocretin neurons in the lateral hypothalamus of the rat. More recently, Yamanaka et al. (2006) showed that TH-reactive neurons can be found near hypocretin neurons. Moreover, application of DA to sections of lateral hypothalamus can hyperpolarize or stimulate hypocretin neurons, depending on the dose (Alberto et al., 2006; Yamanaka et al., 2006). Hence, these findings argue in favour of indirect effects of dopaminergic compounds on S/W regulation.

Conclusion

The present study highlights the role of DA in the regulation of both wakefulness and sleep. Moreover, our results suggest that partial DA depletion may modulate the effect of dopaminergic compounds on S/W in mice (even though there were relatively few inter-group differences). In fact, the putative role of DA lesions in the occurrence of S/W disorders may explain why the latter can occur in the early stages of PD (i.e. before the motor symptoms become apparent). Further studies are needed to better define the neurotransmission system(s) responsible for DA’s effects on S/W. To this end, it would be informative to test the effect of noradrenalin, hypocretin or acetylcholine neurotransmission blockade (via the use of antagonists, receptor knock-out or enzyme inhibitors) prior to administration of dopaminergic drugs.

Acknowledgements

We wish to thank the French Parkinson’s Disease Foundation for funding this work and Dr David Fraser for improving the English in the manuscript.

Ancillary