مدارک و شواهد از ارتباط بین خواب و دیسکینزی ناشی از لوودوپا در مدل حیوانی بیماری پارکینسون

Levodopa-induced dyskinesia (LID) represents a major challenge for clinicians treating patients affected by Parkinson's disease (PD). Although levodopa is the most effective treatment for PD, the remodeling effects induced by disease progression and the pharmacologic treatment itself cause a narrowing of the therapeutic window because of the development of LID. Although animal models of PD provide strong evidence that striatal plasticity underlies the development of dyskinetic movements, the pathogenesis of LID is not entirely understood. In recent years, slow homeostatic adjustment of intrinsic excitability occurring during sleep has been considered fundamental for network stabilization by gradually modifying plasticity thresholds. So far, how sleep affects on LID has not been investigated. Therefore, we measured synaptic downscaling across sleep episodes in a parkinsonian animal model showing dyskinetic movements similar to LID. Our electrophysiological, molecular, and behavioral results are consistent with an impaired synaptic homeostasis during sleep in animals showing dyskinesia. Accordingly, sleep deprivation causes an anticipation and worsening of LID supporting a link between sleep and the development of LID.

Levodopa (LD)–induced dyskinesia (LID) is difficult to treat, negatively affects quality of life, and increases the treatment costs of Parkinson's disease (PD) patients (Dodel et al., 2001). Although LD currently represents the most effective treatment for PD patients (Poewe et al., 2010), ameliorating cardinal signs such as bradykinesia, akinesia and rigidity (Katzenschlager and Lees, 2002), these benefits are in some measure overshadowed by the emergence of LID (Jankovic and Stacy, 2007). In the early stages, PD patients usually experience an acceptable quality of life that is impaired in the advanced stages by the emergence of these involuntary movements that frequently occur at the peak of the LD effect (Berthet and Bezard, 2009 and Nadjar et al., 2009). In other words, when the patients experience motor complications such as a shortening motor response and the development of dyskinesia, the delivery of LD without inducing dyskinesia becomes increasingly difficult (Jankovic and Stacy, 2007 and Olanow et al., 2006). Several attempts have been made to find “pure” anti-dyskinetic drugs, able to uncouple the anti-akinetic effect from the dyskinetic one of LD and other dopamine (DA) agonists. As a consequence, great efforts have been made to clarify LID pathogenesis, emphasizing the role of pulsative stimulation of striatal receptors by dopaminergic treatment (Olanow et al., 2006), and also changes of postsynaptic (Calabresi et al., 2010 and Jenner, 2008) or presynaptic processes at the striatal level (Lunardi et al., 2009). Although no conclusive results on LID pathogenesis have been achieved, disease duration, that is, the degree of dopaminergic degeneration, along with long-term use of LD, seems to play a crucial role (Berthet and Bezard, 2009 and Nadjar et al., 2009). This is perceivable in 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine–induced parkinsonism in humans in whom the extended dopaminergic damage caused within a few days of LD treatment the development of LID undistinguishable from those in the idiopathic form (Langston and Ballard, 1984). Promising results with respect to LID pathogenesis come from animal models and PD patients demonstrating an uncontrolled long-term potentiation (LTP) of corticostriatal synapses in association with the development of LID (Calabresi et al., 2000 and Picconi et al., 2003). Striatal medium spiny neurons of dyskinetic rats have an impaired reversal of the previously induced LTP, that is, “depotentiation” (Picconi et al., 2003). Furthermore, PD patients showing LID were unresponsive to the depotentiation paradigm suggesting that depotentiation is abnormal in the motor cortex of patients with LID (Huang et al., 2011). Dyskinetic movements can be interpreted as intruding motor programs in selected neuronal responses because of unrestrained LTP. Physiological control of spike-timing–dependent plasticity (STDP) such as LTP has a crucial role in network stability (Turrigiano and Nelson, 2004). Recently, an important role for sleep and in particular for electroencephalogic (EEG) slow-wave activity (SWA) has been proposed. According to the concept of synaptic homeostasis (SH) hypothesis, SWA is believed to underlie an overnight homeostatic process, by which neuronal circuit activity is not driven toward runaway excitation by LTP (Tononi and Cirelli, 2014). Moreover, SWA spectral density analysis represents a reliable measure of synaptic downscaling during nonrapid eye movement (NREM) sleep (Tononi and Cirelli, 2014 and Vyazovskiy et al., 2007) and recent compelling evidence associated SWA to cortical synaptic strength and therefore measuring the reduction in SWA from early (the first 1 hour at the beginning of the light period) to late (the last 1 hour of the light period) NREM sleep has often been used as a direct measure of SH (Vyazovskiy et al., 2007). STDP and homeostatic plasticity involve, in different ways, the trafficking of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR) by different calcium dynamics. Some signaling molecules are linked to postsynaptic expression of AMPARs, and among these, the Arc protein seems to be a critical mediator of this cellular phenomenon (Béïque et al., 2011). A physiological decrease in Arc has been demonstrated during sleep (Hanlon et al., 2009), and an increased level of the Arc protein has been shown in the striatum of dyskinetic rats (Sgambato-Faure et al., 2005). Indeed, sleep disturbances are frequently associated with PD, often preceding the motor onset of the disease itself and revealing the intimate role of DA in regulating the sleep-wake cycle (Dzirasa et al., 2006). Furthermore, improvement of motor performance in the morning is frequently mentioned by patients, principally in those with a long history of the disease and motor fluctuations (Bateman et al., 1999). Parallel to the disease progression, a progressive change in the physiological sleep architecture is observed, which includes a significant decrease in the amount of NREM sleep (Diederich et al., 2005 and Rye and Jankovic, 2002). In the present article, we investigated whether the development of LID might be associated with a critical impairment of synaptic downscaling during NREM, implying sustained retention of aberrant synaptic memories within the corticobasal ganglia-thalamic loop. To examine our hypothesis, we measured the synaptic strength in the following 5 animal groups: (i) sham-lesioned drug-naive rats (sham/LD−), (ii) sham-lesioned LD-treated rats (sham/LD+), (iii) 6-hydroxydopamine (6-OHDA)–lesioned drug-naive rats (PD/LD−), (iv) 6-OHDA–lesioned LD-treated rats without dyskinesia (PD/DYS−), and (v) 6-OHDA–lesioned LD-treated rats with dyskinesia (PD/DYS+). We found that PD/DYS− animals had a physiological sleep-induced reduction of synaptic strength, whereas PD/DYS+ animals do not manifest an adequate reduction in SWA, thus possibly implying reduced synaptic downscaling during sleep. Concordantly, behavioral experiments demonstrated that sleep deprivation (SD) was able to enhance the emergence of LID during LD treatment (Cam et al., 2013).

Our data provide the first evidence of an impaired SH in parkinsonian animals with dyskinesia, supporting the idea that sleep and its underlying mechanisms play a crucial role in the appearance of LID. During the course of PD, the occurrence of an inadequate downscaling of corticocortical and corticostriatal synaptic strength may alter the physiological function of basal ganglia. As a consequence, there may be an imbalanced information storage within the corticobasal ganglia-thalamic loop that produces persisting and widespread activity presenting as multiple interfering motor programs. From a clinical point of view, dyskinetic PD patients may benefit from compounds enhancing synaptic downscaling during NREM sleep.