An increasing body of experimental evidence demonstrates that the glutamatergic system is involved in the genesis of
l-3,4-dihydroxyphenylalanine (
L-DOPA)-induced
dyskinesia (LID). Indeed, the
N-methyl-d-aspartate (
NMDA) receptor antagonist
amantadine is the only anti-dyskinetic compound used in patients, albeit with limited efficacy and side effects. In this study, we investigated the anti-dyskinetic properties of
memantine, a non-competitive
NMDA receptor antagonist in clinical use for the treatment of
dementia, in the 6-hydroxy-dopamine (6-OHDA)-lesion rat model of
Parkinson's disease. For comparison, parallel experiments were also performed with
amantadine. First, we investigated the acute effect of different doses of
memantine (5, 10, 15 and 20mg/kg), and
amantadine (10, 20, 40, 60mg/kg) on established
dyskinesia induced by
L-DOPA (6mg/kg plus
benserazide). Results showed that both
memantine and
amantadine produced a significant reduction of LID. Afterward, drug-naïve and
L-DOPA-primed 6-OHDA-lesioned rats were sub-chronically treated with daily
injections of
L-DOPA (6mg/kg plus
benserazide) alone, or in combination with the effective doses of
memantine, while
amantadine was tested in already dyskinetic rats. Results showed that
memantine significantly dampened
dyskinesia in both drug-naïve and
L-DOPA-primed rats, but only during the first few days of administration. In fact, the anti-dyskinetic effect of
memantine was completely lost already at the fifth administration, indicating a rapid induction of tolerance. Interestingly, a 3-week washout period was not sufficient to restore the anti-dyskinetic effect of the drug. Similarly,
amantadine was able to dampen already established
dyskinesia only during the first day of administration. Moreover,
memantine partially decreased the
therapeutic effect of
L-DOPA, as showed by the result of the stepping test. Finally, loss of the anti-dyskinetic effect of
memantine was associated to increased synaptic GluN2A/GluN2B ratio at striatal synaptic membranes. Our results are in line with clinical observations suggesting that
NMDA receptor blockade may only be transiently effective against LID in PD patients.