Address correspondence and reprint requests to M. A. Cenci, Department of Exp. Medical Science, Basal Ganglia Pathophysiology Unit, Lund University, BMC F11, 22184 Lund, Sweden. E-mail: Angela.Cenci_Nilsson@med.lu.se
We explored possible differences in the peripheral and central pharmacokinetics of l-DOPA as a basis for individual variation in the liability to dyskinesia. Unilaterally, 6-hydroxydopamine (6-OHDA) lesioned rats were treated chronically with l-DOPA for an induction and monitoring of abnormal involuntary movements (AIMs). Comparisons between dyskinetic and non-dyskinetic cases were then carried out with regard to plasma and striatal l-DOPA concentrations, tissue levels of dopamine (DA), DA metabolites, and serotonin. After a single intraperitoneal injection of l-DOPA, plasma l-DOPA concentrations did not differ between dyskinetic and non-dyskinetic animals, whereas peak levels of l-DOPA in the striatal extracellular fluid were about fivefold larger in the former compared with the latter group. Interestingly, the time course of the AIMs paralleled the surge in striatal l-DOPA levels. Intrastriatal infusion of l-DOPA by reverse dialysis concentration dependently induced AIMs in all 6-OHDA lesioned rats, regardless of a previous priming for dyskinesia. Steady-state levels of DA and its metabolites in striatal and cortical tissue did not differ between dyskinetic and non-dyskinetic animals, indicating that the observed difference in motor response to l-DOPA did not depend on the extent of lesion-induced DA depletion. These results show that an elevation of l-DOPA levels in the striatal extracellular fluid is necessary and sufficient for the occurrence of dyskinesia. Individual differences in the central bioavailability of l-DOPA may provide a clue to the varying susceptibility to dyskinesia in Parkinson's disease.
The dopamine (DA) precursor, l-DOPA, is still the mainstay of the treatment of Parkinson's disease (PD). Unfortunately, after a few years from treatment initiation, l-DOPA loses its ability to provide a stable symptomatic control, and most patients start to exhibit pronounced oscillations in mobility (‘wearing-off’ and ‘on–off’ fluctuations) and abnormal involuntary movements (dyskinesia) (for review see Marsden et al. 1981; Nutt 1992). There is ample agreement that dyskinesias and complex, unpredictable fluctuations are caused by plastic changes produced by l-DOPA downstream of the nigrostriatal DA neuron (Chase 1998; Obeso et al. 2004). This contention is strongly supported by findings obtained in 6-hydroxydopamine (6-OHDA) lesioned rats treated with l-DOPA. In this animal model, the development of dyskinesia is positively correlated with gene and protein expression changes in striatal neurons, and shows a less tight, non-linear relationship with the extent of nigral DA cell loss (Cenci 2002; Winkler et al. 2002). Interestingly, rats with the same degree of DA denervation, and receiving the same dose of l-DOPA, may show a large variability in developing dyskinesia and the associated plastic changes (Andersson et al. 1999; Winkler et al. 2002; Picconi et al. 2003). This variability suggests that some DA-denervated subjects are relatively protected from the dyskinetic action of l-DOPA by some yet unknown mechanisms. Recent studies in PD patients suggest that a predisposition to l-DOPA-induced dyskinesia may depend on presynaptic factors, such as the dynamics of striatal DA release following peripheral l-DOPA administration (de la Fuente-Fernandez et al. 2004) and/or the metabolic rate of putaminal DA (Rajput et al. 2004). The role of these factors, however, remains to be examined under tightly controlled experimental conditions.
The present study was undertaken in order to dissect pharmacokinetic and/or biochemical parameters that may account for a differential liability to l-DOPA-induced dyskinesia among rats. Unilateral injections of 6-OHDA in the ascending nigrostriatal DA pathway were followed by a course of l-DOPA treatment, during which animals were sorted in two motor response categories based on the presence or the absence of dyskinesia. Dyskinetic and non-dyskinetic animals were compared with regard to plasma and striatal l-DOPA concentrations after i.p. drug administration. In addition, tissue levels of DA, DA metabolites and serotonin (5-HT) were measured in the terminal fields of the nigrostriatal and mesocortical DA projection after a period of l-DOPA treatment washout. The most striking result emerging from these comparisons was a much larger increase in striatal l-DOPA levels in dyskinetic rats compared with non-dyskinetic cases following a peripheral injection of l-DOPA. This prompted us to verify the dyskinetic effects of intrastriatal l-DOPA delivery in a separate group of animals, comprising both l-DOPA-treated and l-DOPA-naïve subjects.
Materials and methods
The study was performed on female Sprague–Dawley rats (B&K Universal, Stockholm, Sweden; ∼225 g body weight at the beginning of the experiment). The animals were housed under a 12-h light/dark cycle. The treatment of the animals and their conditions had been approved by the Malmö-Lund Ethical Committee on Animal Research (permit no. M66/04).
Treatment groups and experimental design
This study was performed on rats with nearly complete (> 95%) unilateral 6-OHDA lesions of the ascending nigrostriatal DA pathway. The general time course of the experiments is illustrated in Fig. 1. The rats were treated chronically with l-DOPA (6 mg/kg methyl l-DOPA plus 12 mg/kg benserazide) for 2 weeks for an induction and monitoring of abnormal involuntary movements (AIMs). At the end of the treatment period, different groups of rats (each consisting of dyskinetic and non-dyskinetic cases) were allocated to three subexperiments. One group of rats (n = 10) was used for measuring changes in l-DOPA concentrations in the striatal extracellular fluid after a single i.p. injection of l-DOPA (central pharmacokinetics). A second group (n = 13) was used for studying l-DOPA concentrations in the blood (peripheral pharmacokinetics). Thereafter, this group of animals was given a 3-day l-DOPA washout prior to determining steady-state levels of DA, DA metabolites and 5-HT in the cortex and the striatum. A third group of rats (n = 6) was used for studying AIMs induction by intrastriatally infused l-DOPA (reverse microdialysis). The latter experiment also included a control group of drug-naïve 6-OHDA lesioned animals (n = 5).
l-DOPA methyl ester and the peripheral DOPA decarboxylase inhibitor benserazide-HCl were purchased from Sigma-Aldrich (Stockholm, Sweden). They were dissolved in physiological saline, and administered within 1 h at a volume of 1.0 mL/kg body weight by i.p. injection. The dose/injection used throughout the experiment was 6 mg/kg and 12 mg/kg for l-DOPA and benserazide, respectively. For the reverse microdialysis experiment, l-DOPA methyl ester was dissolved in Ringer's solution at four different concentrations, i.e. 10 μm, 1, 10 and 100 mm. Each concentration was infused intrastriatally for 60 min. The lowest concentration was infused for 45 min only, as it did not evoke any behavioural change.
Lesion surgery and amphetamine rotation
All the animals in this study were subjected to unilateral 6-OHDA lesions according to a standard procedure (see, for example, Cenci et al. 1998). Briefly, 6-OHDA-HCl (Sigma-Aldrich) was dissolved in 0.02% ascorbate–saline at a concentration of 3 μg/μL, and injected into the right medial forebrain bundle at the following coordinates (in mm) relative to bregma and the dural surface: (i) A = −4.4, L = −1.2, V = −7.8, tooth bar at −2.4 (7.5 μg); (ii) A = −4.0, L = −0.8, V = −8.0, tooth bar at +3.4 (6 μg). At 2 weeks post-lesion, rats were tested for amphetamine-induced rotation (2.5 mg/kg d-amphetamine i.p.; 90 min testing), and only animals showing individual means > 5 full turns per min in the direction ipsilateral to the lesion were selected for the study. This rotational score has been shown to correspond to > 90% depletion of DA fibre terminals in the striatum (Winkler et al. 2002).
Quantification of the abnormal involuntary movements (AIMs) induced by l-DOPA was carried out during the light hours (i.e. between 09.00 and 17.00 h) using a previously described rating scale (Lundblad et al. 2002; Winkler et al. 2002). Briefly, this scale considers three topographic subtypes of dyskinetic movements: (i) axial AIMs, i.e. twisting movements of the neck and upper body toward the side contralateral to the lesion; (ii) forelimb AIMs, i.e. repetitive jerks or dystonic posturing of the contralateral forelimb, and/or purposeless grabbing movement of the contralateral paw; (iii) orolingual AIMs, i.e. empty jaw movements and contralateral tongue protrusion. For each observation period (1 min), each AIM subtype is rated on a severity scale from 0 to 4 based on its duration and persistence (1 = occasional; 2 = frequent; 3 = continuous but interrupted by sensory distraction; 4 = continuous, severe, not interrupted by sensory distraction). An additional score is given to the amplitude of limb and axial dyskinesia. The amplitude of axial AIMs is rated according to the lateral deviation (or torsion) of an animal's neck and upper trunk from the longitudinal axis of the body. The amplitude of limb AIMs is scored based on the extent of limb translocation and on the visible involvement of proximal muscle groups. In addition to rating the above dyskinetic behaviours, we also scored asymmetric locomotion (“locomotive aims”) on the scale from 0–4, which is based on the frequency and duration of the observed behaviour. We have previously shown that locomotive aims do not provide a specific measure of dyskinesia in the rat (Lundblad et al. 2002).
In order to select dyskinetic and non-dyskinetic rats, AIMs were rated on five occasions during the 2 weeks of daily l-DOPA treatment (Fig. 1). During these sessions, AIMs were rated every 20 min for 3 h following i.p. l-DOPA injection. At the end of the treatment period, rats were classified as either ‘dyskinetic’ or ‘non-dyskinetic’ prior to further investigation. Non-dyskinetic cases comprised animals that had scored 0 on the rat AIM scale in all testing sessions. Rats classified as dyskinetic had reached a severity grade ≥ 2 on at a least one topographic subtype of AIMs (axial, limb, or orolingual) by the last testing session. In the central pharmacokinetics study, rats were rated on the AIMs scale as per our standard procedure (i.e. 1 min observation period every 20 min) simultaneously with the collection of the microdialysis samples. In the reverse microdialysis experiment, AIMs were rated every 5th min for the entire duration of the experiment (approximately 4 h). A global AIM score was calculated by multiplying the severity and amplitude scores of each AIM subtype at every time point, and then adding these products for the entire testing session.
In vivo microdialysis
Microdialysis probes (Agnthos Microdialysis, Stockholm, Sweden) had a concentric design, an outer diameter of 0.5 mm, and 4 mm of exposed membrane length. The in vitro recovery for l-DOPA was approximately 30%. For the central pharmacokinetic study, probe-implantation surgery was performed using gaseous anaesthesia (1.2% halothane air mixture). For the reverse dialysis experiment, probe implantation was performed using injectable anaesthetics, namely, a mixture of Hypnorm (Janssen Pharmaceutical, Bersee, Belgium) and Dormicum (Hoffman-La Roche, Basel, Switzerland) in sterile water (1 : 1 : 2; 2.7 mL/kg i.p.) as a method of anaesthesia. Upon completing surgery, rats were given analgesic (Temgesic, 0.3 mg/kg s.c.) and anti-sedative treatment (Anti-sedan, 1 mg/kg s.c.; Apoteksbolaget AB, Stockholm, Sweden). In preliminary experiments we had found that these two methods of anaesthesia did not reduce each rat's individual AIM scores to a standard dose of l-DOPA (6 mg/kg, i.p.), when dyskinesia was tested at one or two days post-surgery, respectively (data not shown). Probes were implanted in the right caudate–putamen at the following coordinates (in mm, relative to bregma and to the dura surface), A = +0.6, L =−3.5, V = −6.0 (tooth bar at −3.3). The probe was fixed to the cranial bone with jeweller screws and dental cement. The microdialysis sessions were carried out 1 or 2 days after probe implantation in the central pharmacokinetic study and in the reverse microdialysis experiment, respectively. On the day of the experiments, the rats were placed in custom-made plexiglass boxes (38 × 38 × 40 cm) and the probes were connected through a swivel to a syringe infusion pump (CMA/100; CMA/Microdialysis, Solna, Sweden) via a length of polyethylene tubing. Probes were perfused with Ringer's solution (3 mm KCl; 145 mm NaCl; 1.3 mm CaCl2; 1 mm MgCl2; pH 6.0) at a constant rate of 2 μL/min or 4 μL/min for the central pharmacokinetics study and reverse microdialysis study, respectively. For the central pharmacokinetics study, three dialysate samples (20 min each) were collected prior to an i.p. injection of l-DOPA in order to provide baseline values. After the administration of l-DOPA (6 mg/kg; combined with benserazide, 12 mg/kg), samples were collected, and dyskinesia was scored, every 20 min for 120 min.
For reverse microdialysis of l-DOPA, the probes were first rinsed for 45 min with Ringer's solution, and then perfused with four increasing concentrations of l-DOPA methyl ester (10 μm, 45 min; 1 mm, 60 min; 10 mm, 60 min; and 100 mm, 60 min) in Ringer's solution. After discontinuing the l-DOPA infusion, the probes were rinsed with Ringer's solution for 20 min before allowing the animals to recover in their home cages.
Another group of animals were anaesthetized using light gas anaesthesia (1.2% halothane/air mixture), and the jugular vein was carefully exposed on the right side of the neck. A silicone tube was introduced for about 5 cm through a small incision in the vein wall, placing its ending in the vena cava. When this catheter was correctly positioned, heparine (50 U/kg body weight; Apoteksbolaget) was given through the catheter to prevent blood coagulation. After 5 min, one blood sample was taken to provide baseline determinations. Immediately thereafter, 6 mg/kg l-DOPA (in combination with benserazide, 12 mg/kg) was given i.p. and blood samples were collected at 10, 20, 40, 80 and 120 min after the injection. Every sample was 200 μL in volume. An additional dose of 50 U/kg heparine was given at 60 min after the l-DOPA injection.
Every sample was immediately centrifuged at 10 000 g for 1 min and instantly frozen on powdered dry ice. After the 120th-min sample, the animals were given temgesic (0.3 mg/kg s.c.) for post-operative analgesia. The jugular vein was then ligated, and the skin wound sutured, before letting the rats recover from the gas anaesthesia.
Rats from the microdialysis experiments were killed 1–2 days after the microdialysis sessions in order to verify the placement of the probes. The animals were deeply anaesthetized with sodium pentobarbitale (240 mg/kg i.p; Apoteksbolaget) and transcardially perfused with 50 mL saline at room temperature (18–20°C), followed by 250 mL ice-cold, phosphate-buffered 4% paraformaldehyde (pH 7.2–7.4). The brains were post-fixed for 2 h in the same fixative, and transferred to 25% sucrose for 24 h. Coronal sections through the striatum (40 μm) were cut on a sliding microtome, mounted on chromalum-coated slides and stained with Cresyl violet. The probes were found to be correctly positioned in the lateral striatum in all cases. Animals that had been used for the peripheral pharmacokinetics study were given 3 days of l-DOPA treatment washout, and were then killed by decapitation. The brains were put in a ‘precision brain slicer’ (BS-4000S, Braintree Scientific, Braintree, MA, USA) to obtain coronal sections of the brain at a standardized thickness. The following slices were taken, 0–7 mm from the anterior pole of the brain (prefrontal cortex); 7–11 mm (striatum); and 14–17 mm (substantia nigra; data from the latter structure will not be presented because of technical failure). The structure of interest was rapidly dissected out from the corresponding slice (on ice) and tissue samples were kept frozen (−80°C) until analysis.
Determination of levels of l-DOPA in brain extracellular fluid and blood
Blood samples were prepared as described in Igarashi et al. (2003). In short, 60 μL of 0.4 m perchloric acid and 40 μL internal standard (α-methyldopa, 100 ng/mL in 0.4 m perchloric acid) was added to 50 μL plasma. The tubes were shaken for 1 min and then centrifuged for 10 min, and filtered (0.45 μm pore filter; Sartorius AG, Goettingen, Germany).
Twenty-five microlitres of each microdialysis and plasma sample were injected into an HPLC apparatus coupled to an electrochemical detection system (Antec Leyden, Zoeterwoud, the Netherlands). The flow cell was equipped with a glass carbon working electrode, and an Ag/AgCl reference electrode. The potential was set at +0.65 V. The mobile phase [0.1 m NaH2PO4 adjusted to pH 2.5 with H3PO4, EDTA 30 mg/L, 550 mg/L (w/v) octyl sulfate sodium, 15% (v/v) methanol] was delivered at a flow rate of 230 μL/min to a reverse phase column (4.6 mm Ø, 150 mm length; Chrompack, Oxford, UK). This method specifically measures l-DOPA levels while dopamine is not detected. The assay's detection limit for l-DOPA was 25 fmol.
Determination of dopamine, DOPAC and 5-HT in tissue samples
Levels of dopamine, dihydroxyphenylacetic acid (DOPAC) and 5-HT in tissue samples were also determined by HPLC coupled with electrochemical detection. First, the samples were sonicated in 300 μL of cold 0.4 m perchloric acid containing dihydroxybenzylamine as an internal standard. The samples were centrifuged for 5 min at 15 000 g and the supernatant transferred to 0.22-μm micropure separators (Amicon, Beverly, MA, USA) and spun at 15 000 g for 1 min. Filtrate (10 μL) was injected into a reverse phase C18 column (4.6 mm Ø, 150 mm length, Chrompack). The mobile phase (5 g/L of sodium acetate, 100 mg/L octanesulfonic acid, 30 mg/L EDTA, 12% methanol, pH 4.2) was delivered at a rate of 0.5 mL/min. The detection limits of the assay were 5, 8 and 16 fmol for DA, DOPAC and 5-HT, respectively.
In the pharmacokinetic studies, changes in l-DOPA levels at different post-injection intervals were compared between dyskinetic and non-dyskinetic rats using repeated measure anova and a post-hoc Tukey test. In the reverse microdialysis of l-DOPA, global AIMs scores were computed for each l-DOPA concentration and infusion period (60 min), and compared using two-factor anova and a post-hoc Tukey test. In all the remaining biochemical assays, comparisons were carried out where appropriate using the Student's t-test. Correlations between AIM scores and biochemical parameters were examined using linear regression. Statistical significance level was set at p < 0.05. All data are expressed as group means ± 1 SEM.
Peripheral and central pharmacokinetics of l-DOPA
Well-matched sets of animals, both consisting of dyskinetic and non-dyskinetic cases, were used for l-DOPA determinations in the venous blood and in the striatal extracellular fluid, respectively.
After an i.p. injection of l-DOPA (6 mg/kg, combined with 12 mg/kg benserazide) plasma concentrations of l-DOPA showed a significant increase (time effect, p < 0.0001) and reached peak levels (approximately 1200 nm) at 10–20 min (Fig. 2a). At these time points, plasma l-DOPA concentrations differed significantly from baseline in both the dyskinetic and the non-dyskinetic rats (Fig. 2b). The peripheral concentrations of l-DOPA did not differ between the two groups of animals at any of the time points examined (group effect, p = 0.926; time and group interaction, p = 0.352).
Under baseline conditions, the extracellular levels of l-DOPA in the striatum were below the limit of detection of the assay (namely 25 fmol). After an i.p. drug injection (6 mg/kg methyl l-DOPA combined with 12 mg/kg benserazide), the extracellular l-DOPA levels showed a detectable rise in both dyskinetic and non-dyskinetic rats, reaching peak values within 40 min (time effect, p = 0.0047). The increase in striatal l-DOPA levels was much more pronounced and sustained in animals affected by dyskinesia compared with non-dyskinetic cases (group effect, p = 0.0061; time and group interaction, p = 0.0474; Fig. 3a). At the peak of the l-DOPA surge (i.e. 40 min post-injection) the levels of this amino acid amounted to, on average, 250 nm and 50 nm in the dyskinetic and non-dyskinetic group, respectively (p < 0.001). The levels of l-DOPA declined rapidly, but were still significantly detectable in the striatal extracellular fluid at 120 min post-injection. Interestingly, the change in striatal l-DOPA levels in the dyskinetic animals paralleled closely the time course of their abnormal involuntary movements (Fig. 3b).
Reverse dialysis of l-DOPA
We next asked whether an increase in striatal l-DOPA levels can by itself trigger abnormal movements, and whether such an effect would be conditioned by a previous priming for dyskinesia. To address this question, we infused increasing concentrations of l-DOPA in the lateral (sensorimotor) part of the striatum in rats that had previously been treated with l-DOPA (sorted in dyskinetic and non-dyskinetic subgroups) and in saline-treated, 6-OHDA lesioned controls. The intrastriatal infusion of l-DOPA induced dyskinetic movements in all the tested animals in a concentration-dependent manner (Fig. 4a; concentration effect, p < 0.0001; concentration and group interaction, p = 0.2595), although the AIM scores were overall larger in the dyskinetic group (group effect, p = 0.003). The relationship between l-DOPA concentration and AIMs induction showed a similar pattern in the three experimental groups. The lowest concentration of l-DOPA (10 μm) was subthreshold for inducing dyskinesia in any of the groups, and only one animal exhibited some occasional and mild orolingual and locomotive AIMs (severity grade 1) by 45 min of 10 μm l-DOPA infusion. After increasing the l-DOPA concentration to 1 mm, all animals started to show mild dyskinesia (grade 1–2) and the AIM scores reached a plateau by approximately 50 min of infusion. The first signs of dyskinesia had a stereotypic-like character. They affected mainly the head, neck and forepaw and included head nodding, contralateral deviation of the head and neck, and forepaw-tapping movements (Video S1). The AIMs increased in severity, amplitude and duration with increasing l-DOPA concentrations, extending from the head, neck and forepaw to also involve the rest of the body (Video S2). This pattern was seen in all experimental groups, although dyskinetic animals already showed an overall greater AIM severity and pronounced axial dyskinesia at lower l-DOPA concentrations (i.e. 1 mm; Fig. 4b). At a 100-mm concentration, the l-DOPA-induced AIMs had prominent dystonic features that affected the whole trunk, forelimb and hindlimb on the side contralateral to the infusion (Video S3).
In both dyskinetic and non-dyskinetic animals, tissue levels of DA and its metabolite, DOPAC, were reduced by over 98% on the side of the striatum ipsilateral to the lesion compared with the contralateral intact side (p < 0.01 in all comparisons; Table 1). These data are in agreement with the strong DA-depleting effect produced by the type of 6-OHDA lesion used here (Schmidt et al. 1983). Striatal levels of DA, DOPAC, or DA/metabolite ratio (which is an index of DA turnover) did not differ significantly between dyskinetic and non-dyskinetic rats on either side of the brain. Striatal 5-HT levels were not significantly different between the lesioned and the intact side of the striatum, nor between dyskinetic and non-dyskinetic rats (Table 1). Biochemical determinations were carried out also in the prefrontal cortex in order to sample one important terminal field of the mesocorticolimbic projection originating from the ventral tegmental area [which is partially affected by our 6-OHDA lesion procedure; (Andersson et al. 1999)]. While cortical levels of DA did not differ significantly between dyskinetic and non-dyskinetic rats, 5-HT levels were threefold larger in the former group compared with the latter one (p < 0.05), and showed a significant, positive correlation with the rats' AIM scores (R = 0.71; p = 0.0070; Fig. 5a). By contrast, no significant relationship was found between the rat AIM scores and 5-HT levels in the striatum (Fig. 5b).
Table 1. Striatal levels of dopamine, dihydroxyphenylacetic acid (DOPAC), and serotonin (5-HT) and cortical levels of 5-HT and dopamine (cortical DOPAC levels were below the limits of detection of the assay)
Dyskinetic (n = 9)
Non-dyskinetic (n = 5)
All the rats in the study had received unilateral injections of 6-OHDA in the right medial forebrain bundle. Tissue samples were collected after 3 days of l-DOPA treatment washout in the same rats that had been used for the peripheral pharmacokinetics study. *p < 0.05 versus left (intact) side within the same group (paired t-test); †p < 0.05 versus right (lesioned) side in the non-dyskinetic group (unpaired t-test).
Unilaterally, 6-OHDA lesioned rats treated with l-DOPA show a marked individual variability in their predisposition to develop abnormal involuntary movements. We have shown that this variability in motor response is closely related to different levels of up-regulation of ΔFosB and prodynorphin mRNA in striatal neurons (Cenci et al. 1998; Andersson et al. 1999). However, the question has remained unsolved as to why different rats may or may not develop molecular and behavioural signs of dyskinesia in response to the same pharmacological treatment. The present study was undertaken in order to establish the role of presynaptic factors in the motor response profile to l-DOPA. Rats with 6-OHDA lesions were treated chronically with l-DOPA and evaluated on the rat AIM scale. Dyskinetic and non-dyskinetic rats were then used to examine the concentrations of l-DOPA in the venous blood or in the striatum following a challenge drug injection. While peripheral levels of l-DOPA did not differ between dyskinetic and non-dyskinetic rats, peak levels of l-DOPA in the striatal extracellular fluid were about fivefold larger in animals that expressed abnormal movements compared with animals that did not. Interestingly, the time course of the dyskinetic movements paralleled the surge in striatal l-DOPA levels. Prompted by these observations, we asked whether a rise in striatal l-DOPA levels is by itself sufficient to trigger dyskinetic movements. We found that the intrastriatal infusion of l-DOPA induced AIMs in all the tested rats in a concentration-dependent manner, regardless of any previous exposure to l-DOPA and priming for dyskinesia. Taken together, the present data indicate that an elevation of striatal l-DOPA levels is both necessary and sufficient for the expression of dyskinesia.
Following peripheral administration, the central pharmacokinetics of l-DOPA is potentially influenced by many factors, such as the capacity for intestinal or mucous absorption of the amino acid, its transport from plasma to brain across the blood–brain barrier, and the rate of central l-DOPA decarboxylation and metabolism (Pardridge and Oldendorf 1975; Wade and Katzman 1975a,b; Horne et al. 1984; Contin et al. 1993). As plasma levels of l-DOPA did not differ between dyskinetic and non-dyskinetic animals, we can exclude that the observed difference in central l-DOPA levels was caused by a differential absorption of the amino acid from the mucous peritoneal cavity to the blood. The passage of l-DOPA from blood to brain is controlled by a saturable carrier for large neutral amino acid that is present at the brain capillary endothelium (Pardridge and Oldendorf 1975; Wade and Katzman 1975b). Once in the brain, l-DOPA is rapidly converted to DA by the enzyme, aromatic l-amino-acid decarboxylase (Horne et al. 1984). Even large nigrostriatal DA lesions do not compromise the brain capacity for l-DOPA decarboxylation, which occurs not only in nigrostriatal neurons, but also in serotonergic terminals (Lopez et al. 2001) and in cells intrinsic to the striatum (Melamed et al. 1981). The rate of l-DOPA decarboxylation does not depend on the concentration of l-DOPA in the brain extracellullar fluid (Wade and Katzman 1975b), and is unlikely to explain the large difference in striatal l-DOPA levels between dyskinetic and non-dyskinetic rats. Indeed, if the higher l-DOPA levels in the dyskinetic group were because of a slower rate of DOPA decarboxylation, striatal DA levels would have to be lower in the same group. In contrast to this prediction, extracellular DA levels are much larger in dyskinetic rats compared with non-dyskinetic cases following i.p. administration of a dose of l-DOPA similar to that used in the present study (Ahn et al. 2004). We therefore suggest that the higher striatal l-DOPA levels found in dyskinetic animals compared with non-dyskinetic cases are likely to reflect a greater capacity for l-DOPA transport from blood to brain.
The present data provide further validation to the concept that l-DOPA-induced rat AIMs are the functional equivalent of peak-dose dyskinesia in Parkinson's disease (for a discussion see Cenci et al. 2002). Peak-dose dyskinesia appears when brain levels of l-DOPA and DA are high (Nutt 1992; Contin et al. 1993). In a central pharmacokinetic study in patients, the onset of dyskinesia occurred simultaneously with a rise in l-DOPA concentrations in the ventricular cerebrospinal fluid (Olanow et al. 1991). These data are in line with our results, showing a very close match in the temporal evolution of rat AIMs and striatal l-DOPA concentrations after an i.p. drug injection.
The causal role of striatal l-DOPA in dyskinesia was demonstrated in the reverse microdialysis study. This study showed, in addition, a relationship between the physical features of dyskinesia and the amount of l-DOPA that was infused in the striatum. Lower l-DOPA concentrations elicited stereotypic-like movements of the head, neck and forepaw, while higher l-DOPA concentrations caused overtly abnormal movements with a broader body distribution and a pronounced dystonic character. Also, in patients with Parkinson's disease, brain levels of l-DOPA and DA seem to condition the phenomenology of dyskinesia. Peak-dose dyskinesias are typically characterized by choreiform and choreo-dystonic movements with a broad distribution, which are alleviated by decreasing the dose of l-DOPA (for review see Nutt 1992; Fahn 2000). Biphasic dyskinesias, which appear at the beginning and the end of the l-DOPA concentration-effect curve, most typically manifest as repetitive, stereotypic movements affecting the legs (Luquin et al. 1992). The finding that lower intrastriatal concentrations of l-DOPA elicited stereotypic-like movements in the rat fits with our previous (unpublished) observation of stereotypic gnawing and grooming being particularly common up to 20 min, and from 100 min after l-DOPA administration, i.e. when striatal l-DOPA levels are rising or declining. In our experience, l-DOPA-induced stereotypes are not associated with a significant striatal up-regulation of ΔFosB, which provides a hallmark of dyskinesia in the rat, and in other animal models of PD (reviewed in Cenci et al. 2002). The concentration of intrastriatally infused l-DOPA did not have an impact on the relative representation of turning behaviour (locomotive AIMs), i.e. the main motor manifestation observed in 6-OHDA lesioned rats in studies using higher systemic doses of l-DOPA (Papa et al. 1994). The relatively scarce representation of locomotive AIMs in our experiment may depend on the location of the microdialysis probe, which was placed in a region controlling limb and head movements as opposed to locomotion (see Andersson et al. 1999 for a further illustration of this concept). Interestingly, l-DOPA-treated and drug-naïve (saline-treated) animals showed similar patterns of dyskinetic movements at different l-DOPA concentrations, although the evoked AIMs were overall most severe in dyskinesia-primed cases. In particular, the dyskinetic animals showed a broad involvement of axial and proximal limb muscles at a relatively low l-DOPA concentrations (1 mm). A higher sensitivity to l-DOPA in dyskinesia-primed rats is not surprising, as the development of dyskinesia is associated with an increased sensitivity of DA receptor signalling (Aubert et al. 2005), and with an up-regulation of ‘pro-dyskinetic’ genes in the striatum (Cenci et al. 1998; Andersson et al. 1999). Taken together, our results indicate that a previous priming for dyskinesia lowers the threshold for the dyskinetic action of l-DOPA (i.e. it shifts the l-DOPA–dyskinesia response curve to the left), but is not an absolute requirement for this movement disorder, provided that l-DOPA is delivered directly to the striatum.
In addition to the pharmacokinetic and microdialysis studies, we examined tissue levels of DA, DOPAC and 5-HT in dyskinetic and non-dyskinetic rats in the striatum and the frontal cortex. As this analysis was carried out after a period of l-DOPA washout, the measured parameters are assumed to reflect steady-state monoamine levels and DA metabolism in the tissue. DA levels in the striatum and the frontal cortex (which are a target of nigrostriatal and mesocorticolimbic DA projections, respectively) were found not to differ between dyskinetic and non-dyskinetic rats. These data are in agreement with counts of tyrosine–hydroxylase-positive cells in the midbrain DA cell groups, showing that the pattern and/or extent of DA denervation does not differ significantly between dyskinetic and non-dyskinetic rats in the same animal model (Andersson et al. 1999). The DA metabolite, DOPAC, provides an index of the metabolic rate of released DA, and the DOPAC/DA ratio is regarded as a measure of DA turnover (Zigmond et al. 1990). Striatal DOPAC levels and DOPAC/DA ratio tended to be higher in dyskinetic rats than in non-dyskinetic cases, although this trend did not reach significance. Unexpectedly, the only biochemical marker that differed significantly between the two experimental groups was the content of 5-HT in the frontal cortex. On the side ipsilateral to the lesion, cortical levels of 5-HT were threefold larger in the dyskinetic group compared with the non-dyskinetic cases, showing a significant, positive correlation with the rats' AIM scores. The frontal cortex receives an abundant, and predominantly ipsilateral projection from both the dorsal and the median raphe nuclei (for review see Jacobs and Azmitia 1992). Ascending 5-HT projections run in close proximity of, and partly within, the medial forebrain bundle (Jacobs and Azmitia 1992), and are affected by the 6-OHDA lesion procedure used here (not shown). The present data may therefore suggest that non-dyskinetic animals had a higher degree of lesion-induced 5-HT damage. However, this suggestion remains speculative at this point, particularly because striatal 5-HT levels did not differ significantly between dyskinetic and non-dyskinetic rats. Further experimental work is therefore required to examine the plasticity of ascending 5-HT projections in 6-OHDA-lesioned rats with or without l-DOPA treatment.
In conclusion, the present study indicates that the amount of l-DOPA present in the striatal extracellular fluid is the single most important ‘risk factor’ for the occurrence of dyskinesia in this experimental model. Our data fit with the results of a recent positron emission tomography study in human Parkinson's disease, showing that the increase in putaminal DA release 1 h post-l-DOPA administration is much more pronounced in dyskinetic patients compared with non-dyskinetic cases (de la Fuente-Fernandez et al. 2004). We propose that individual differences in the central bioavailability of l-DOPA may provide a clue to the varying susceptibility to dyskinesia in Parkinson's disease.
The authors warmly thank Kerstin Beirup for her excellent technical supervision in the blood sampling experiment, and Prof. Tadeusz Wieloch for his generous help. The study was supported by grants from the Swedish National Research Council, Elsa and Thorsten Segerfalk Foundation, The Swedish Parkinson Association, and the Craaford Foundation (to MAC). MC was supported by a Marie Curie Host Training Fellowship.