l-DOPA-induced dopamine efflux in the striatum and the substantia nigra in a rat model of Parkinson’s disease: temporal and quantitative relationship to the expression of dyskinesia


  • 1

    These authors shared first authorship.

Address correspondence and reprint requests to Angela Cenci-Nilsson, Basal ganglia Pathophysiology, BMC F11, SE-221 84 Lund, Sweden. E-mail: angela.cenci_nilsson@med.lu.se


J. Neurochem. (2010) 112, 1465–1476.


l-DOPA-induced dyskinesia in Parkinson’s disease is associated with large increases in brain dopamine (DA) levels following drug dosing, but the precise significance of this phenomenon is not understood. Here we compare DA efflux and metabolism in the striatum and the substantia nigra in dyskinetic and non-dyskinetic animals following a standard dose of l-DOPA. Rats with 6-hydroxydopamine lesions were treated chronically with l-DOPA, monitored on the abnormal involuntary movements scale, and then subjected to intracerebral microdialysis under freely-moving conditions. Following s.c. l-DOPA injection, peak extracellular DA levels in both striatum and substantia nigra were about twice as large in dyskinetic animals compared to non-dyskinetic rats. This effect was not attributable to differences in DOPA levels or DA metabolism. The larger DA efflux in dyskinetic animals was blunted by 5-HT1A/5-HT1B receptor agonists and tetrodotoxin infusion, reflecting release from serotonin neurons. Striatal levels of serotonin and its main metabolite, 5-hydroxyindolacetic acid were indeed elevated in dyskinetic animals compared to non-dyskinetic rats, indicating a larger serotonergic innervation density in the former group. High DA release was, however, not sufficient to explain dyskinesia. The ‘abnormal involuntary movements output’ per unit concentration of striatal extracellular DA was indeed much larger in dyskinetic animals compared to non-dyskinetic cases at most time points examined. The present results indicate that both a high DA release post-l-DOPA administration and an increased responsiveness to DA must coexist for a full expression of dyskinesia.

Abbreviations used:



5-Hydroxyindoleacetic acid






abnormal involuntary movements




DA transporter


3,4-dihydroxyphenylacetic acid


homovanillic acid


l-DOPA-induced dyskinesia


median forebrain bundle


Parkinson’s disease


substantia nigra




ventral tegmental area

Treatment of Parkinson’s disease (PD) with l-DOPA is associated with a high incidence of dyskinesia (abnormal involuntary movements, AIMs) and motor fluctuations (reviewed in Obeso et al. 2000; Cenci and Lundblad 2006; Cenci and Odin 2009). Large and transient increases in brain dopamine (DA) levels induced by the medications have been attributed a prime causative role in these motor complications (Chase 1998; Bezard et al. 2001; Olanow 2004; Cenci and Lindgren 2007). The ensuing pulsatile stimulation of DA receptors would determine abnormal postsynaptic responses, such as changes in striatal gene and protein expression (Andersson et al. 1999; Tel et al. 2002; Konradi et al. 2004) and abnormal firing patterns (Alonso-Frech et al. 2006; Meissner et al. 2006). In support of this hypothesis, high extracellular levels of DOPA (Carta et al. 2006) and DA (Meissner et al. 2006; Lee et al. 2008) have been measured in the striatum in l-DOPA-treated, dyskinetic rats relative to untreated control animals. Moreover, following administration of a standard l-DOPA dose, significantly higher DA efflux has been seen in the putamen in PD patients exhibiting l-DOPA-induced dyskinesia (LID) compared to non-dyskinetic subjects (de la Fuente-Fernandez et al. 2004; Pavese et al. 2006). In post-mortem biochemical studies, indexes of DA turnover are significantly elevated in PD patients with motor complications compared to stable l-DOPA-responders (Rajput et al. 2004). Taken together, these data provide strong evidence that LID is associated with pre-synaptic abnormalities in DA release and/or metabolism. The mechanisms of such abnormalities may rely on the uptake and conversion of exogenous l-DOPA by serotonin neurons (Carta et al. 2007), which can store and release DA, but lack pre-synaptic autoregulatory mechanisms for DA (reviewed in Cenci and Lundblad 2006).

In this study, we have used intracerebral microdialysis and biochemical assays in order to compare indexes of DA and serotonin (5-HT) neurotransmission in 6-hydroxydopamine (6-OHDA)-lesioned and chronically l-DOPA-treated rats that had, or had not, developed AIMs during the treatment. In addition to the striatum, we investigated the substantia nigra pars reticulata (SN), one of the output nuclei of the basal ganglia showing pronounced changes in neural firing patterns and microvascular remodelling in this animal model of LID (Meissner et al. 2006; Westin et al. 2006). The microdialysis experiments were carried out in awake and freely moving rats in order to correlate the neurochemical data to the behavioural response to l-DOPA.

Materials and methods


Female Sprague-Dawley rats (220–225 g; Harlan Laboratories, An Venray, The Netherlands) were housed in a 12 h light/dark cycle with food and water ad libitum. The experiments were approved by local ethical committees on animal research in the Malmö-Lund and Gothenburg districts. A total of 79 rats with stable 6-OHDA lesions (8–10 weeks post-surgery) and eight intact rats were used in the study.

6-Hydroxydopamine lesions

6-Hydroxydopamine (6-OHDA-HCl; Sigma-Aldrich Sweden AB) was injected into the right median forebrain bundle (MFB) at two coordinates according to our established procedures (Cenci et al. 1998). Two weeks postoperatively, amphetamine-induced rotation (2.5 mg/kg d-amphetamine, i.p., Sigma-Aldrich Sweden AB) was applied to select rats with > 90% striatal DA depletion (corresponding to > 5 full turns/min ipsilateral to the lesion in a 90-min session; Winkler et al. 2002; Carta et al. 2006). The lesion extent was verified in the analysis of striatal DA tissue contents by HPLC.

Experimental design

A total of 57 rats with 6-OHDA lesions as well as eight intact rats were used in two microdialysis experiments. The first experiment explored differences in extracellular and tissue concentrations of DA and 5-HT, and their corresponding metabolites between l-DOPA-treated dyskinetic (n = 24) and non-dyskinetic animals (n = 8), comparing them to unlesioned drug-naïve rats (‘normal controls’, n = 8). The second experiment addressed the mechanisms of l-DOPA-induced DA efflux in dyskinetic animals (n = 25) using 5-HT1A/1B receptor agonists and/or tetrodotoxin (TTX) infusion. Two additional groups of 6-OHDA-lesioned, l-DOPA/saline-treated rats were used to verify the extent of midbrain DA and 5-HT denervation by immunohistochemistry (n = 16), or to examine the effects of DA transporter (DAT) blockade on l-DOPA-induced DA efflux and AIMs (n = 6). Results from these experiments are shown in Appendix S2 and S4, respectively.

Drug treatments

l-DOPA (L-3, 4-dihydroxyphenylalanine methyl ester hydrochloride), benserazide-HCl, and TTX were purchased from Sigma-Aldrich Sweden AB. The 5-HT1A and B agonist, 8-Hydroxy-N,N-dipropyl-2-aminotetralin (8-OH-DPAT) and CP94253, respectively, were bought from Tocris Bioscience, Bristol, UK. All drugs but TTX were dissolved in physiological saline, and administered at the volume of 1.0 mL/kg body weight by s.c. injection. Throughout the study, l-DOPA was given at the dose of 6 mg/kg, combined with 12 mg/kg benserazide, in single daily injections for 14 days. With this treatment paradigm, approximately 80% of the rats are expected to develop AIMs during the treatment period, whereas 20% will remain non-dyskinetic (reviewed in Cenci and Lundblad 2007). Drug-naïve intact controls were treated with saline (0.9% NaCl) for the same treatment period. In experiment 2, the 5-HT1A and -B agonists were given at three different dose combinations that had been reported to reduce LID in this rat model, i.e. 0.1 and 1.75 mg/kg of 8-OH-DPAT and CP94253 (high dose, n = 3); 0.05 and 1.0 mg/kg of 8-OH-DPAT and CP94253 (medium dose, n = 6), or 0.035 and 0.75 mg/kg of 8-OH-DPAT and CP94253 (low dose, n = 6) (Carta et al. 2007). The highest dose of the 5-HT1A/B agonists (0.1 and 1.75 mg/kg of 8-OH-DPAT and CP94253) induced a severe serotonin syndrome, i.e. flat body posture associated with motor depression (attributable to over-stimulation of post-synaptic 5-HT1A receptors (Goodwin et al. 1986; Yamada et al. 1988), which lasted for approximately 60 min. Thereafter the rats started to move and displayed dyskinetic movements (severity grade 2–3) that lasted for the remaining testing period. For this reason, the highest dose combination was not tested further, and only data from 0.05/1.0 mg/kg (‘medium dose’) and 0.035/0.75 mg/kg (‘low dose’) will be reported.

Dyskinesia ratings

The development of l-DOPA-induced AIMs was monitored during 14 days of l-DOPA treatment according to our standard procedures. Following drug administration, each rat was observed every 20th minute for 3 h and scored on the basic AIM severity scale (Cenci et al. 1998; Lundblad et al. 2002) from 0 to 4 on three subtypes of dyskinesia (axial, limb, and orolingual). In addition, the amplitude of these AIMs subtypes was rated on a scale from 0 to 4 (Cenci and Lundblad 2007). On each monitoring period, the basic severity score was multiplied by the amplitude score for each of the three AIM subtype (thus, the maximum possible AIM score per monitoring period was 48). Global AIMs were defined as the sum of these products for an entire testing session (the theoretical maximum value being 432). l-DOPA-treated rats were classified as either ‘dyskinetic’ or ‘non-dyskinetic’ prior to the microdialysis experiment. The non-dyskinetic group comprised animals with either none or mild and occasional AIMs (basic severity grade 0–1 on each AIM subtype), whereas rats classified as dyskinetic had developed moderate to severe AIMs (basic severity grade ≥ 2 on at least two of the three AIM subtypes). These classification criteria conform to those applied in our previous studies, in which we have shown that animals with AIM severity grade ≤ 1 do not exhibit molecular markers of LID, such as the activation of extracellular signal-regulated kinases 1/2, and the up-regulation of ΔFosB and prodynorphin mRNA in striatal neurons (Andersson et al. 1999; Westin et al. 2007). In addition, rats with AIM severity grade ≤ 1 do not exhibit a significant angiogenic response to l-DOPA treatment (Westin et al. 2006).

Microdialysis probes and implantation surgery

Following 14 days of l-DOPA (or saline) treatment, rats were anaesthetized with isoflurane (Forene®, Abbot Scandinavia AB, Sweden), and mounted on a stereotaxic frame for the implantation of microdialysis probes in the striatum (AP: +0.6, ML: −3.5, DV: −6.0) and SN (AP: −5.3, ML: −2.2, DV: −8.6) on the side ipsilateral to the lesion. The dialysis membrane (20 000 kDa cut-off; Filtral 16, Hospal Ind., Meyzien, France) had an active length of 3 or 2 mm in the striatum and the SN, respectively. Rats received post-surgical analgesia with ketoprofen (5 mg/kg s.c.; Romefen® Vet, Merial, France) and were allowed to recover single-housed before the microdialysis sessions.

In vivo microdialysis

Two days (∼48 h) after probe implantation, rats were placed in custom-made plexiglass boxes (38 × 38 × 40 cm). The probes were connected through a swivel to a syringe infusion pump (CMA/Microdialysis AB, Stockholm, Sweden) via a length of polyethylene tubing. Probes were perfused with Ringer’s solution (140 mM NaCl, 3.0 mM KCl, 1.2 mM CaCl2 and 1.0 mM MgCl2) at a constant rate of 2 μL/min, and perfusate samples were collected every 20 min (40 μL/sample). Probes were rinsed for 1 h, and then three baseline samples were collected prior to l-DOPA injection. After administering l-DOPA (6 mg/kg, combined with benserazide, 12 mg/kg, s.c.), AIMs were scored every 20th minute simultaneously with dialysate collection during the drug active period (3 h). In the second experiment, rats were pre-treated with the 5-HT agonists, 8-OH-DPAT and CP94253, 5 min before l-DOPA. In an additional group of rats, the striatal and nigral probes were perfused with TTX (1 μM in Ringer solution) from 60 min before through 3 h after the administration of l-DOPA. These rats also received co-treatment with the low dose of the 5-HT1A/1B agonists (i.e. 0.035/0.75 mg/kg 8-OH-DPAT/CP94253). Twenty-four hours after completing the microdialysis session, rats were decapitated, brains were sliced and macroscopically inspected to exclude intracerebral haemorrhages or aberrant probe locations. In experiment 2, where the SN region was dissected out for biochemical assays, correct nigral probe placement was inferred from the position of the probe track in the remaining midbrain tissue. Perfusate fractions and striatal and nigral tissue samples were analysed through HPLC as described in Appendix S1. The detection limit of the assay, estimated as 2 × noise level, was 0.01–0.03 and 3–5 fmol/μL, respectively, fmol/μL for DA and 5-HT, and 0.2–0.4 fmol/μL for their corresponding metabolites.

Statistical analysis

All data are presented as group means ± SEM. The extracellular concentrations of DOPA, DA, 5-HT, and their corresponding metabolites are expressed as fmol/μL in the perfusate samples without correction for recovery across the dialysis membrane. Behavioural data and perfusate concentrations were analyzed using repeated-measures anova. In the first experiment, group comparisons were carried out on the following time points: baseline (average of three baseline samples), peak (40 min after l-DOPA administration) and end phase (average of 140–180 min) of the l-DOPA action-curve. In the second experiment, the effects of pharmacological treatments on l-DOPA-induced AIMs and DA efflux were examined with repeated measures anova on all time points (20–180 min) post-l-DOPA injection. Global AIM scores and cumulative DA concentrations were calculated for the 0–180 min interval post-l-DOPA injection, and analysed by one-factor anova. Tissue monoamine contents of DA, 5-HT and 5-Hydroxyindoleacetic acid (5-HIAA) were compared among groups and hemispheres by two-factor anova. In each data set, relevant pair-wise comparisons were carried out using post hoc Tukey′s honestly significant difference (HSD) test. The threshold for statistical significance was set at α level, 0.05.


Effects of l-DOPA administration on dyskinetic behaviour

The AIM scores following l-DOPA injection during the first microdialysis experiment are shown in Fig. 1(a). Rats previously classified as dyskinetic promptly exhibited severe AIMs, whereas non-dyskinetic animals showed motor activation occasionally associated with AIMs of low severity, and normal controls remained still or asleep during most of the experiment (Fig. 1a: F2,37 = 27.39, pgroup < 0.0001, ptime < 0.001, and pinteraction < 0.001).

Figure 1.

 The effect of l-DOPA administration on dyskinetic behaviour and DA efflux. (a) Time course of l-DOPA-induced AIMs during the microdialysis experiment. Dyskinetic rats (n = 24) showed pronounced AIMs from 20 to 140-min post-injection, whereas non-dyskinetic cases (n = 8) only displayed occasional and mild AIMs, and did not differ significantly from intact rats (‘normal’; n = 8). §< 0.05 vs. normal animals, #< 0.05 vs. non-dyskinetic group (n = 8) (Repeated Measures anova and post hoc Tukey’s HSD test). (b) The difference between peak and baseline DA levels (ΔDA) was larger in dyskinetic animals compared to non-dyskinetic rats in both the striatum and the SN. §< 0.05 vs. normal animals, #< 0.05 vs. non-dyskinetic group (one-factor anova and post hoc Newman-Keuls’ test). (c) and (d) DA efflux in the striatum and the SN. The peak of extracellular DA (40-min post-l-DOPA-injection) was larger in dyskinetic animals compared non-dyskinetic rats in both the striatum (c) and the SN (d). *< 0.05 vs. baseline within the same group, #< 0.05 vs. non-dyskinetic animals and §< 0.05 vs. normal animals. (e) and (f) Relationship between AIMs and DA efflux (AIM scores per unit concentration of extracellular DA). For the same amount of extracellular DA in striatum (e) and SN (f), dyskinetic rats showed a significantly larger dyskinetic motor output than did non-dyskinetic animals. *< 0.05 vs. 20 min, #< 0.05 vs. non-dyskinetic animals (Repeated Measures anova and post hoc Tukey’s HSD test).

Effects of l-DOPA administration on striatal and nigral DA efflux

Extracellular DA concentrations differed significantly between groups and time points in both the striatum (Fig. 1c; F2,37 = 12.03, pgroup < 0.001, ptime < 0.001, pinteraction = 0.001) and the SN (Fig. 1d; F2,37 = 3.28, pgroup = 0.049, ptime < 0.001, pinteraction = 0.030). In the striatum, baseline DA concentrations were reduced by approximately 99% in both dyskinetic and non-dyskinetic animals (∼0.04 fmol/μL) compared to intact control animals (∼4 fmol/μL) (Fig. 1c). In the SN, on the other hand, baseline extracellular DA did not differ significantly between any of the groups, amounting to 0.1–0.2 fmol/μL (Fig. 1d). After s.c. l-DOPA injection, extracellular DA concentrations increased in both structures in all the three groups examined, reaching a peak by 40 min. An increase in nigral and striatal DA levels was seen also in normal animals, which is in line with findings from other groups (Orosz and Bennett 1992; Jonkers et al. 2001; Bianco et al. 2008; Lee et al. 2008) (but see Miller and Abercrombie 1999). At the peak time point, extracellular DA levels in both structures were approximately 80% larger in dyskinetic animals compared to non-dyskinetic rats (< 0.05; Fig. 1c and d). In the striatum, peak extracellular DA concentrations were similar in dyskinetic rats and normal controls, while being significantly lower than control values in the non-dyskinetic group (Fig. 1c). The difference between peak and baseline DA levels (ΔDA; Fig. 1b) was significantly larger in the dyskinetic animals compared to the other two groups (one factor anova, F2,37 6.65, = 0.003, < 0.05 for dys. vs. both non-dys and normal animals). The situation was different in the SN, where peak DA concentrations tended to be higher than normal in the dyskinetic rats, and approached normal values in the non-dyskinetic animals (Fig. 1d). The increment in DA levels between baseline and peak values (ΔDA; Fig. 1b) was significantly larger in dyskinetic rats compared to non-dyskinetic animals, but it did not differ from normal controls values in either of the groups (one factor anova, F2,36 = 3.96, = 0.028).

Extracellular concentrations of DOPA, DOPAC, Homovanillic acid (HVA), and the DA turnover index are shown in Appendix S3. Briefly, these parameters did not differ significantly between dyskinetic and non-dyskinetic rats in either the striatum or the SN (figure in Appendix S3).

After the microdialysis experiment, the extent of lesion-induced DA depletion was assessed in striatal and nigral tissue samples from dyskinetic and non-dyskinetic animals. This analysis revealed a reduction of DA levels by > 99% and 95% in the striatum and the SN, respectively, on the side ipsilateral to the 6-OHDA lesion compared to control values (p < 0.05 for dys. and non-dys. cases vs. normal rats). There was however no difference between dyskinetic and non-dyskinetic rats in either structure (further information in the footnote of Table 1).

Table 1.   Striatal and nigral tissue levels of DA, 5-HT and 5-HIAA
  1. Tissue samples were collected 24 h after the end of the microdialysis session. The 6-OHDA lesion had produced a nearly complete DA depletion in the ipsilateral striatum in both dyskinetic and non-dyskinetic animals compared to both the contralateral intact side and normal control values (pgroup < 0.001, pside < 0.001, and pinteraction < 0.001, F2,96 = 51.37). Tissue samples from normal controls were not available for the SN, but a comparison with the intact side revealed > 90% DA depletion on the lesioned side in both dyskinetic and non-dyskinetic rats, with no significant difference between groups (pgroup = 0.48, pside < 0.001, and pinteraction = 0.37, F1,79 = 0.82). The 6-OHDA lesion also had caused partial damage to the ascending serotonin projections. In both dyskinetic and non-dyskinetic animals, striatal tissue contents of 5-HT and 5-HIAA were significantly reduced on the side ipsilateral to the lesion compared to both the contralateral side and normal control values (striatum, 5-HT: pgroup < 0.027, pside < 0.001 and pinteraction < 0.001, F2,96 = 16.37; 5-HIAA: pgroup = 0.001, pside = 0.001 and pinteraction < 0.001, F2,96 = 14.46). In the SN, the levels of 5-HT were reduced on the side ipsilateral to the lesion in both dyskinetic and non-dyskinetic animals, with no significant difference between the two groups (pgroup = 0.69, pside = 0.009, and pinteraction = 0.75, F1,79 = 0.10). In contrast to 5-HT, levels of 5-HIAA were not reduced on the lesioned side in the SN (pgroup = 0.78, pside = 0.59, and pinteraction = 0.55, F1,79 = 0.36). *< 0.05 vs. intact side in the same group, #< 0.05 vs. non-dyskinetic animals, and §< 0.05 vs. normal animals (pairwise group comparisons on the same side; Two-factor anova and post hoc Tukey’s HSD test).

 DA0.08 ± 0.03*§10.37 ± 0.620.04 ± 0.01*§8.80 ± 0.7110.71 ± 0.5110.45 ± 0.64
 5-HT0.30 ± 0.03*§#0.44 ± 0.020.11 ± 0.03*§0.50 ± 0.020.39 ± 0.020.40 ± 0.02
 5-HIAA0.47 ± 0.05#0.57 ± 0.030.13 ± 0.06*§0.56 ± 0.030.63 ± 0.040.50 ± 0.02
 DA0.11 ± 0.01*2.04 ± 0.620.15 ± 0.03*1.69 ± 0.13  
 5-HT1.38 ± 0.162.29 ± 0.311.61 ± 0.192.32 ± 0.27  
 5-HIAA2.53 ± 0.303.00 ± 0.292.96 ± 0.433.06 ± 0.26  

Relationship between DA efflux and dyskinetic behaviour

To define the ‘dyskinesia output’ per unit concentration of extracellular DA, the AIM score per time point was divided by the DA levels in the corresponding perfusate fraction. This parameter showed very pronounced differences between groups and time points (Fig. 1e and f; striatum: F1,16 = 36.13, SN: F1,12 = 16.59; pgroup < 0.001, ptime < 0.001, pinteraction < 0.001 in both structures). Dyskinetic rats exhibited very severe AIMs as soon as DA levels started to increase (p < 0.05 vs. non-dys. group in the striatum at the 20-min point), and continued to show severe dyskinesia up to 80–160 min post-l-DOPA injection, when extracellular DA concentrations decreased rapidly in all animals. At these later time points, the difference between dyskinetic and non-dyskinetic rats was largest (3- to 15-fold; < 0.05 in both structures).

Biochemical indexes of serotonergic innervation in dyskinetic and non-dyskinetic animals

Extracellular concentrations of 5-HT and its metabolite, 5-HIAA, showed significant overall group differences in both the striatum (Fig. 2a, 54-HT: F2,37 = 13.75, pgroup < 0.001, ptime = 0.39, pinteraction = 0.82; Fig. 2c, 5-HIAA: F2,37 = 4.99, pgroup = 0.012, ptime = 0.001, pinteraction = 0.39) and the SN (Fig. 2b, 5-HT: F2,37 = 6.48, pgroup = 0.004, ptime = 0.004, pinteraction = 0.90; Fig. 2d, 5-HIAA: F2,37 = 7.48, pgroup = 0.002, ptime = 0.003, pinteraction = 0.17). Compared to non-dyskinetic rats, dyskinetic animals showed significantly higher extracellular levels of both 5-HT and 5-HIAA in the striatum (Fig. 2a and c; p < 0.05 for dys vs. non-dys animals), though not in the SN (Fig. 2b and d). Also in the tissue samples collected at the end of the study (Table 1), striatal levels of 5-HT and 5-HIAA were higher in dyskinetic rats (70–80% of control values) than in non-dyskinetic animals (20–30%; p < 0.05). In contrast, 5-HT and 5-HIAA levels did not differ significantly between the two groups in the SN (further information in the footnote of Table 1).

Figure 2.

 Extracellular levels of 5-HT and 5-HIAA. In both the striatum (a) and the SN (b), extracellular levels of 5-HT were overall lower in intact animals (n = 8) compared to all the rats with 6-OHDA lesions (dyskinetic or non-dyskinetic rats, n = 24 and 8, respectively). The opposite applied to 5-HIAA, whose levels were higher in normal controls in both structures (striatum, c; SN, d). Interestingly, striatal levels of both 5-HT and 5-HIAA were higher in dyskinetic animals compared to non-dyskinetic cases (a, c) whereas no significant difference was seen in the SN (b, d). *< 0.05 vs. baseline, #< 0.05 vs. non-dyskinetic animals, § and §§< 0.05 vs. normal animals (§§ indicates that both dyskinetic and nondyskinetic animals differed significantly from the normal controls) (Repeated Measures anova and post hoc Tukey’s HSD test).

Modulation of l-DOPA-induced DA efflux by 5-HT1A and B receptor agonists

Pharmacological experiments were undertaken in order to ascertain the origin of l-DOPA-induced DA efflux in dyskinetic animals. A DAT blocker (GBR12909, 20 mg/kg s.c.) had no effect on extracellular DA levels post-l-DOPA (Appendix S4). Treatment with the 5-HT1A and 5-HT1B receptor agonists, 8-OH-DPAT and CP94253 was tested based on its purported capacity to attenuate l-DOPA-induced DA efflux (Carta et al. 2007). These agonists were given at either of two dose combinations (‘low dose’: 0.035/0.75 mg/kg; ‘medium dose’: 0.05/1.0 mg/kg of 8-OH-DPAT and CP94253) prior to a s.c. injection of l-DOPA.

The time course and severity of l-DOPA-induced AIMs were markedly affected by the 5-HT1A/B agonists (Fig. 3a; F3,21 = 6.95, ptreatment = 0.004, ptime < 0.001, and pinteraction < 0.001). Both dose combinations were equally effective in inhibiting the expression of dyskinesia at 20- to 100-min post-injection (p < 0.05 for l-DOPA-only vs. the 5-HT agonists), but not at later time points (Fig. 3a). The overall effect of the 5-HT1A/B agonists is shown in Fig. 3(b). Both dose combinations reduced the global AIM scores by approx. 60% compared to l-DOPA-only treatment (Fig. 3b; one-factor anova: F3,21 = 6.64, = 0.003).

Figure 3.

l-DOPA-induced AIMs upon 5-HT1A/1B agonist co-treatment. The 5-HT1A and 5-HT1B agonists, 8-OH-DPAT (‘dpat’) and CP94253 (‘cp’), at two dose combinations (low dose, 0.035 and 0.75 mg/kg; medium dose, 0.05 and 1.0 mg/kg, of dpat and cp, respectively) reduced the AIMs scores 20–100 min after l-DOPA administration (a). The global AIM score during the testing session also was significantly reduced by both dose combinations (b). Reverse dialysis of TTX (1 μM in the perfusion medium) further reduced the AIMs peak at 40-min post-l-DOPA (a), but not the global AIM scores (b) when compared to 5-HT1A/B agonist treatment alone. In (a): *< 0.05 vs. 20 min, #< 0.05 vs. low dose cp/dpat (n = 6), §< 0.05 vs. medium dose cp/dpat (n = 6), and °< 0.05 vs. cp/dpat low + TTX group (n = 4) (Repeated Measures anova and post hoc Tukey’s HSD test). In (b): ♦< 0.05 vs. l-DOPA-only group (n = 9) (one-factor anova and post hoc Student-Newman-Keuls’ test).

In parallel to attenuating dyskinesia, treatment with the 5-HT1A/B agonists significantly reduced l-DOPA-induced DA efflux in both the striatum and the SN (Fig. 4). The effect was more pronounced in the striatum (F3,21 = 6.10, ptreatment = 0.004, ptime < 0.001, and pinteraction < 0.001), where both dose combinations significantly lowered extracellular DA concentrations at 40- to 60-min post-l-DOPA injection (Fig 4a; ∼50% and 80% reduction with ‘low’ and ‘medium’ dose, respectively, at 40 min; < 0.05 for each dose combination vs. l-DOPA-only), whereas DA levels at 80- to 180-min post-injection were not significantly reduced. Extracellular DA concentrations also were affected in the SN (Fig. 4c; F3,21 = 3.21, ptreatment = 0.044, ptime < 0.001, and pinteraction = 0.011). However, the low dose of the 5-HT1A/B agonists was without effect in this structure, while the medium dose significantly reduced nigral DA concentrations at 40 and 60-min post-injection to an extent similar to that observed in the striatum (< 0.05 for medium dose vs. l-DOPA-only) (Fig. 4c). The overall effects of the 5-HT1A/B agonists on l-DOPA-induced DA efflux are shown in Fig. 4(b) (striatum: F3,21 = 6.10, = 0.004) and 4(d) (SN: F3,22 = 3.21, = 0.04).

Figure 4.

 Striatal and nigral DA efflux upon 5-HT1A/B agonist co-treatment and TTX infusion. A combination of the 5-HT1A and 5-HT1B agonists, 8-OH-DPAT (‘dpat’) and CP94253 (‘cp’), respectively, dose-dependently reduced the surge of extracellular DA between 40 and 60- to 80-min post-l-DOPA administration (striatum, a; SN, c). In the SN, the reduction was however significant only with the higher dose combination (‘medium dose’, 0.05 and 1.0 mg/kg dpat and cp, respectively). Reverse dialysis with TTX (1 μm) further reduced DA levels at 40–60 min compared with the low dose of the 5-HT1A/1B agonists, but did not differ significantly from the higher 5-HT1A/1B dose combination. The overall effect of the treatment on the cumulative DA efflux post-l-DOPA in shown in (b) (striatum) and (d) (SN). In (a) and (c): *< 0.05 vs. baseline, #< 0.05 vs. low dose cp/dpat (n = 6), §< 0.05 vs. medium dose cp/dpat (n = 6), and °< 0.05 vs. cp/dpat low + TTX group (n = 4) (Repeated Measures anova and post hoc Tukey’s HSD test). In (b) and (d): ♦p < 0.05 vs. l-DOPA-only group (n = 9), °p < 0.05 vs. cp/dpat low + TTX group (one-factor anova and post hoc Student-Newman-Keuls’ test).

We next asked if the residual DA efflux that occurred in the presence of 8-OH-DPAT/CP94253 reflected impulse-dependent DA release. We therefore infused the sodium channel blocker, TTX (1 μM) in both the striatum and the SN during the microdialysis experiment in a group of rats that were pre-treated with the low dose of the 5-HT1A/B agonists (0.035 and 0.75 mg/kg of 8-OH-DPAT and CP94253, respectively) before l-DOPA administration. Extracellular DA concentrations in both striatum and SN were further reduced by the combined TTX infusion and agonist-treatment (< 0.05 for TTX vs. low-dose 5-HT 1A/B agonists or l-DOPA-only at 40- to 80-min post-injection; Fig. 4a and b). However, in neither structure did TTX reduce extracellular DA concentrations significantly below the levels measured after treatment with the higher dose of the agonists (0.05/1.0 mg/kg 8-OH-DPAT/CP94253) (Fig. 4a and c). Moreover, TTX infusion did not significantly reduce the cumulative DA efflux compared to the 5-HT1A/1B agonist treatment alone (Fig. 4b and d).


Positron-emission tomography imaging studies in PD patients have highlighted the relationship between LID and large changes in putaminal DA levels following l-DOPA administration (de la Fuente-Fernandez et al. 2004; Pavese et al. 2006). In keeping with the human data, microdialysis studies in l-DOPA-treated dyskinetic rats have described high striatal levels of DA during the expression of AIMs, although comparisons with non-dyskinetic animals were not reported (Meissner et al. 2006; Lee et al. 2008). The mechanisms contributing to a large rise in extracellular DA levels ‘on’l-DOPA have not been resolved, and several suggestions have been put forward by previous studies, including loss (Miller and Abercrombie 1999) or reversal/dysfunction (Ahn et al. 2004; Lee et al. 2008; Sossi et al. 2009) of the DAT, an altered rate of DA metabolism (Meissner et al. 2006), increased passage of l-DOPA from blood to brain (Carta et al. 2006; Westin et al. 2006), and non-regulated DA efflux from serotonin neurons (Kannari et al. 2001; Carta et al. 2007).

This study was undertaken in order to clarify the relationship between LID and DA release and metabolism in the brain. Although this topic has been the focus of other microdialysis studies, our work presents many important novel aspects, namely: (i) a comparison between animals that show AIMs and those that remain free from dyskinesia with the same regime of l-DOPA treatment; (ii) an analysis of the temporal and quantitative relationships between AIM scores and brain DA efflux; (iii) a comparison between indexes of DA and 5-HT transmission in the same samples of extracellular fluid or brain tissue; (iv) a comparison between the striatum and the SN. Indeed, while only the striatum was examined in previous microdialysis study on LID, motor effects of l-DOPA are mediated at both striatal (Carta et al. 2006) and nigral levels (Robertson and Robertson 1989). Our results show that a peripheral injection of l-DOPA produces a large surge of extracellular DA in both the striatum and the SN, and that peak DA concentrations are significantly higher in dyskinetic compared to non-dyskinetic rats. This large difference in peak DA concentrations could not be ascribed to group differences in either central DOPA levels or DA turnover rate (cf. figure in Appendix S3), nor to a reversal of DAT function, previously proposed by Ahn et al. (2004) (cf. figure in Appendix S4). Even though we have previously found higher striatal concentrations of l-DOPA in dyskinetic compared to non-dyskinetic rats (Carta et al. 2006), this finding was not reproduced here. We attribute the discrepancy to technical differences between the studies (different animals, types of probes, analytical procedures, and routes of l-DOPA administration). While higher central levels of l-DOPA would obviously result in larger production of DA, the present data demonstrate that differences in extracellular DA levels between dyskinetic and non-dyskinetic animals need not reflect a differential bioavailability of l-DOPA in the brain extracellular fluid, pointing to the existence of more complex regulatory mechanisms.

A growing body of literature indicates that, when nigrostriatal DA neurons are damaged, 5-HT neurons become the main contributor to the conversion of exogenous l-DOPA to DA (Arai et al. 1994, 1995). Indeed, serotonin neurons can decarboxylate l-DOPA to DA, store DA in synaptic vesicles, and release it together with 5-HT, but they lack both DA autoreceptors and the DAT (Arai et al. 1995; Tanaka et al. 1999; Kitahama et al. 2007). Neurotransmitter release from 5-HT neurons is strictly regulated by 5-HT1A and 5-HT1B autoreceptors, which are located on the cell bodies in the raphe nucleus (5-HT1A) or on 5-HT axonal terminals (5-HT1B) (reviewed in Blier et al. 1998). Because these autoreceptors cannot sense changes in extracellular DA concentrations, DA release from 5-HT neurons is not subjected to any homeostatic control mechanism, pre-disposing to motor complications and LID (Carta et al. 2007; Dupre et al. 2007, 2008; Eskow et al. 2007). The importance of serotonergic neurons to the production and release of DA has been extensively documented in rats with 6-OHDA lesions (Tanaka et al. 1999; Kannari et al. 2001, 2006). In this animal model, a lesion of the serotonergic system results in an 80% attenuation of striatal DA concentrations after acute l-DOPA injection (Tanaka et al. 1999). Moreover, pre-treatment with the 5-HT1A autoreceptor agonist, 8-OH-DPAT, reduces l-DOPA-derived DA efflux in the striatum (Kannari et al. 2001). We examined the combined effect of 5-HT1A and 5-HT1B receptor agonists on l-DOPA-induced DA efflux at doses that have been proposed to reduce LID via a pre-synaptic mechanism, although this suggestion was not directly verified (Carta et al. 2007). The 5-HT1A/1B-agonists significantly reduced extracellular DA levels at 40- to 80-min post-l-DOPA administration in both the striatum and the SN, and attenuated the AIM scores between 20 and 100-min post-drug injection (though not at later time points). Because 5-HT1A/1B-agonists reduce the activity of 5-HT neurons (reviewed in Blier et al. 1998), these results demonstrate a predominant serotonergic origin of l-DOPA-induced DA efflux, in line with the results of lesion studies (Tanaka et al. 1999). To verify whether DA efflux reflected impulse-dependent release, we infused TTX, a blocker of voltage-sensitive sodium channels, in both the striatum and the SN during the microdialysis experiment. Infusion of TTX significantly decreased extracellular DA levels by 70% and 60% in the striatum and the SN, respectively, at 40 to 80–100 min post-l-DOPA injection. The magnitude of this effect did not, however, differ significantly from that produced by the higher dose of the 5-HT1A/B agonists. The effect of TTX in our study is in agreement with previous reports from 6-OHDA-lesioned rats, where striatal TTX infusion was found to attenuate peak DA concentrations ‘on l-DOPA’ by 80% (Miller and Abercrombie 1999). A similar extent of DA reduction was reported following treatment with reserpine (Kannari et al. 2000). The fact that l-DOPA-induced DA efflux was not completely blocked by either 5-HT1A/1B treatment or TTX suggests that a non-neuronal source may account for part of striatal and nigral DA levels post-l-DOPA. Accordingly, it has been shown that astrocytes can serve as a temporary l-DOPA storage site and can convert l-DOPA to DA, suggesting expression of amino acid decarboxylase in this cell type (Tsai and Lee 1996).

Having verified that the largest portion of l-DOPA-induced DA efflux reflects active release from serotonergic neurons, we suggest that the higher peak of extracellular DA in dyskinetic rats compared to non-dyskinetic cases reflects properties of the serotonergic system in the former group. In line with this suggestion, extracellular and tissue concentrations of 5-HT and 5-HIAA were significantly larger in the striatum in dyskinetic rats compared to non-dyskinetic animals, which is indicative of a denser striatal serotonergic innervation in the former group. Conceivably, a denser 5-HT innervation results in a larger capacity for DA release following l-DOPA administration.

In contrast to the striatum, nigral extracellular and tissue concentrations of 5-HT and 5-HIAA did not differ between dyskinetic and non-dyskinetic rats. Yet, treatment with the 5-HT1A/B receptor agonists (at least with the higher dose used) significantly reduced DA efflux also in this structure. In line with previous reports (Bergquist et al. 2003; Sarre et al. 2004), this study showed that 6-OHDA injections in the MFB did not reduce extracellular DA concentrations in the ipsilateral SN (Fig. 1c), despite a pronounced loss of DA tissue content in the same structure (Table 1). The reasons for this phenomenon are not known, but it has been suggested that the maintenance of nigral extracellular DA levels following MFB lesion partly depends on DA diffusion from the adjacent ventral tegmental area (VTA) (Sarre et al. 2004), in which a substantial number of DA neurons are spared (Andersson et al. 1999) (cf. fig. 1a and c in Appendix S2). High-affinity DA reuptake is less active in the VTA compared to the SN (Cragg et al. 2001), resulting in a larger sphere of DA diffusion following release (Rice and Cragg 2008). Both the SN and the VTA receive a large 5-HT innervation (Herve et al. 1987; Moukhles et al. 1997). While interrupting ascending serotonergic pathways, a transection of the MFB has been shown to increase the density of 5-HT-immunoreactive fibres in the VTA and the medial part of the SN (Revuelta et al. 1999). A similar phenomenon occurred in the present study, where a dense serotonergic innervation was apparent in the 6-OHDA-lesioned VTA region, particularly in dyskinetic rats (fig. 1e in Appendix S2). Although nigral levels of 5-HT and 5-HIAA did not differ between dyskinetic and non-dyskinetic animals, the midbrain region sampled for microdialysis and biochemical assays in our study did not encompass the VTA. Based on the immunohistochemical observations and on the effects of 5-HT1A/1B agonist treatment, we thus suggest that a large part of l-DOPA-derived DA efflux in the SN may reflect release from serotonergic fibres located in the VTA (figs 1and 2 in Appendix S2).

In line with clinical observations (de la Fuente-Fernandez et al. 2004; Pavese et al. 2006), the present results establish that l-DOPA-induced DA efflux is larger in dyskinetic subjects compared to non-dyskinetic cases and indicate that the predominant contributor to this difference is the amount of DA released from serotonergic axons. At the same time our data show that the relationship between DA release and AIMs is not straightforward. This conclusion is based on two types of observations. First, in the experiment using 5-HT1A/1B receptor agonists, both of the tested doses reduced the AIM scores to the same extent, although the lower dose achieved a smaller reduction in DA levels in the striatum, and did not reduce DA efflux in the SN. Second, the quantitative and temporal relationship between AIMs and DA levels demonstrated that the responsiveness to extracellular DA was much increased in dyskinetic animals, which readily expressed AIMs when DA levels started to rise, and continued to exhibit severe AIMs even when DA levels had declined. The pathophysiological model emerging from these results is illustrated in Fig. 5. According to this model, two causative factors need to interact for a full expression of LID, namely, a large DA efflux post-dosing and an increased post-synaptic sensitivity in dopaminoceptive neurons (the latter has been demonstrated in a large number of previous studies reviewed in Cenci and Lundblad 2006; Cenci and Lindgren 2007; see also Berthet et al. 2009). Pharmacological agents that can blunt either of these phenomena (Carta et al. 2007; Mela et al. 2007; Munoz et al. 2008; Rylander et al. 2009) may have great utility in PD, where the treatment of LID still represents a major unmet need (Rascol et al. 2003; Olanow 2004; Cenci and Odin 2009).

Figure 5.

 Extent and sources of DA efflux following l-DOPA administration, and their significance to LID. While nigrostriatal projections provide the main site of DA formation and release in a normal brain (a), serotonergic neurons acquire a preponderant role in a severely DA-denervated brain (‘PD’, b and c). This pre-synaptic change is, however, not sufficient to cause LID (b). Dyskinetic movements are generated when both the sensitivity of post-synaptic neurons and the amount of DA released ‘on’l-DOPA exceed a critical threshold level. In this drawing, DA is represented by blue circles, and the number of circles reflects the peak increment in extracellular DA levels relative to baseline values (see ΔDA in Fig. 1b). Post-synaptic sensitivity is represented by a colour shift from yellow to orange in the dopaminoceptive striatal neurons, and by an increased number/sensitivity of glutamate (Glu-R) and DA receptors (DA-R) at the plasma membrane (documented in, e.g. Samadi et al. 2008; Berthet et al. 2009). These receptor classes regulate both the immediate electrophysiological response to DA and the activation of nuclear signaling pathways producing long-lasting adaptations (reviewed in Cenci and Lundblad 2006; Cenci and Lindgren 2007).


The skilful technical assistance of Gunilla Jonason is gratefully acknowledged. We also would like to thank Dr. Jonas Björk (Competence Centre for Clinical Research, Lund University Hospital) for his advice on statistics. The study was supported by grants from the following sources: the Swedish Research Council (HN and MAC), Åhlén’s Foundation (HN), Lundgren’s Foundation (HN), Arvid Carlsson Foundation (HN), Michael J Fox Foundation for Parkinson’s Research (MAC), The Swedish Parkinson Foundation (MAC), The Crafoord Foundation (MAC), and The Royal Physiographic Society (HSL).

Disclosure statement

The authors do not have financial or personal conflicts of interest associated with this work.