SEARCH

SEARCH BY CITATION

Keywords:

  • 5-HT1A receptor agonist;
  • 6-hydroxydopamine;
  • l-DOPA induced dyskinesia;
  • micro-positron emission tomography and microdialysis;
  • Parkinson’s disease

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

J. Neurochem (2012) 120, 806–817.

Abstract

Recent studies suggest that l-3,4 dihydroxyphenylalanine (l-DOPA)-induced dyskinesia (LID), a severe complication of conventional l-DOPA therapy of Parkinson’s disease, may be caused by dopamine (DA) release originating in serotonergic neurons. To evaluate the in vivo effect of a 5-HT1A agonist [(±)-8-hydroxy-2-(dipropylamino) tetralin hydrobromide, 8-OHDPAT] on the l-DOPA-induced increase in extracellular DA and decrease in [11C]raclopride binding in an animal model of advanced Parkinson’s disease and LID, we measured extracellular DA in response to l-DOPA or a combination of l-DOPA and the 5-HT1A agonist, 8-OHDPAT, with microdialysis, and determined [11C]raclopride binding to DA receptors, with micro-positron emission tomography, as the surrogate marker of DA release. Rats with unilateral 6-hydroxydopamine lesions had micro-positron emission tomography scans with [11C]raclopride at baseline and after two pharmacological challenges with l-DOPA + benserazide with or without 8-OHDPAT co-treatment. Identical challenge regimens were used with the subsequent microdialysis concomitant with ratings of LID severity. The baseline increase of [11C]raclopride-binding potential (BPND) in lesioned striatum was eliminated by the l-DOPA challenge, while the concurrent administration of 8-OHDPAT prevented this l-DOPA-induced displacement of [11C]raclopride significantly in lesioned ventral striatum and near significantly in the dorsal striatum. With microdialysis, the l-DOPA challenge raised the extracellular DA in parallel with the emergence of strong LID. Co-treatment with 8-OHDPAT significantly attenuated the release of extracellular DA and LID. The 8-OHDPAT co-treatment reversed the l-DOPA-induced decrease of [11C]raclopride binding and increase of extracellular DA and reduced the severity of LID. The reversal of the effect of l-DOPA on [11C]raclopride binding, extracellular DA and LID by 5-HT agonist administration is consistent with the notion that part of the DA increase associated with LID originates in serotonergic neurons.

Abbreviations used:
5-HT

serotonin

6-OHDA

6-hydroxydoapmine

8-OHDPAT

(±)-8-hydroxy-2-(dipropylamino) tetralin hydrobromide

ANOVA

analysis of variance

AP

anterior–posterior

AUC

area under curve

BPND

binding potential

DA

dopamine

DAT

DA transporter

DV

dorso-ventral

ERLiBiRD

Estimation of Reversible Ligand Binding and Receptor Density

i.p.

intraperitoneal

l-DOPA

l-3,4 dihydroxyphenylalanine

LID

l-DOPA-induced dyskinesia

ML

medio-lalteral

NA

noradrenaline

PD

Parkinson’s disease

PET

positron emission tomography

ROI

region of interest

s.c.

subcutaneous

VOI

volume of interest

Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by both motor and non-motor symptoms. Among the motor symptoms, tremor, rigidity, bradykinesia and postural instability are the most common that follow the loss of neurons of the dopaminergic nigrostriatal projection and dopamine (DA) in the striatum. The standard treatment of PD is replacement therapy with the precursor of DA, l-3,4 dihydroxyphenylalanine (l-DOPA), administered in combination with an inhibitor of peripheral[extracerebral] aromatic amino acid decarboxylase such as carbidopa or benserazide. Although direct DA agonists are often used in the early stages of the disease to reduce or delay the need for l-DOPA, ultimately l-DOPA must be added to any form of continuing treatment. The major drawback of l-DOPA therapy is the eventual development, within 5–8 years, of l-DOPA-induced motor fluctuations and dyskinesia (Ahlskog and Muenter 2001). Peak-dose dyskinesia is the most common and disabling symptom, consisting of involuntary, rapid and choreic-type movements (Jankovic 2005). Although abnormalities in DA turnover and DA transporter function likely play some role, repeated abnormal stimulation of both the DA D2 and D1 receptors, in response to rapid fluctuations in the extracellular (Abercrombie and Bonatz 1990) and synaptic (Pavese et al. 2006) and (de la Fuente-Fernández et al. 2004) concentrations of DA following a therapeutic dose of l-DOPA, are the conventional explanations of l-DOPA-induced dyskinesia (LID) with an emphasis on the role of processes downstream to the DA D1 receptor activation (Berthet and Bezard 2009) and (Cenci and Konradi 2010). Positron emission tomography (PET) with [11C] raclopride, a tracer of the DA D2/3 receptors, the binding of which responds to changes in synaptic DA concentrations, revealed a greater rapid increase of DA concentration 1 h after administration of a clinical dose of l-DOPA in dyskinetic patients compared with stable responders (de la Fuente-Fernández et al. 2004).

Although increased DA synthesis coupled with decreased storage capacity and fewer DA transporters in dopaminergic terminals may tentatively explain the non-physiological concentrations of DA measured in response to a therapeutic dose of l-DOPA in PD patients and animals with severe nigrostriatal lesions, a number of reports suggest that unregulated conversion of l-DOPA into DA and subsequent uncontrolled release can occur in non-DA cells, especially from serotonergic (Everett and Borcherding 1970) and (Ng et al. 1970) and noradrenergic neurons and from glial cells (Cenci and Lundblad 2006). Centrally, aromatic amino acid decarboxylase or DOPA decarboxylase is active in many cells, including dopaminergic neurons, but also in serotonergic and noradrenergic neurons (Cenci and Lundblad 2006). l-DOPA is an amino acid that undergoes facilitated diffusion across cell membranes, and the activity of the enzymes regulate the fractions of l-DOPA that are converted to DA in different cellular compartments of the brain (Gjedde et al. 1993) and (Cumming et al. 1995). When the DOPA decarboxylase activity is reduced in PD, larger than normal fractions of the amino acid are subject to conversion in cells other than the dopaminergic neurons of the nigrostriatal pathway.

The existence of the enzyme complement necessary for the synthesis of DA from l-DOPA in serotonergic terminals raises the possibility of a role of these neurons in the increased l-DOPA-induced DA release in advanced PD patients. Immunohistochemical studies also demonstrate that l-DOPA treatment can induce the formation and presence of DA in serotonergic neurons (Yamada et al. 2007). The mechanism of release of DA from serotonergic cells is not known but there is evidence that this release is inhibited by the pre-synaptic action of serotonin (5-HT) and stimulation of raphe 5HT1A autoreceptors (Kannari et al. 2001). LID is reduced by co-treatment with 5-HT1A receptor agonists, presumably through autoreceptor-induced decrease in joint 5-HT and DA release from serotonergic neurons (Eskow et al. 2009). Lesions to both the dopaminergic and serotonergic systems reduce fluctuations of extracellular DA in response to l-DOPA and decrease LID (Tanaka et al. 1999; Carta et al. 2007). These findings underline the essential role that serotonergic terminals may play in the striatal metabolism of l-DOPA and in the induction of LID in patients with advanced PD as well as in animals with severe dopaminergic nigrostriatal lesions. By means of diffusion-based volume transmission (Cragg and Rice 2004), DA released at serotonergic terminals after a dose of l-DOPA, in the absence of DA transporters, could, in theory, diffuse widely and rapidly throughout the extracellular and synaptic spaces in the striatum, and this could contribute significantly to increased synaptic DA concentration and receptor stimulation.

These mechanistic interactions between serotonergic and dopaminergic neurons in advanced PD cannot easily be investigated non-invasively in human subjects. Although in vivo imaging with raclopride as a surrogate marker of DA release can provide some clues to the changes of striatal DA concentrations after pharmacological challenge, actual measurements of DA concentration must confirm the changes of DA receptor occupancy determined with raclopride imaging. In the present study, we determined the extracellular concentrations of striatal DA in concert with behavioural assessment and in vivo PET measurement of synaptic DA occupancy in rats lesioned unilaterally with 6-hydroxydopamine (6-OHDA) as an experimental model of PD and LID in response to l-DOPA challenges with and without a 5-HT1A agonist.

By doing so, we tested two specific hypotheses that reveal the role of serotonergic innervation of the striatum in LID:

(i) Co-administration of (±)-8-hydroxy-2-(dipropylamino) tetralin hydrobromide (8-OHDPAT), a 5-HT1A agonist, with l-DOPA prevents the l-DOPA-induced displacement of [11C]raclopride and the instance of LID, without affecting the l-DOPA-induced motor effect in the unilaterally 6-OHDA lesioned rat.

(ii) The l-DOPA induced decrease of [11C]raclopride binding is more pronounced in the ventral striatum than in the dorsal striatum, in reflection of the dorsoventral gradient of serotonergic innervation of the striatum.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Experimental animals

Twenty-one female Sprague–Dawley rats (Taconic, Sjaelland, Denmark) weighing 250–270 g at the beginning of the experiment were housed at a 12-h light/dark cycle with free access to food and water during the entire period. All experiments were performed according to the Danish Experimental Animal Inspectorate.

Drugs

l-3,4-Dihydroxyphenylalanine methyl ester hydrochloride, benserazide hydrochloride, (±)-8-hydroxy-2-(dipropylamino) tetralin hydrobromide, desipramine hydrochloride and 6-hydroxydopamine hydrochloride were purchased from Sigma-Aldrich (Brøndby, Denmark). Dormicum was purchased from Midazolam (Herlev, Denmark); Hypnorm from VetaPharma (Leeds, UK) and Rimadyl from Aarhus University Hospital Pharmacy (Aarhus, Denmark).

Experimental parkinsonism

The rats were unilaterally lesioned with 6-OHDA, injected into the medial forebrain bundle as described in detail elsewhere (Lee et al. 2000). Briefly, Hypnorm (fentanyl, 0.315 mg/mL and fluanisone, 10 mg/mL) and Dormicum (midazolam, 5 mg/mL) were separately dissolved in 1 : 1 solutions in sterile saline. Anaesthesia was induced with Dormicum (4.5 mg/kg midazolam) and Hypnorm (9 mg/kg fluanisone and 0.2835 mg/kg fentanyl) administered in a volume of 1.8 mL/kg and was maintained with a third of the initial dose every 30–60 min. Desipramine hydrochloride (25 mg/kg i.p.) was given at least 60 min prior to anaesthesia to protect noradrenergic terminals. Animals were positioned in a stereotaxic frame (David Kopf, Tujunga, CA, USA) and received two infusions of 6-OHDA of 6 and 7.5 μg 6-OHDA in a concentration of 3 μg/μL dissolved in a solution of 0.05%l(+)-ascorbic acid in sterile saline at the rate of 1 μL/min. Injections were performed into the right medial forebrain bundle with a Hamilton syringe directed at the following coordinates: (i) tooth bar at +3.4, anterior–posterior (AP) −4.0, medial–lateral (ML) −0.8, dorsal–ventral (DV) −8.0 (6 μg 6-OHDA), and (ii) tooth bar at −2.3, AP −4.4, ML −1.2, DV −7.8 (7.5 μg 6-OHDA) in mm from bregma, midline and the dural surface, respectively. Before cautious removal, the Hamilton syringe was left at the site of injection for an additional 3 min to allow for diffusion of 6-OHDA. The rats received post-operative analgesia with Rimadyl (4 mg/kg) and additional saline was administered subcutaneously (s.c.) for hydration. The rats were carefully monitored until fully awake.

Dyskinesia and motor activity

After a few days of recovery from surgery, the animals started to receive daily injections of l-DOPA (8 mg/kg, s.c.) and benserazide (15 mg/kg, s.c.) in a volume of 0.3 mL saline. This dose has been found to induce stable expression of dyskinetic features in rats with severe 6-OHDA unilateral lesions (Cenci et al. 1998; Carta et al. 2007; Lindgren and Andersson 2010). Of the original group, 13 rats developed LID within a few days of initiation of l-DOPA + benserazide treatment and were included in the rest of the study. We rated the severity of LID according to three topographical manifestations, using the abnormal involuntary movement rating scale (Cenci et al. 1998) with small modifications. Briefly, we observed and scored LID for 2 min every 20th minute for a total period of 300 min after l-DOPA or l-DOPA + 8-OHDPAT administration. The topographical manifestations included: (i) limb: choreiform limb movements, stereotypic and repetitive movements on the side contralateral to the lesion; (ii) axial: deviation or torsion of the head, neck and trunk contralateral to the lesion; (iii) oral: repetitive chewing movements and tongue protrusion. l-DOPA-induced motor effects were considered to be drug-induced rotations in the form of full body turns contralateral to the lesioned striatum and were counted with the same schedule as LID. Each of the three LID manifestations and the drug-induced rotations were rated as: 0 = absent; 1 = occasional, that is, present during less than 50% of the observation time; 2 = frequent, that is, present during more than 50% of the observation time; 3 = present during the entire observation time but could be disrupted by strong external stimulation (sound for example) and 4 = continuous and not interrupted by strong sensory stimuli. The final motor score and LID score in this study, for the purpose of comparison to micro-PET and microdialysis data, were obtained during the performance of the microdialysis study. The final LID score used for analysis consists of the average of the three topographic manifestations.

Positron emission tomography

The rats underwent micro-PET after 1.5–3 weeks of daily l-DOPA treatment and the identification of reproducible LID as the state of rats with stable dyskinesia. Each rat underwent three micro-PET sessions with [11C]raclopride, at baseline before pharmacological challenge, 25–68 min after a dose of l-DOPA + benserazide (50 mg/kg + 25 mg/kg s.c.), and 26–63 min after injection of l-DOPA + benserazide immediately followed by injection of 8-OHDPAT (0.6 mg/kg i.p.). The high dose of 50 mg/kg of l-DOPA for the PET challenge was chosen to permit comparison with earlier work (Abercrombie and Bonatz 1990; Miller and Abercrombie 1999; Tanaka et al. 1999; Kannari et al. 2001). The dose of 0.6 mg/kg of 8-OHDPAT was selected as an intermediate to the doses used by Kannari et al. (2001), to induce a significant effect, measurable within the timeframe of our micro-PET design, while avoiding unwanted side effects of 5-HT agonism. The baseline and one challenge session with either l-DOPA or l-DOPA + 8-OHDPAT were performed on day one, and the session with the second challenge was performed within a week, that is, always on different days to reduce possible interactions between the two challenges. The animals did not receive its routine dose of l-DOPA at the regular time on the day of the scan as the drug injections were done as needed prior to injection of the radiotracer. Thus, for the baseline scan and the second and third scan of the series completed on different days, the animal received the last dose of l-DOPA approximately 26–28 h prior to the onset of the session. We used the same small animal tomograph (micro-PET R4; CTI Concorde, Knoxville, TN, USA) for all scans. Anaesthesia was induced in a chamber filled with 5% isoflurane in a mixture of oxygen (O2) (0.4 L/min) and air (1.5 L/min). After the induction of anaesthesia, we positioned the head of the animal in a custom-built Plexiglas head holder and maintained the anaesthesia with a cone mask delivering isoflurane (1.8–2.0%) in O2 (0.4 L/min) and air (1.5 L/min) fitted to the head holder. A catheter was inserted transcutaneously into the tail vein for injection of [11C]raclopride in saline. Rectal temperature was maintained close to 36.5°C with a heat lamp regulated by a rectal thermometer. The saturation and heart rates were monitored with the ABL520 system (Radiometer, Copenhagen, Denmark). At the beginning of each PET session, we obtained a 10-min attenuation scan with a 68Ge point source. Dynamic emission recordings were initiated upon injection of a dose of [11C]raclopride, between 7 and 8 MBq/100 g body weight, followed by a 60 (in three scans 90) min long emission recording of 23 frames increasing in duration from 15 s to 10 min. After the PET sessions, animals woke up spontaneously and were returned to the housing cage. Data processing and data analysis procedures are described in detail elsewhere (Pedersen et al. 2007). Briefly, attenuation-corrected dynamic emission images were reconstructed by 3-D-filtered back projection resulting in a 128 × 128 × 63 matrix. Summed emission recordings were manually registered to a digital high-resolution atlas of the rat brain (Toga et al. 1995; Rubins et al. 2003). Time activity curves were extracted by the resampling of five volumes of interest composed of left and right dorsal striatum (40 mm3), left and right ventral striatum (7 mm3) and cerebellum (314 mm3) templates in the native space. We calculated the binding potential BPND relative to ‘non-displaceable’ (i.e. non-saturable) radioligand in the tissue by means of the Logan reference tissue model (Logan et al. 1996) with cerebellum as the non-displaceable binding reference tissue in the period between 30 and 60 min. For comparison, we also used the non-linear regression method, Estimation of Reversible Ligand Binding and Receptor Density (ERLiBiRD) method using the time frames between 1 and 60 min (Møller et al. 2007). Only the results from the Logan reference tissue model are shown here, as the results with ERLiBiRD analysis completely supported the results with Logan reference tissue model. Three [11C] raclopride doses accidently were injected interstitially and hence the corresponding results were not used for further analysis.

Microdialysis

Following the completion of all micro-PET sessions, a subset of the animals (N = 7–8) were used to measure extracellular DA in parallel with the rating of the severity of LID. The animals were allowed a few days of recovery after the completion of the PET studies. They were then moved to the animal housing area of the microdialysis laboratory and were allowed a few days to acclimatize to the new environment. They did not receive l-DOPA during this period (10–14 days) after which daily l-DOPA was re-initiated. Microdialysis studies were performed about a week after the expression of stable abnormal involuntary movements was reinstated, that is, about 3.0–3.5 weeks after the last PET study. The microdialysis proceeded as described by Wegener et al. (2000). Under anaesthesia (Hypnorm 0.05 mL/100 g s.c. and Midazolam 0.5 mg/100 g i.p.), two microdialysis probes (concentric probes of rigid design with 3 mm active membrane, Goodfellow) were implanted stereotactically in both dorsal striata through a burr hole in the skull (AP + 0.5, ML ± 3.0, DV −6.0 from the dural surface and bregma, respectively, tooth bar −3.3). The probes were fastened by dental cement (GC Fuji PLUS, Alsip, IL, USA) and the rats were allowed to recover for 2 days before the microdialysis experiments were conducted. Approximately 24 h prior to the dialysis experiment, the rats were brought to the testing cage and connected to the wires in the microdialysis setup and to a swivel joint allowing unrestricted movement within the testing cage and the probes were perfused continuously with distilled water (Maxima) at a rate of 2 μL/min overnight. On the day of the experiment, the probes were perfused continuously with Ringer′s solution (147 mM NaCl, 4 mM KCl, 2.3 mM CaCl2) at a flow rate of 2 μL/min at least half an hour prior to collecting samples to make sure that the wires were filled with Ringers solution. During the experiment, dialysate was collected at 20-min intervals (40 μL/sample). After collection, 5 μL of perchloric acid (0.005 M) was added to 30 μL of each sample to reduce further metabolism and degradation. After centrifugation, the samples were stored at −80°C until analysis. On the day of analysis, we used high pressure liquid chromatography with electrochemical detection (ESA 5014B Coulometric and Coulochem III, ESA, Chelmsford, MA, USA) to analyze the samples. The mobile phase (NaH2PO4 13,40 g/L, Na2EDTA 15.0 mg/L, octane sulfonic acid 400 mg/L and acetonitrile 90 mL/L, pH 2.6 adjusted by phosphoric acid) was delivered at a flow rate of 400 μL/min. Using Chromeleon chromatographic software (Dionex, Chelmsford, MA, USA), the samples were analyzed for the content of DA. The detection limit was 0.1 nM. We acquired baseline data for 3 h, after which l-DOPA + benserazide (50 mg/kg + 25 mg/kg s.c.) was administered and sampling continued for a further 5 h. On the following day, after acquisition of new baseline data for 3 h to ensure that the previous dose of l-DOPA did not affect the results, the rats received l-DOPA + benserazide (50 mg/kg + 25 mg/kg s.c.) + 8-OHDPAT (0.6 mg/kg i.p.) and sampling continued for 5 h. The order of challenge drugs to be administered on day 1 or on day 2 were randomly chosen. Following the end of the microdialysis study the rats were rapidly decapitated with a guillotine for verification of lesion severity with autoradiography and probe location.

Autoradiography

The brains were removed, immediately frozen in isopentane and stored at −80°C until processing. The brains were cut into 20 μm coronal sections on the Vibratome ULTRApro 5000, mounted on poly-l-lysine slides (Menzel Gläzer Polysine, TM) and again stored at −80°C until further processing. We assessed the abundance of DA transporters by autoradiography of the DA transporter ligand [3H]WIN 35,428 to evaluate the ipsilateral loss of DA terminals by comparison with the contralateral side. Briefly, the slides were defrosted at 25°C for 30 min and pre-incubated for 5 min in buffer 1 (50 mM Tris–HCl, 120 mM NaCl and 5 mM KCl dissolved in distilled water at pH 7.9 and temperature 4°C) followed by assessment of the total DA transporter (DAT) binding by incubation of the slides with the DAT-tracer [3H]WIN 35,428 (15 nM) in buffer 2 (50 mM Tris–HCl, 300 mM NaCl and 5 mM KCl dissolved in distilled water pH 7.9 at 4°C) for 40 min. We assessed non-specific binding on adjacent slides by the addition of a NA-DA reuptake inhibitor (Nomifensine 10 mM) to buffer 2, of course with the DA transporter ligand [3H]WIN 35,428 (15 nM), in a separate tray. After the incubation, the slides were rinsed from excess incubation buffer by a post-wash twice for 1 min in buffer 2 and then dipped in distilled water at 4°C. The slides were dried under a stream of cold air, placed in a vacuum desiccator with paraformaldehyde for 24 h and exposed to tritium-sensitive imaging phosphor screens (Fujifilm) for 5 days. After exposure, the imaging phosphors were scanned with the BAS-5000 scanner. The autoradiograms were analyzed using Image Gauge 4.03 (FujiFilm). We measured digital light units by placing regions of interest on the sections based on three criteria: most intact tissue, darkest binding and outlines of the striatum. An average of non-specific binding, in units of digital light units, was measured from the slides that had been incubated with both [3H ]WIN 35,428 and nomifensine. The specific binding was calculated by subtracting the non-specific binding from the total binding.

Statistical analysis

Statistical analyses used Graphpad Prism version 5.0. Results are reported as means with the standard error of mean indicated by the symbol (±). Probability of less than 0.05 was considered statistically significant.

The severity of the lesion was calculated as the percent binding of [3H]WIN 35,428 in the lesioned compared with the intact striatum. Two-way analysis of variance (anova) followed by Bonferroni’s post-tests were used in the comparisons of [11C]raclopride binding at baseline and after l-DOPA or l-DOPA + 8-OHDPAT treatment, using lesion and challenge as factors. The changes in DA receptor occupancy after l-DOPA and l-DOPA + 8-OHDPAT challenges compared with baseline were estimated from the changes of [11C]raclopride BPND according to conventions (Gjedde and Wong 1987; Wong et al. 1997; Innis et al. 2007):

  • image

One-tailed, paired t-tests compared the changes of radioligand binding at baseline and during challenge conditions.

We used the average of orolingual, limb and axial dyskinesia scores as a measure of total dyskinesia, while the drug induced rotations reflected the motor effect of l-DOPA and hence a separate index of therapeutic efficacy. Using time and challenge with l-DOPA or l-DOPA + 8-OHDPAT as factors, two-way anova separately assessed the roles of total dyskinesia (as index of serotonergic mechanism) and drug-induced rotation (as index of dopaminergic mechanism).

We calculated the area under curve (AUC) to quantify the effects of drug challenge on extracellular DA concentrations following challenge with l-DOPA or l-DOPA + 8-OHDPAT. The microdialysis data only illustrate the effects of challenge drugs on extracellular DA as only a subset of animals that were used in the micro-PET study were also used for microdialysis.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Denervation

Compared with the intact striatum, DA transporter density evaluated by [3H]WIN 35,428 declined by 97% (95% CI 94.08–99.61) on the lesioned side, indicating a high degree of lesion and severe loss of nigrostriatal dopaminergic projections. Two-way anova showed significant increase (p < 0.0001) of [11C]raclopride BPND in dorsal and ventral striatum on the 6-OHDA injected side, determined with either method of analysis (Fig. 1), compared with the contralateral striatum.

image

Figure 1.  [11C] Raclopride BPND at baseline and after l-DOPA or l-DOPA + 8-OHDPAT challenge, calculated with Logan reference tissue model, in the dorsal (a) and ventral (b) striatum. The significance of the difference between baseline and challenge was tested with two-way anova’s, seperately for dorsal and ventral striatum (see text) (**p < 0.01; ***p < 0.001).

Download figure to PowerPoint

Dyskinesia and motor activity

Two-way anova showed significant effects of challenge drugs (p < 0.0001) and of time of administration of challenge drugs (p = 0.0004) and a significant interaction (p < 0.0001) between challenge drugs and of time of administration of challenge drugs on the severity of expression of LID. Bonferroni post-tests showed that dyskinesia severity significantly declined with co-administration of 8-OHDPAT until 180 min post-administration (Fig. 2a). There was a significant effect of time of administration of the challenge drugs (p < 0.0001) and of the challenge drugs themselves (p = 0.0072) but no significant interaction (p = 0.3763) between the time of administration and the kind of challenge drugs on the magnitude of drug-induced rotations. Bonferroni post-tests revealed significant difference in drug-induced rotations between the l-DOPA and l-DOPA + 8-OHDPAT drug challenges during the first 20 min post-administration (p < 0.001). In the following 280 min, the drug-induced rotations were similar for l-DOPA and l-DOPA + 8-OHDPAT treatment (Fig. 2b).

image

Figure 2.  Total dyskinesia scores were calculated from mean of orolingual, limb and axial dyskinesia scores. Co-administration of 8-OHDPAT significantly reduced total dyskinesia from 20 to 100 min (p < 0.001), 120 to 140 min (p < 0.01) and 160 to 180 min (p < 0.05) after drug challenge (a). Co-administration of 8-OHDPAT only reduced drug induced rotations in the first 20 min (p < 0.0001) after drug challenge. l-DOPA challenge (circles) and l-DOPA + 8-OHDPAT challenge (squares) (b).

Download figure to PowerPoint

Dopamine receptor availability

To take into account the possible effect of the injected mass of raclopride on [11C]raclopride BPND, the total injected dose of raclopride per body weight (kg) was calculated separately for the three PET-sessions and tested for correlation to [11C]raclopride BPND in all volumes of interest in the three PET sessions (the mass was known in 24 scans). The mean injected raclopride mass per bodyweight (kg) was 3.067 nmol/kg (± 0.4211) in the baseline session, 5.414 nmol/kg (± 0.9364) in the l-DOPA session and 4.203 nmol/kg (± 0.8234) in the l-DOPA + 8-OHDPAT challenge session. There was no significant correlation between the injected raclopride mass and [11C]raclopride BPND in the three PET sessions, except, after l-DOPA + 8-OHDPAT challenge, in the lesioned ventral striatum (p = 0.0313).

Two-way anova revealed significant effect of lesion on [11C]raclopride binding at baseline in both ventral and dorsal striatum (p < 0.0001) and significant effect of drug challenge in the ventral striatum (p < 0.0001) and close to significant effect in the dorsal striatum (p = 0.0996). The interaction between lesion and drug challenge was significant in both the dorsal and ventral striatum (p = 0.0444 and 0.0409, respectively). Bonferroni post-tests revealed a significant decrease of baseline [11C]raclopride binding after the l-DOPA challenge (p < 0.01 and p < 0.001 in the dorsal and ventral striatum, respectively). Co-administration of 8-OHDPAT reversed the l-DOPA-induced decrease of [11C]raclopride binding in the ventral striatum (p < 0.01) and showed a trend in the dorsal striatum, although the latter did not reach significance (p > 0.05) as shown in Fig. 1 of the [11C]raclopride-binding potentials. The ERLiBiRD non-linear method showed similar results (not shown here). However, as in Pedersen et al. (2007), [11C]raclopride BPND were somewhat lower with ERLiBiRD than with the Logan reference tissue model.

Dopamine receptor occupancy

We calculated the DA receptor occupancy from the change of baseline [11C]raclopride binding after the drug challenge, with one-tailed t-tests of the significance of random change of DA receptor occupancy after l-DOPA or l-DOPA + 8-OHDPAT drug challenge. The dorsal striatum of the intact hemisphere sustained 7% and 5% increases in DA receptor occupancy after l-DOPA and l-DOPA + 8-OHDPAT treatments, respectively (non-significant p = 0.2811). In the dorsal striatum of the lesioned hemisphere, the l-DOPA challenge increased the DA receptor occupancy by 18%, while the DA receptor occupancy marginally rose by 4% with the l-DOPA + 8-OHDPAT treatment (p = 0.0682). In the intact ventral striatum, l-DOPA treatment raised the DA receptor occupancy by 14%, and 8-OHDPAT co-treatment reversed the change to 0 (p = 0.213). In the lesioned ventral striatum, l-DOPA treatment raised DA receptor occupancy by 29% that significantly declined to 3% with 8-OHDPAT co-treatment (p = 0.0165), as shown in Figs 3 and 4.

image

Figure 3.  DA receptor occupancy after drug challenge relative to baseline in dorsal (a) and ventral (b) striatum was calculated according to conventions. One tailed t-tests were performed to test the significance of the difference in DA receptor occupancy between the two challenge conditions[(*p < 0.05; and a trend towards significance is indicated with (*)].

Download figure to PowerPoint

Micro-dialysis measurements of dopamine

By placing microdialysis probes bilaterally in both the right and left dorsal striatum, we were able to measure extracellular DA concentrations, before and after challenge, in both the intact and lesioned striata. Following administration of l-DOPA, the AUC of the DA concentrations on the lesioned side was much larger, than that of the AUC of the intact side, magnitudes averaging 5854 units and 1179 units, respectively for lesioned and intact hemispheres. The co-administration of 8-OHDPAT dramatically reduced DA concentrations in both lesioned and intact striatum, to 110.5 and 32.54 units, respectively, for the lesioned and intact sides, as shown in Fig. 5.

image

Figure 5.  Measurements of extracellular DA at baseline and after l-DOPA or l-DOPA + 8-OHDPAT challenge (time of administration of drug challenge is indicated by the dotted line). After l-DOPA administration the increase in extracellular DA was larger in the lesioned side compared with the intact side. 8-OHDPAT reduced the release of DA in both the intact and lesioned striata.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

In this animal model of advanced PD and LID, we combined in vivo micro-PET and microdialysis to assess the contributions of serotonergic neurons to the effects of exogenous l-DOPA in the lesioned brain. The unilaterally lesioned rats had three PET sessions with the DA receptor radioligand [11C]raclopride, once at baseline and then twice after both l-DOPA and l-DOPA + 8-OHDPAT challenges, accompanied by microdialysis of extracellular DA and rating of LID. As predicted on the basis of previous results, the co-administration of the 5-HT agonist, 8-OHDPAT, did significantly inhibit the increase of extracellular DA in the lesioned hemisphere and the displacement of the DA receptor radioligand from its receptors, and the effect was greater in ventral than in dorsal striatum, as predicted from the anatomy of serotonergic innervation of the striatum.

Baseline

The baseline of parkinsonism showed elevated [11C]raclopride binding in both dorsal and ventral striatum of the lesioned hemisphere, irrespective of analysis method. This finding is consistent with the abundant evidence from untreated symptomatic PD patients (Rinne et al. 1993), symptomatic MPTP-lesioned monkeys (Doudet et al. 2002) and 6-OHDA-lesioned rodents (Sossi et al. 2009). In these subjects, the increase of [11C]raclopride binding appears to reflect the combined changes of increased number of receptors (Bmax) and decreased DA receptor occupancy (Rinne et al. 1995) and (Doudet et al. 2002) and is consistent with the emergence of motor deficits from loss of nigrostriatal DA innervation where the loss exceeds 75% of baseline (Sossi et al. 2009). In this study, we had more than 90% loss of terminal density, confirmed by autoradiography of DA transporters.

In PD patients, chronic l-DOPA treatment reverses the increase of [11C]raclopride binding (Turjanski et al. 1997), such that no significant differences of [11C]raclopride binding are found in patients with advanced PD compared with age-matched control subjects. However, in this model of the disease in rats, the increase of [11C]raclopride binding persisted with the daily treatments with l-DOPA + benserazide at the time of the baseline PET session. The discrepancy may relate to the different frequency of treatment, as human subjects receive l-DOPA four times every day, while the animals were treated only once a day. The short half-life of l-DOPA and more rapid metabolism in rodents compared with primates may reduce the DA receptor stimulation and hence attenuate the compensatory down-regulation of DA receptors in response to l-DOPA in the animal model. Also, the PET session required that the rats received the last dose of l-DOPA at least 24 h prior to the tomography.

  • image(4)

[  We defined volumes of interest (VOIs) for the dorsal and ventral striatum in the digital rat atlas (a). Blue and green VOIs indicate the intact, left, dorsal and ventral striata, respectively. Yellow and white VOIs indicate the lesioned, right dorsal and ventral straita, respectively. (b) Parametric images of [11C] raclopride BPND at baseline (upper image) and after l-DOPA (middle image) or l-DOPA + 8-OHDPAT (lower image) were created using the Logan reference tissue model (arrow shows lesioned side). As seen in the upper and middle image and in support of our hypothesis, l-DOPA administration reduces the increase in baseline [11C] raclopride binding. As seen in the lower image, co-treatment with 8-OHDPAT reverses the effect of l-DOPA on [11C] raclopride binding. ]

l-DOPA challenge

Raclopride is a benzamide antagonist of the DA D2/3 receptors. As for other benzamides, it is believed that changes in radioligand binding in the challenge situation reflect changes of synaptic DA, that is, an increase in synaptic DA inhibits radioligand binding (Cumming et al. 2002; Ginovart 2005), although other mechanisms affecting receptor availability may also be involved (Doudet and Holden 2003; Ginovart 2005; Gjedde et al. 2005).

The l-DOPA challenge had no significant effect on the [11C]raclopride binding in the intact striatum, and the microdialysis revealed much smaller increases of the extracellular DA in this hemisphere than in the lesioned hemisphere after the l-DOPA challenge, as observed in previous studies with microdialysis (Abercrombie and Bonatz 1990) and PET (Sossi et al. 2009). The acute l-DOPA challenge significantly inhibited the [11C]raclopride binding in both dorsal and ventral striatum of the lesioned hemisphere, which is consistent with the prediction of increased extracellular DA after the challenge, and with the evidence from microdialysis. With micro PET, the increase of DA receptor occupancy in dorsal and ventral striatum of the lesioned hemisphere approximated 18% and 29%, respectively. Similar increases in DA occupancy have been reported in the striatum of patients with advanced PD and LID (de la Fuente-Fernández et al. 2004; Pavese et al. 2006) and in previous micro PET studies of 6-OHDA lesioned rats (Sossi et al. 2009). The change in radioligand binding after the l-DOPA challenge was more pronounced in patients with LID than in stable responders (Tedroff et al. 1996; de la Fuente-Fernández et al. 2004; Pavese et al. 2006).

With respect to the possible metabolism of l-DOPA in, and release of DA from, serotonergic neurons (Cenci and Lundblad 2006; Carta et al. 2007), we predicted that the l-DOPA-induced DA release, as measured with micro-PET, would be larger in the lesioned ventral striatum than in the lesioned dorsal striatum, as suggested by the increasing density of serotonergic innervation from dorsal to ventral striatum. The prediction was upheld by the observation that the l-DOPA-induced increase of DA receptor occupancy in the lesioned hemisphere numerically was larger in ventral striatum (29%) than in dorsal striatum (18%), although the difference was not statistically significant. The effect of partial voluming in this study also needs to be taken into consideration: It has been suggested that in humans 30% of the raclopride radioactivity in the ventral striatum originates from putamen and caudate (Mawlawi et al. 2001). Partial voluming may play an even greater role in the small rat brain and the poor resolution scanner used and thus, the greater serotonergic DA release in ventral than in dorsal striatum could be under-estimated.

In support of a gradient of serotonergic release of DA in advanced PD, recent microdialysis data demonstrated that the density of serotonergic innervation in- and outside striatum predicted the magnitude of l-DOPA-induced changes in extracellular DA concentration (Navailles et al. 2011). We could not test this observation in the present microdialysis study, as we only placed one probe in the middle of each dorsal striatum.

l-DOPA + 8-OHDPAT challenge

The possible involvement of the serotonergic system in PD and LID has been the focus of considerable attention, and has led to the suggestion of adjunctive 5-HT-selective drugs to the treatment of PD patients with LID (Cenci and Lundblad 2006; Carta et al. 2007; Fox and Chuang 2008). However, despite this special focus on the role of 5-HT1A receptors in the management of LID, few imaging studies have actually addressed the interactions between 5-HT and DA in advanced PD with LID (Esposito and Di Matteo 2008).

The 5-HT1A receptors are located post-synaptically in forebrain regions and pre-synaptically on the soma and dendrites of the serotonergic cell bodies in the raphé nuclei. The 8-OHDPAT agonist of 5-HT1A receptors is highly selective and potent (Stamford et al. 2000). By stimulation of pre-synaptic 5-HT1A autoreceptors, the agonist lowers 5-HT synthesis and neuronal firing as well as the subsequent release of 5-HT from terminals in the striatum of neurons that originate in the dorsal raphé nucleus (Barnes and Sharp 1999). Pharmacological inhibition of the activity of 5-HT neurons or lesioning of the 5-HT system in animal models of advanced PD both limit the l-DOPA-induced increase of extracellular DA in striatum and the severity of the accompanying LID (Tanaka et al. 1999; Kannari et al. 2001; Carta et al. 2007; Lindgren and Andersson 2010). The non-physiological release of DA from serotonergic neurons would lead to non-physiological stimulation of DA receptors. Hence, we tested the receptor occupancy of DA receptors in response to l-DOPA co-administered with 8-OHDPAT, and the co-administration attenuated the l-DOPA-induced changes both of extracellular DA and DA receptor occupancy in the lesioned striatum as measured by microdialysis and microPET with raclopride as the surrogate marker of DA occupancy. When 8-OHDPAT was co-administered, [11C] raclopride significantly returned to baseline in the ventral striatum but only showed a trend in the dorsal striatum, which is in contrast with our microdialysis results that show a dramatic decrease in extracellular DA in the dorsal striatum following 8-OHDPAT co-administration. It is however in line with previous studies (reviewed by Laruelle 2000) that show that the binding of benzamides, such as raclopride, more readily respond to increases in intrasynaptic DA compared with increases in extrasynaptic/extracellular DA. Breier et al. (1997) showed that an increase in extracellular DA by 1365%, after amphetamine challenge, decreased [11C] raclopride binding with 21%. However, Kim et al. (1998) showed that a challenge with nicotine that increases DA neuronal activity, without blocking DAT, increased extracellular DA by only 29% but reduced [11C] raclopride binding by 21%. Laruelle (2000) suggests that this apparent discrepancy between changes in extracellular DA on one hand and [11C] raclopride binding on the other hand support the hypothesis that the binding of benzamides reacts more readily to changes in intrasynaptic DA (Kim et al. 1998) compared with changes in extracellular DA (Breier et al. 1997). Although our dialysis data suggest a minimal effect of l-DOPA co-administered with 8-OHDPAT on the extracellular concentrations of DA, it is possible that this effect was intrasynaptically sufficient to induce some competition between raclopride and the endogenous DA released. In our dialysis study, an average 600% increase in l-DOPA induced extracellular DA led to about 18% change in raclopride binding in the dorsal striatum. The extracellular DA concentrations after l-DOPA + 8-OHDPAT showed only a marginal increase and the change in raclopride binding following l-DOPA + 8-OHDPAT in the dorsal striatum was in fact very close to the test–retest reproducibility value of the method in our hands (6%), suggesting that the apparent lack of significance may result from variability in a small sample of animals. It is however remarkable that the reversal of l-DOPA induced release of DA when co-administeed with 8-OHDPAT can be measured non-invasively in a small subject sample, especially when taking into consideration the technical limitations of small animal imaging and the effect of partial voluming in small brain regions. This suggests that similar studies may be performed successfully in a reasonable subset of human subjects, arguing favourably for the translational application of such pharmacological challenges.

In addition to the inhibition in serotonergic neurons of l-DOPA metabolism and DA release, stimulation of 5-HT1A receptors also may affect motor activity in animals with impaired dopaminergic neurotransmission. This effect may be mediated by stimulation of post-synaptic 5-HT1A receptors which modulate the glutamatergic cortico-striatal innervation that affects general activity (Reith et al. 1998; Mignon and Wolf 2005).

Dyskinesia and therapeutic motor effects

In this study, we scored three manifestations of dyskinesia (limb, axial and oral dyskinesia) and one of motor activity (rotation). We averaged the dyskinesia scores for all three manifestations as we observed no significant differences in the effect of the pharmacological challenges on the individual scores. In contrast, there were marked differences between the effects of the challenges on the dyskinesia on the one hand and the motor score on the other. As expected, the effects of l-DOPA challenge occurred rapidly and were observed within 20 min of administration, most likely because of an immediate increase of extracellular DA and were followed by emergence of LID and rotational behavior and decline of [11C]raclopride binding. In the interval that matched the duration of a micro-PET session, total LID scores fell significantly with the l-DOPA + 8-OHDPAT treatment, compared with the l-DOPA treatment alone. While co-administration of 8-OHDPAT significantly reduced the total dyskinesia throughout the entire 5-h post-challenge period, l-DOPA + 8-OHDPAT administration only delayed the onset of drug-induced rotations by up to 40 min.

Co-treatment with 8-OHDPAT not only reduced LID as shown by others (Carta et al. 2007) but it also altered the temporal development of the LID. While maximum LID appeared within 20 min of the injection and remained maximal for the next 4 h, slowly decreasing by the end of the experiment, observable dyskinesia after l-DOPA + 8-OHDPAT did not emerge until an hour post-injection and progressed slowly towards a peak during the last hour of the study, never reaching the intensity observed after l-DOPA administration. This pattern suggests that co-administration of 8-OHDPAT may lower the firing of serotonergic neurons sufficiently to delay the DA release. As measured with micro-PET, co-treatment with 8-OHDPAT significantly prevented or reduced the l-DOPA induced increase of DA occupancy that lowered the binding of [11C] raclopride in the ventral striatum.

It is generally accepted that the beneficial and detrimental effects of l-DOPA are present with the same frequency in patients with advanced PD that receive long-term l-DOPA treatment. This relationship between positive and negative effects of l-DOPA complicates the treatment of PD since it is very difficult to limit the LID without losing the beneficial effects of l-DOPA. The present study suggests that non-physiological activation of D2 receptors by DA, inappropriately released from serotonergic neurons, may play an important role in the development of LID.

The present study also suggests that adjunct therapy with a 5HT1A receptor agonist, optimized for dose and time of administration, may help maintain the beneficial motor effects of l-DOPA therapy, at least in some patients as it reduces the occurrence of the unwanted dyskinesia.

In conclusion, this study demonstrates that evaluating the effect(s) of pharmacological challenges by one or multiple drugs may be a feasible goal in human subjects. Although studies similar to ours, demonstate the effects of co-administration of l-DOPA with 5HT1A agonist on DA synthesis and release have been published, most have used a combination of behavioral assessment with an invasive method such as post-mortem examination or microdialysis, making them ethically not acceptable for a human population. This study add a significant translational component to earlier studies as we demonstrate that PET, using a specific tracer of DA release (in our case raclopride but it could be feasible with tracers of extrastriatal DA function such as FLB457 or fallypride) may help investigate the potential corrective effect of adjunct therapies as well as their mechanism of action in subjects suffering from Parkinson’s Disease.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This study received financial support from Aarhus PET-Centre, Aarhus University Hospitals, Aarhus University, Denmark and from a grant from the Danish Medical Research Council. The authors have no conflicting interests. We are most grateful to Dr Elissa Strome for teaching us how to carry out the surgical procedures.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References