Chronic intermittent L-DOPA treatment induces changes in dopamine release


Address correspondence and reprint requests to Dr. Ingrid Strömberg, Department of Integrative Medical Biology, Umeå University, S-901 87 Umeå, Sweden. E-mail:


3,4-Dihydroxyphenyl-l-alanine (l-DOPA)-induced dyskinesia often develops as a side effect of chronic l-DOPA therapy. This study was undertaken to investigate dopamine (DA) release upon l-DOPA treatment. Chronoamperometric measurements were performed in unilaterally DA-depleted rats, chronically treated with l-DOPA, resulting in dyskinetic and non-dyskinetic animals. Normal and lesioned l-DOPA naïve animals were used as controls. Potassium-evoked DA releases were significantly reduced in intact sides of animals undertaken chronic l-DOPA treatment, independent on dyskinetic behavior. Acute l-DOPA further attenuated the amplitude of the DA release in the control sides. In DA-depleted striata, no difference was found in potassium-evoked DA releases, and acute l-DOPA did not affect the amplitude. While immunoreactivity to serotonin uptake transporter was higher in lesioned striata of animals displaying dyskinetic behavior, no correlation could be documented between serotonin transporter-positive nerve fiber density and the amplitude of released DA. In conclusions, the amplitude of potassium-evoked DA release is attenuated in intact striatum after chronic intermittent l-DOPA treatment. No change in amplitude was found in DA-denervated sides of either dyskinetic or non-dyskinetic animals, while release kinetics were changed. This indicates the importance of studying DA release dynamics for the understanding of both beneficial and adverse effects of l-DOPA replacement therapy.

Abbreviations used

5-hydroxytryptamine or serotonin




abnormal involuntary movement






phosphate-buffered saline


Parkinson’s disease


serotonin transporter


tyrosine hydroxylase

The dopamine (DA) precursor 3,4-dihydroxyphenyl-l-alanine (l-DOPA) has routinely been used for pharmacological treatment of the symptoms in Parkinson’s disease (PD) since the late 1960s (Cotzias et al. 1967). After long-term l-DOPA treatment, patients develop motor complications, which involve motor fluctuations and abnormal involuntary movements (AIMs) or l-DOPA-induced dyskinesia (Yahr 1970). Approximately 40% of the parkinsonian patients become dyskinetic within 4 years of l-DOPA treatment, and the number of affected patients increases over time (Ahlskog and Muenter 2001).

In animal models of l-DOPA-induced dyskinesia, it has been demonstrated that a subpopulation of animals with unilateral DA denervation develops AIMs after low-dose, chronic l-DOPA treatment (Cenci et al. 1998; Lundblad et al. 2002). Although l-DOPA is known to be converted to DA and to increase the extracellular level of DA in rodent models of PD (Abercrombie et al. 1990; Di Monte et al. 1996; Finberg et al. 1998; Miller and Abercrombie 1999; Rodríguez et al. 2007), the DA physiology has not yet been extensively studied such that it can explain why some of the l-DOPA-treated animals never develop the dyskinetic behavior while others do. There are several locations where systemically given l-DOPA can be converted to DA in the DA-depleted brain (Cenci and Lundblad 2006). For instance, l-DOPA conversion can occur in non-neuronal cells (Melamed et al. 1980; Juorio et al. 1993). Furthermore, 5-hydroxytryptamine or serotonin (5-HT) neurons contribute to extracellular DA after systemic l-DOPA injection in the rat (Arai et al. 1995; Tanaka et al. 1999), and indeed, dyskinesia is improved by compounds acting on serotonergic receptors (Kannari et al. 2001; Durif et al. 2004; Bara-Jimenez et al. 2005; Taylor et al. 2006; Carta et al. 2007; Dekuny et al. 2007; Zhang et al. 2008). Moreover, serotonergic hyperinnervation, as revealed by grafting of serotonergic neurons into DA-depleted striatum, exacerbates dyskinesia, while lesioning of both the DA and 5-HT nerve fibers in the striatum, dampen dyskinetic behavior (Carlsson et al. 2007; Carta et al. 2007). Thus, there are strong indications that dyskinetic behavior is related to l-DOPA being converted to DA and released from the serotonergic rather than from the dopaminergic nerve fibers. The actual release of converted l-DOPA in DA-depleted striatum has, however, not yet been explored in dyskinetic versus non-dyskinetic animals.

The release mechanism of DA, synthesized from exogenous l-DOPA, is also likely to play an important role to achieve an appropriate function in DA signaling. Microdialysis studies have been undertaken to elucidate whether DA can be released from serotonergic nerve terminals in the striatum, but the results are contradictory. Miller and Abercrombie (1999) found the release to be sensitive to tetrodotoxin, which indicates that the release occurs from neural cells. Other microdialysis studies have shown the increased extracellular concentration of DA to be insensitive to tetrodotoxin, suggesting that the extracellular DA levels originate from non-neuronal sources (Sarre et al. 1994). Microdialysis, while very sensitive to low concentrations of DA, has low time resolution (1 sample/20 min), which excludes the possibility of studying the kinetics of DA release. In vivo chronoamperometry is another method to monitor KCl-evoked DA releases. This technique offers subsecond time resolution, and allows, in addition to peak concentration of KCl-evoked release, to also monitor changes in uptake rate and clearance time (van Horne et al. 1992; Zahniser et al. 1998). Therefore, in this study, in vivo chronoamperometry was used to monitor DA release in normal, 6-hydroxydopamine (6-OHDA) lesioned, and DA-depleted chronically l-DOPA treated animals with or without l-DOPA-induced AIMs. Measurements of potassium-evoked releases were made before and after systemic injection of l-DOPA. After electrochemical detections, the brains were processed for immunohistological evaluations of nerve fiber densities in the striatum.

Materials and methods


Female Sprague–Dawley rats (B&K, Sollentuna, Sweden), weighing 200–225 g in the beginning of the experiments were used. The animals were housed under 12 h light/dark conditions with access to food and water ad libitum. The treatment of the animals and their conditions were in accordance with internationally accepted guidelines, and had been approved by local ethics committee.


3,4-Dihydroxyphenyl-l-alanine and the DOPA decarboxylase inhibitor benserazide-hydrochloride (Sigma-Aldrich, Stockholm, Sweden) were dissolved together in physiological saline, and administered s.c. within 1 h after preparation in a volume of 2.0 mL/kg body weight. The dose/injection used throughout the experiment was 4 and 15 mg/kg for l-DOPA and benserazide, respectively. This dose of pure l-DOPA is equal to 5 mg/kg l-DOPA methyl-ester.

Dopamine lesions and amphetamine-induced rotations

Animals were subjected to unilateral 6-OHDA lesions according to a standard procedure (e.g. see Cenci et al. 1998). Rats were anesthetized by initial exposure to isoflurane® in an induction chamber and subsequently placed in a stereotaxic frame and maintained anesthetized in 1.5% isoflurane/oxygen for inhalation. 6-OHDA-HCl (3 μg/μL; Sigma-Aldrich) was dissolved in saline containing 0.02% ascorbic acid and injected into the right medial forebrain bundle at the following coordinates (in mm) relative to bregma and the dural surface: (i) AP: −4.4, ML: 1.2, V: −7.8, tooth bar set at −2.3 (7.5 μg 6-OHDA) and (ii) AP: −4.0, ML: 0.8, V: −8.0, tooth bar set at +3.4 (6 μg 6-OHDA). At 2 weeks post-lesion, rats were tested for amphetamine-induced rotational behavior (2.5 mg/kg d-amphetamine i.p. over a time period of 90 min), and only animals showing individual mean > 5.0 full turns per min in the direction ipsilateral to the lesion were selected for the study. This rotational score has been shown to correspond to > 90% depletion of DA fiber terminals in the striatum (Lundblad et al. 2002).

Ratings of abnormal involuntary movements

After screening the rats for high degree of DA depletion, selected animals were either chronically treated, once per day, with injections of l-DOPA 4 mg/kg combined with 15 mg/kg benserazide for 14 days or with saline, to serve as controls. AIMs were rated on days 1, 4, 7, 10, and 14 according to the validated rating scale used in previous studies (Cenci et al. 1998; Lee et al. 2000; Lundblad et al. 2002), with the addition of a recently developed amplitude rating scale (Carta et al. 2006). The AIMs are divided in three categories namely (i) axial AIMs, i.e. twisting movements of the neck and upper body contralateral to the lesion; (ii) forelimb AIMs, i.e. repetitive jerks or dystonic posturing of the contralateral forelimb, and/or purposeless grabbing movement of the contralateral paw; (iii) orolingual AIMs, i.e. empty jaw movements and contralateral tongue protrusion. Each AIM subtype was rated on a severity scale from 0 to 4 based on its duration and persistence (1 = occasional; 2 = frequent; 3 = continuous but interrupted by sensory distraction; and 4 = continuous, severe, not interrupted by sensory distraction) over 1 min time period. Ratings were made every 20th minute starting 20 min after l-DOPA injection and the last observation was made 180 min after l-DOPA. The amplitude of axial AIMs was rated according to the lateral deviation (or torsion) of the animal’s neck and upper trunk from the longitudinal axis of the body. The amplitude of limb AIMs was scored based on the extent of limb translocation and on the visible involvement of proximal muscle groups. Orolingual AIMs amplitude was based on the extent of involvement of facial, masticatory, and tongue muscles. The standard AIM score was always given before the amplitude score to ensure that each rated movement was verified to be dyskinesia, and not normal movements. A global AIM score was calculated by multiplying the standard and amplitude scores for each AIM subtype at every time point, and then adding these products for the entire testing session (Carta et al. 2006; Mela et al. 2007). After the AIMs ratings during chronic l-DOPA treatment, the animals were divided into two groups according to their individual AIM score: animals were ranked as dyskinetic when the global AIM score/session was > 50 and non-dyskinetic when the global AIM score/session was < 5 (Westin et al. 2006). Animals with AIM score between 5 and 50 were excluded from further studies. In addition, locomotive hyperactivity was induced by l-DOPA injection in both non-dyskinetic and dyskinetic animals with no difference monitored, as previously described (Winkler et al. 2002; Picconi et al. 2003). However, this behavior has been shown to correspond to drug-induced rotation and does not provide a specific measure of dyskinesia in the rat (Lundblad et al. 2002).

After the 14-day chronic treatment period, l-DOPA treated animals received two to three l-DOPA injections/week to maintain the dyskinetic behavior at a plateau level (Lundblad et al. 2002). At the last day of the 10-day plateau period, the AIMs were tested to verify that the dyskinetic status of the animals remained stable. The plateau regimen of l-DOPA continued during the time when measurements were performed. Animals received l-DOPA on different days so that chronoamperometric measurements were made 2 days after the last l-DOPA injection in all animals. Chronoamperometric measurements were performed during a 19-day period starting 2 days after the 10-day plateau l-DOPA administration regimen (Fig. 1).

Figure 1.

 Flow scheme over animal treatment and study design.

Electrochemical detection

High-speed chronoamperometric measurements (5 Hz) of extracellular DA levels were performed using a Pentium-IV microcomputer-controlled instrument (FAST-12; Quanteon L.L.C., Nicholasville, KY, USA) as previously described (Hoffman and Gerhardt 1998). Briefly, an oxidation potential was applied (+0.55 V; resting 0.0 V vs. Ag/AgCl reference) and the resulting oxidation and subsequent reduction currents from the microelectrodes were integrated during the final 80% of each 100-ms pulse. Both oxidation and reduction currents were continually recorded and integrated over 1 s.

Single carbon-fiber electrodes (Quanteon L.L.C) sealed in a glass capillary (fiber diameter 30 μm; exposed length 120–150 μm) and coated with nafion were dried at high temperature (two to three coats at 200°C for 3 min; Sigma-Aldrich) before use. The electrodes were dried for 3 min at 200°C before coating was applied. Nafion coating prevents detection of anionic substances and increases the selectivity for DA. The electrode sensitivity and linearity were determined by generating calibration curves in 0.1 M phosphate-buffered saline (PBS) solutions, pH 7.4, at 20°C for each recording electrode. Electrode responses were regarded as linear for 2–10 μM increments of DA when r2 > 0.997. The electrodes showed high sensitivity to DA but were insensitive to ascorbic acid, with an average selectively ratio of DA to ascorbic acid of 3767 ± 482 to 1 (n = 24). Limit of detection, defined as signal to noise ratio of 3 : 1, was 0.013 ± 0.0030 μM.

Pilot studies revealed that the electrochemical properties were different between the commonly used methyl-ester form of l-DOPA and the l-DOPA used in this study, as the methyl-ester form of l-DOPA was detected by the carbon fiber electrode prepared for detection of DA with a ratio of about 1 : 1 (data not shown) and a red/ox ratio of about 0.8, which cannot be separated from DA red/ox ratio of about 0.7. To avoid interference from the methyl-ester form of l-DOPA, the pure form was used during the entire study. The pure form of l-DOPA is not electrochemically active and could not be detected with the electrodes used in this study.

The electrodes were mounted together with a single glass micropipette with outer tip diameters of 10–15 μm using sticky wax. The distance between the electrode and micropipette was 130–160 μm. The micropipette was filled with KCl (120 mM, pH 7.4) and connected to a micropressure system (BH-2; Medical System, Washington, DC, USA). The volume applied was determined using a stereomicroscope fitted with a reticule in one eyepiece to measure the movement of the meniscus in the micropipette (Friedemann and Gerhardt 1992). An Ag wire was used as reference electrode, which was prepared by plating for 30 min in 1 M HCl solution saturated with NaCl.

Experimental animals

Based on the results from the AIMs ratings, made during the chronic treatment period, the animals were assigned as dyskinetic (n = 6, global AIMs > 50) or non-dyskinetic (n = 4, global AIMs = 0–5). Animals that developed mild or moderate dyskinesia (global AIMs 5–50) were excluded from the study. Rats with 6-OHDA lesions (n = 4) and normal animals (n = 5), without any l-DOPA treatment, served as controls. In addition, another set of five normal animals were used in a separate control experiment to monitor the effects of long-term anesthesia and/or electrode sensitivity change over time.

Electrode implantation and recording procedures

The animals were anesthetized with urethane (Sigma-Aldrich) at the dose of 1.25–1.5 g/kg body weight and placed in a stereotaxic frame. The rats were tracheotomized and allowed to breath spontaneously. Body temperature was maintained at 37°C with a thermostatic heating pad coupled to a rectal thermometer. The skull and dura overlying the striatum were bilaterally removed. The Ag/AgCl reference electrode was implanted approximately 3 mm into the left hemisphere 4–5 mm caudally to bregma. The electrode/micropipette assembly was lowered into the dorsal striatum using the following coordinates, calculated from bregma: AP: + 1.0 and 0.0 mm, ML: ±3.4 mm, V: −3.0 to −5.0 mm (five sites of measurement with 0.5 mm separation per track). Two consecutive KCl ejections were made at each recording site with an interval of 2 min. The first KCl ejection causes release from cells in a resting state. After 2 min, the signal has returned to baseline but the cells have not fully recovered and thus, the second KCl ejection represents release from a challenge state. The electrode/micropipette assembly was implanted and left in place at ventral site −3.0 mm for 60 min before measurements were initiated. During this time period, the anesthesia level and the baseline is stabilized (data not shown). Once the electrode was in position, a calibrated volume of KCl was applied by pressure ejection (240 nL, 1–20 psi for 0.1–4 s) at 2-min intervals. Recordings were made from both the left and the right striata at the coordinates listed above. The first measurements were alternated between the hemispheres and balanced within each group with respect to which side was first measured. After measuring both hemispheres, the rat was injected with 4 mg/kg l-DOPA combined with 15 mg/kg benserazide (same dose as in the chronic treatment period). Measurements, performed after acute l-DOPA treatment, were initiated 40 min after the injection at different anterior/posterior coordinates from what was used before giving l-DOPA. Thus, no measurements were performed twice at the same recording site. The entire measuring procedure lasted for 6–7 h starting from the urethane injection.


Immediately after the electrochemical measurements were terminated, the animals were perfused with 4%p-formaldehyde. After rinsing in 10% sucrose diluted in 0.1 M phosphate buffer, coronal brain sections were collected from the striatum and mounted on slides. The sections were processed for indirect immunohistochemistry using primary antibodies against tyrosine hydroxylase (TH; diluted 1 : 1500; Diasorin Inc., Stillwater, OK, USA) or serotonin transporter (SERT; diluted 1 : 400; Chemicon Int. Inc., Cambridge, UK). Incubations in primary antibodies were performed in 48 h at 4°C. After rinsing in PBS sections were incubated in Alexa 594 secondary antibodies (Molecular Probes, VWR Int., Stockholm, Sweden) for 1 h at 20°C. All antibodies were diluted in PBS containing 0.3% Triton X-100. The sections were rinsed and mounted in 90% glycerin in PBS. Densities of TH- and SERT-positive nerve fibers were measured as optical density in the dorsal striatum and expressed as gray values. Measurements were performed on blind-coded slides, and based on images captured with a CCD camera (ProgRes C14; Jenaoptik, Jena, Germany). NIH image software (National Institute of Health, was used to produce the gray density from binary images. Mean values for each brain and side were calculated from three images.

Statistical analysis

Data from the AIMs recording are presented as global AIM score ± SEM. Repeated measures anova with Tukey post hoc test was used to assess the development of AIMs over time.

In the in vivo electrochemical measurements, two different anterior–posterior coordinates in the lateral striatum were chosen to enable measurements of both pre- and post-administration of l-DOPA and to avoid the use of the same coordinate twice. In normal striatum, no statistically significant difference in amplitude was found between the rostral (1 mm anterior to bregma) and caudal (0 mm relative to bregma) positions (position effect p = 0.143, two-factor anova, measurement sequence order and position as factors), no difference was found between the measurement sequence order (measurement sequence order p = 0.47, two-factor anova), and no interaction was found between measurement sequence order and position (p = 0.956, measurement sequence order and position interaction, two-factor anova). Similar analysis was performed for the different dorsoventral recording sites and no significant differences could be detected. To control for a potential time-dependent change in sensitivity of the electrode or changes in animal physiology because of long-term anesthesia, statistical analysis revealed no significant changes in amplitude of released DA or any of the temporal parameters (T50 and Trise), as a result of duration of recording time (one-factor anova, F = p = 0.20, p = 0.68, p = 0.58, respectively).

Based on the results above, average values for the different parameters were calculated for measurements made on the intact and the lesioned sides, pre- and post-injection of l-DOPA, irrespective to location of the detecting electrode. Thus, each average value represents the mean of measurements made on five depths in the lateral striatum at either AP ±0 or +1 mm, relative to bregma. Separate average values on all parameters were calculated for the first and second KCl-evoked releases. All statistical analysis was made using one- or two-factor anova and Tukey post hoc test with 95% confidence interval. Data are presented as mean ± SEM. Four different parameters describing the kinetics and the maximal concentration of DA after KCl-evoked release are presented. Amplitude, the highest DA concentration (in μM) recorded after the KCl-evoked release, Trise, the time (in seconds) between the KCl ejection and the maximal concentration is reached, T50, the time from the maximal concentration to 50% of the maximal concentration is reached, and uptake rate, the peak amplitude concentration (μM) multiplied with the Michaelis–Menten first-order decay constant (1/s) were determined.

On sections processed for TH- and SERT-immunohistochemistry, statistical analysis was performed on mean gray values from each animal and hemisphere using one-factor anova and Tukey post hoc test, and expressed as mean ± SEM.


Abnormal involuntary movements

l-DOPA-induced dyskinesia was monitored over the time period for chronic treatment using the rat AIM rating scale. During the first week of l-DOPA treatment a majority of the animals developed increasingly severe AIMs (Fig. 2). At the end of the chronic treatment period, animals were divided into dyskinetic or non-dyskinetic behavior based on the results of the AIMs ratings. The dyskinetic animals had at this time point reached a plateau level of dyskinesia. The AIMs scoring revealed that 78% of the l-DOPA-treated animals developed severe dyskinesia with a global AIM score/session > 50 and were therefore included in the study as dyskinetic rats. Animals that did not develop dyskinesia during the chronic l-DOPA treatment period were assigned to the non-dyskinetic animals (Fig. 2). As previously documented, the dyskinetic and non-dyskinetic animals maintained their global AIMs scores after the daily l-DOPA treatment period (Fig. 2) by 2–3 l-DOPA administrations per week (Lundblad et al. 2002).

Figure 2.

 Global AIM score was monitored in the unilaterally dopamine-depleted animals during l-DOPA treatment. The animals were separated by their response to dyskinetic behavior. Animals that did not express any dyskinetic behavior were regarded as non-dyskinetic, while the animals that improved their dyskinetic parameters over time and reached a plateau after 2 weeks were regarded as dyskinetic. All animals between the two groups were excluded from the study.

Electrochemical detections of released dopamine

Overall changes in dopamine release in 6-OHDA-lesioned striatum

The DA depletion produced an expected and dramatic effect on the amplitude of KCl-evoked DA release when compared with the intact striatum (Fig. 3). The amplitude for released DA was very low in the lesioned striata and measured about 5% of values achieved in control striata (Fig. 3). Furthermore, the temporal pattern of the DA release was also changed in DA-depleted striata, i.e. Trise, the time needed to reach the peak amplitude, and T50, the time measured from maximal amplitude until the concentration had reached 50% of maximal concentration, were significantly increased in the denervated compared with normal striatum (one-factor anova, side effect p = 0.002, F = 12.46 and p = 0.015, F = 6.96, respectively; Fig. 4a and b). Trise was an average of 51% longer in the lesioned striata (from approximately 33 s in normal animals to approximately 50 s in the lesioned striata), and the reuptake was affected such that T50 was increased by approximately 40% (from 57 to 80 s). In the lesioned striatum, the uptake rate (μM DA/s), calculated as the slope constant multiplied with the peak amplitude, was reduced by an average of 95.7% compared with the intact side (two-factor anova, side effect p = 0.022, F = 5.82, animal group effect p = 0.049, F = 2.95, animal group and side interaction p = 0.11).

Figure 3.

 Traces from potassium-evoked (arrows) dopamine releases in normal (a) and dopamine-depleted (b) striata (note the different scale of the y-axis). The peak amplitude was significantly higher in normal compared with dopamine-denervated striatum, while the reuptake was much slower in the denervated side.

Figure 4.

 Electrochemical measurements of potassium-evoked dopamine release in the intact and dopamine-depleted (lesioned) sides of normal, dopamine-depleted l-DOPA naïve (lesion cont.), dopamine-depleted dyskinetic (dyskinetic), and dopamine-depleted non-dyskinetic (non-dyskinetic) animals. Trise (a and c) was measured as the time needed to reach the peak amplitude, and T50 (b and d) was measured as the time from peak amplitude until the concentration had decreased to 50% of maximal concentration after the first (a and b) and second (c and d) potassium-evoked release. The time between the two potassium ejections were 2 min. An overall increase in time was monitored in the lesioned sides compared with intact sides and normal striatum. After the second potassium ejection, Trise was increased in both intact and lesioned sides of non-dyskinetic animals after acute l-DOPA, while it was reduced in the lesioned sides of the dyskinetic animals. In control sides of non-dyskinetic animals, T50 was affected by increased time after acute l-DOPA treatment; *p < 0.05.

Effects of chronic l-DOPA treatment on dopamine release in intact striatum

The amplitude of the first potassium-evoked DA release was significantly reduced on intact sides of both dyskinetic and non-dyskinetic animals, i.e. chronically l-DOPA-treated, when compared with intact side of DA-depleted drug naïve control animals before given acute l-DOPA (two-factor anova where animal group and treatment were used as factors, animal group effect p = 0.003, treatment effect p = 0.048; Fig. 5a). The same trend was also obvious after the second KCl ejection in the control sides, but these changes did not reach statistical significance (p = 0.096 for animal group effect). No changes correlated to chronic l-DOPA treatment were found on release kinetics, i.e. Trise, T50 or uptake rate on the intact sides.

Figure 5.

In vivo electrochemical detection of potassium-evoked dopamine releases from normal (control), dopamine-depleted l-DOPA naïve (lesion/control), dopamine-depleted dyskinetic (dyskinetic), and dopamine-depleted l-DOPA treated non-dyskinetic (non-dyskinetic) animals contralateral (a) and ipsilateral to the lesion (b). Chronic l-DOPA treatment reduced the amplitude of released dopamine in the control sides compared with l-DOPA naïve striatum. The amplitude of released dopamine was neither affected by chronic nor by acute l-DOPA ipsilateral to the lesioned sides; *p < 0.05.

Effects of chronic l-DOPA treatment on dopamine release in the DA-depleted striatum

In the lesioned striatum, no statistically significant differences in the amplitude of released DA were found, as measured after the first or second potassium ejection (Fig. 5b). No changes concerning release kinetics were found after chronic l-DOPA treatment compared with l-DOPA naïve animals in the DA-depleted side (Fig. 4).

Acute l-DOPA effects

After acute l-DOPA administration, the peak amplitude of DA upon the first KCl ejection was significantly reduced by an average of 40% in the intact striatum of all animals independent of treatment (treatment effect F = 4.232, p = 0.048, animal group effect F = 5.95, p = 0.003; Fig. 5a). The second potassium-evoked release revealed a trend for reduced DA levels in the control sides. In the lesioned sides, acute l-DOPA administrations did not significantly affect the amplitude of released DA after the first or second potassium ejection when compared with that found before l-DOPA was given (Fig. 5b).

Acute l-DOPA did not change Trise after the first KCl-evoked release in either control or lesioned sides in any of the animals (Fig. 4a). However, the Trise was significantly increased after the second KCl ejection after acute l-DOPA in the denervated sides of non-dyskinetic and the lesioned control animals (Tukey post hoc test p < 0.05; Fig. 4c), whereas Trise in the dyskinetic animals was significantly reduced after acute l-DOPA injection compared with pre-l-DOPA values on the lesioned side (two-factor anova with treatment and animal group as factors, F = 0.63, p = 0.435 for treatment, F = 0.01, p = 0.990 for animal group and F = 3.80, p = 0.040 for treatment and animal group interaction). Interestingly, Trise after the second potassium ejection was significantly increased after acute l-DOPA in the intact sides of non-dyskinetic animals compared with pre-l-DOPA measurements (Fig. 4c), while Trise was not affected by acute l-DOPA on the intact sides of the other animals (two-factor anova, F = 8.44, p = 0.007 for treatment effect, F = 4.65, p = 0.009 for animal group effect and F = 6.636, p = 0.002 for treatment and animal group interaction, Tukey post hoc test p < 0.05 for non-dyskinetic animal group pre vs. post-l-DOPA). Thus, the response in Trise was unique to the non-dyskinetic animals for both sides.

The reuptake pattern, as measured by T50, was significantly altered in the non-dyskinetic animals in the control side after the second KCl ejection (two-factor anova, treatment effect F = 8.125, p = 0.08, animal group effect F = 4.105, p = 0.015 and treatment and animal group interaction F = 4.76, p = 0.008; Fig. 4d). T50 was increased almost by 100% in the intact side of the non-dyskinetic animals (Tukey post hoc test p < 0.05). Hence, the reuptake time was increased from about 40 to 80 s.

Acute l-DOPA injection reduced the uptake rate in all animals by an average of about 49%, but this effect did not reach statistical significance (two-factor anova treatment effect F = 3.41, p = 0.074, treatment and animal group interaction F = 0.14, p = 0.937).

Tyrosine hydroxylase- and serotonin transporter-immunoreactive nerve fiber density

Immunohistochemical evaluations verified that the electrode positions were correctly located. The density of TH-positive nerve fibers confirmed the amphetamine-induced rotational behavior such that all animals included in the study showed severe reduction (> 98%) of TH-immunoreactive nerve fiber density in the lesioned compared with normal or intact sides of the striatum (data not shown).

Serotonin transporter-immunohistochemistry revealed that nerve fiber density was significantly decreased in DA-depleted striata of lesioned controls and of non-dyskinetic animals when compared with normal striatum (one-factor anovaF = 5.95, p = 0.009, Tukey post hoc test). The SERT-positive nerve fiber density in dyskinetic animals, on the other hand, was not significantly different when compared with normal controls (Fig. 6). No correlation was found between SERT-positive nerve fiber density and the amplitude of potassium-evoked released in the lesioned side of non-dyskinetic or dyskinetic animals (Fig. 6f).

Figure 6.

 SERT-immunohistochemistry in the dopamine-depleted striata of animals with dyskinetic (a) and non-dyskinetic (b) behavior and in dopamine depleted l-DOPA naïve (c) and normal control striatum (d), scale bar, 50 μm. Measurements of SERT-positive nerve fiber density, expressed as gray values, revealed reduced density in non-dyskinetic and lesioned/control striatum when compared with dyskinetic and intact striatum (e). No correlation was detected between SERT-positive nerve fiber density and concentration of potassium-evoked dopamine release in the lesioned striata of dyskinetic or non-dyskinetic animals after acute l-DOPA (f); *p < 0.05 and **p < 0.01.


Although the symptomatic treatment of PD using DA replacement therapy with l-DOPA has been used since the late 1960s, it is still not completely understood how l-DOPA treatment affects the DA release. The results from this study revealed that chronic l-DOPA treatment reduced the amplitude of KCl-evoked DA release in intact sides of the striatum, and that acute l-DOPA further attenuated the level of released DA. The amplitude of released DA ipsilateral to the denervation was, as expected, drastically reduced compared with the control sides, and both the Trise and T50 were significantly prolonged in all denervated sides when compared with intact striatum, as previously demonstrated (van Horne et al. 1992). Acute l-DOPA did not change the amplitude in the denervated side. However, acute l-DOPA decreased the Trise ipsilateral to the lesion in dyskinetic animals after the second KCl-evoked release, while the same parameter was increased in non-dyskinetic rats on both sides. T50 was changed only after the second KCl-evoked release in intact side of non-dyskinetic animals upon acute l-DOPA injection.

Chronic intermittent l-DOPA treatment reduced the peak amplitude (about 60%) of KCl-evoked DA release in the intact side of the striatum. Since the washout time for l-DOPA was 2 days and striatal l-DOPA levels become normalized within 2–3 h after the injection (DeJesus et al. 2000; Carta et al. 2006), the effect was monitored without the presence of l-DOPA. Acute l-DOPA further attenuated the amplitude of released DA. The results were surprising as microdialysis studies have demonstrated that exogenous l-DOPA increases the striatal DA levels in normal side (Miller and Abercrombie 1999). It is important to remember the differences between microdialysis and the current technique. While microdialysis measures the basal concentration of extracellular DA, chronoamperometry determines the change in DA concentration above basal level, evoked by potassium. However, high extracellular DA concentration, as revealed by microdialysis, should normally affect the DA autoreceptor and consequently, reduced release would be expected. This was clearly demonstrated in this study using chronoamperometry. Thus, the present results might explain, beside the loss of DA nerve fibers, why parkinsonian patients need to increase their l-DOPA doses over time.

Dopamine levels are known to be increased in the lesioned striatum after acute l-DOPA injection (Miller and Abercrombie 1999; Carta et al. 2006). Furthermore, l-DOPA levels are higher in lesioned striatum of dyskinetic than in non-dyskinetic animals, indicating that DA levels are higher in dyskinetic striata (Carta et al. 2006). However, in this study no difference was found in the amplitude of the release in the denervated sides, independent of treatment and behavior. As revealed by the TH-positive nerve fiber density in the denervated sides, almost no DA nerve fibers were present to convert l-DOPA to DA. However, l-DOPA can be converted to DA in serotonergic nerve fibers (Arai et al. 1995), and this event has been suggested as a possible mechanism involved in dyskinetic behavior. For instance, drugs acting on serotonergic receptors, can improve dyskinetic behavior (Kannari et al. 2001; Durif et al. 2004; Bara-Jimenez et al. 2005), and serotonergic hyperinnervation exacerbates dyskinesia, as revealed by grafting of serotonergic neurons into DA-depleted striatum (Carlsson et al. 2007). Thus, as the SERT-positive nerve fiber density was higher in dyskinetic than in non-dyskinetic animals, increased amplitude of the release in the dyskinetic striatum was expected but not achieved. Indeed, it has been demonstrated that l-DOPA can be converted in non-monoaminergic neurons in the striatum (Ikemoto et al. 1997). Thus, these neurons may participate in the conversion of l-DOPA in the non-dyskinetic animals.

Locally applied KCl causes a rapid depolarization of nerve terminals in the proximity of the carbon fiber electrode and the neurotransmitter is released into the extracellular space. In this study, two KCl ejections were made at each recording site in the lateral striatum. The rationale for studying two consecutive KCl-ejections was to enable measurements of the resting state release (first ejection) and release in a challenged system (second ejection) where the recently depolarized cells had been allowed to recover only for 120 s, i.e. full recovery had not been achieved between the two time points. The result revealed that the Trise time after acute l-DOPA was longer in the DA-denervated side of non-dyskinetic and lesion/control than in dyskinetic animals after the second KCl-evoked release. Thus, the time to reach maximal amplitude was much faster in dyskinetic than in non-dyskinetic animals, reaching times comparable to normal striatum. As there was no difference in the amplitude of the release in the denervated striata, data indicate that the changes can be attributed to the release and clearance mechanisms. It was clear that the shorter time interval to reach maximal concentration after potassium-evoked release was correlated with dyskinetic behavior.

The serotonergic nerve fiber density, as revealed by SERT-immunohistochemistry, was significantly lower in the non-dyskinetic and lesion/control animals compared with dyskinetic animals, which had a density similar to intact striatum. As serotonergic nerve fibers pass via the medial forebrain bundle, it is tempting to believe that all denervated animals had similar SERT-positive nerve fiber density after the DA denervation, prior to l-DOPA treatment. If so, l-DOPA might exert a trophic influence and induce regeneration. This is strengthened by the fact that dyskinetic behavior is increasing over time with l-DOPA treatment (Ahlskog and Muenter 2001). On the other hand, a variable serotonergic denervation might have predestined the animals to become dyskinetic versus non-dyskinetic after l-DOPA treatment, with prevalence to their degree of serotonergic denervation prior to l-DOPA treatment. However, both 5-HT and SERT levels are reduced in PD, with similar levels measured in dyskinetic as in non-dyskinetic patients (Kish et al. 2008), which does not favor the idea that l-DOPA induces sprouting. Indeed, the serotonergic nerve fibers density was not correlated to the released concentration of DA after acute l-DOPA.

In conclusion, chronic l-DOPA treatment reduced potassium-evoked DA release in intact striatum, an effect that was further strengthened during acute l-DOPA influence, while the DA reuptake was not affected. Interestingly, potassium-evoked DA release was not affected by either chronic or acute l-DOPA in terms of amplitude of the release in the denervated striata of either dyskinetic or non-dyskinetic animals, and the amplitude was very low and not dependent on SERT-positive nerve fiber density. The release kinetics were different for dyskinetic compared with non-dyskinetic animals which suggests that several types of nerve fibers might be involved in the conversion of l-DOPA to DA.


This study was funded by the Swedish Research Council Grant #09917, the Umeå University Medical Faculty Funds, and Konung Gustav V och drottning Victorias fond, Sweden. ML was financed from Parkinsonfonden and Hjärnfonden, Sweden.