L-DOPA-induced dyskinesia in adult rats with a unilateral 6-OHDA lesion of dopamine neurons is paralleled by increased c-fos gene expression in the subthalamic nucleus

Animals and drug treatments

Adult male Sprague–Dawley rats (Charles River, Wilmington, MA, USA) were maintained under a 12 h light : dark cycle with constant temperature and humidity. Food and water was available ad libitum. All experimental procedures were performed in accordance with the Institutional Animal Care and Use Committee guidelines at Boston University School of Medicine. Surgical procedures were performed under anesthesia with a mixture of ketamine (80 mg/kg) and xylazine (10 mg/kg) injected intraperitoneally (i.p.) at 1.3 mL/kg (Phoenix Pharmaceutical, St. Joseph, MO, USA).

Twenty-one rats were unilaterally depleted of dopamine following stereotaxic injections of 8.0 µg of freebase 6-hydroxydopamine (6-OHDA hydrobromide; Sigma Chemical, St. Louis, MO, USA) into the left rostral substantia nigra, pars compacta (SNc) (H, 2.8 mm, AP, 3.4 mm, L, 2.0 mm), and the left median forebrain bundle (H, 1.9 mm, AP, 4.2 mm, L, 1.2 mm) with the incisor bar at 0 mm. Four sham-operated rats were injected at the same stereotaxic coordinates with vehicle (saline plus 0.1% ascorbic acid). Vehicle or 6-OHDA was injected with a Hamilton syringe over 2 min, and the syringe was kept in place after the injection for another 5 min. Three weeks following the surgery, 17 6-OHDA-lesioned and four sham-operated rats received a daily s.c. injection of levodopa (L-DOPA methyl-ester; Sigma) at a dose of 6 mg/kg/day. L-DOPA was dissolved in vehicle solution (saline with 0.1% ascorbic acid) and was injected with the peripheral decarboxylase inhibitor benserazide (Sigma) at a dose of 15 mg/kg. Four 6-OHDA-lesioned rats received only one injection of L-DOPA at the same dose (lesion-acute group). All rats were injected with L-DOPA or vehicle between 09.00 and 11.00 h.

Behavioral testing

On every other day, rats in the chronic groups were injected with vehicle or L-DOPA and were returned to their cage. The occurrence and severity of dyskinetic movements was measured over a period of 3 h post-injection using a scoring method that was adapted from a previous study (Cenci et al., 1998). Each rat was observed for one out of every 20 min, for a total of 9 min in 3 h. Dyskinetic movements were classified into one of four subtypes: (1) locomotor, defined as a rotational behavior; (2) axial, defined as contralateral twisting posture of the neck and upper body away from the side of the lesion; (3) forelimb, defined as a repetitive rhythmic jerk or dystonic posturing of the contralateral forelimb; and (4) orolingual, defined as vacuous chewing movements or tongue protrusion and/or licking with a contralateral bias. Each of these four subtypes was rated by severity on a scale of 0–4 based on the following criteria. A score of 0 was given if the behavior was absent. A score of 1 was given if the behavior was occasional and occurred less than 30 s per minute. A score of 2 was given if the behavior occurred more than 30 s per minute, but was not continuous. A score of 3 was given if the behavior was continuously observed over 1 min, but could be interrupted by sensory distraction. The sensory distraction was a touch/rub with a pen on the animal's flank. A score of 4 was given if the behavior was continuous and was not interrupted by a sensory distraction.

Combining the subtype scores yielded an overall motor score for each animal during each 20-min interval. The maximum score for each 20-min interval was 16 (maximum score per dyskinesia subtype × four dyskinesia subtypes), and the maximum for each session was 144 (9 min × 16). Significant differences between average dyskinesia scores over the course of chronic L-DOPA administration were calculated with a non-parametric Kruskal–Wallis test and a Dunn's Multiple Comparison Test, significance at P < 0.05. All statistical analyses were performed using GraphPad Prism 3.0c for Macintosh (GraphPad Software, San Diego, CA, USA, http://www.graphpad.com). Three hours after the last injection of L-DOPA or vehicle, animals were sedated with CO₂, killed and their brain removed. Ten-micrometer-thick brain sections were produced on a cryostat at the level of the striatum (IA, 10.00), the STN (IA, 4.84) or the substantia nigra (IA, 3.80), according to Paxinos & Watson (1986), thaw-mounted on chromalum gelatin-coated glass slides and stored at −80 °C until further processing.

Tyrosine hydroxylase (TH) immunoreactivity

Immunohistochemistry was performed using a standard peroxidase-based method (Vectastain Elite ABC Kit; Vector Laboratories) on coronal sections obtained from fresh-frozen brains. Slides were fixed in 4% paraformaldehyde for 5 min, quickly washed in 0.2 m KPBS plus Triton X-100 and incubated for 1 h in 0.2 m KPBS with 5% normal goat serum (NGS) at room temperature. Sections were then incubated overnight at 4 °C with a polyclonal rabbit TH antibody (1 : 1000; Chemicon, Temecula, CA, USA) in 0.2 m KPBS, 1% NGS and 0.2% Triton X-100. The next day, sections were washed in 0.2 m KPBS plus 0.2% Triton X-100 and were incubated for 1 h at room temperature in affinity-purified biotinylated goat anti-rabbit IgG (1 : 500; Chemicon). Then, sections were washed in 0.2 m KPBS and processed with the avidin-biotin complex (ABC, Vector Laboratories, Burlingame, CA, USA). Immunolabeling was revealed following incubation of the sections in a DAB/Metal Concentrate (Pierce Laboratories, Rockford, IL, USA). Sections were dehydrated and mounted in Eukitt mounting medium. Quantitative analysis of TH immunolabeling in the SNc and ventral tegmental area (VTA) was carried out on digitized pictures by computerized densitometry utilizing NIH Image 1.61 software. The brain sections were placed on a light table and digitized using an AF Micro Nikkor 60 mm f/2.8D lens and a Sony CCD video-camera connected to a McIntosh computer. Relative optical density (OD) measurements were calculated by standardization against Kodak gelatin filters. Two sections per animal were quantified. The value of TH immunolabeling for each rat was the average from two sections. Differences in immunolabeling intensity between sides was calculated with a non-parametric Mann–Whitney test with P < 0.05 considered significant.

³ H-Mazindol-binding radioautography

³H-Mazindol binding was used to measure the density of dopamine re-uptake sites in the striatum. Ten-micrometer-thick fresh-frozen tissue sections were dried under a flow of air. Two sections per rat were rinsed for 5 min at 4 °C in 50 mm Tris buffer with 120 mm NaCl and 5 mm KCl to wash off endogenous ligand. Sections were then incubated for 40 min at 4 °C in 15 nm³H-mazindol (PerkinElmer Life Sciences, Boston, MA, USA; specific activity 21.0 Ci/mmol) in 50 mm Tris buffer with 300 mm NaCl, 5 mm KCl. Desipramine (0.3 µm; Sigma) was added to block binding to norepinephrine transporters. Non-specific binding was determined in the presence of 30 µm unlabeled benztropine (Sigma). Sections were then quickly rinsed in ice-cold buffer, distilled water and air-dried. All sections were apposed to Kodak BIOMAX MR X-ray films at room temperature for 35–45 days. The films were developed in Kodak D-19 for 3.5 min at 14 °C. Levels of ³H-mazindol labeling on the X-ray film autoradiographs from the striatum were quantified by computerized densitometry with NIH Image 1.55 software (Macintosh; http://www.zippy.nimh.nih.gov). The X-ray films were placed on a light table and digitized using an AF Micro Nikkor 60 mm f/2.8D lens and a Sony CCD video-camera connected to a McIntosh computer. Relative OD measurements were calculated in each region of interest by standardization against Kodak WRATTEN gelatin filters and subtracting the background OD of the film. Differences between the intact and lesioned side were compared with a non-parametric Mann–Whitney test (significance P < 0.05).

In situhybridization histochemistry

A ³⁵S-radiolabeled complementary RNA (cRNA) probe was transcribed in vitro from a c-fos cDNA (Rivest & Rivier, 1994). The cDNA was inserted into the transcription vector Bluescript SK, and was linearized with Sma1 restriction enzyme. Transcription of the antisense cRNA was performed for 2 h at 37 °C in the presence of 2.5 µm³⁵S-UTP (specific activity 1250 Ci/mmol, PerkinElmer Life Sciences) and 10 µm unlabeled UTP with ATP, CTP and GTP in excess. The template was digested with DNAse I. The labeled RNAs were purified by phenol/chloroform extraction and ethanol precipitation. The cRNA probe length was reduced to 100–150 nucleotides by partial alkaline hydrolysis to improve accessibility of the probes (Cox et al., 1984). Coronal sections at the level of the STN were processed for in situ hybridization histochemistry as previously described (Nielsen & Soghomonian, 2003). Sections were hybridized for 4 h at 52 °C with 4.0 ng of radiolabeled RNA probe in hybridization mix [40% formamide, 10% dextran sulfate, 4 × standard sodium citrate (SSC), 10 mm dithiothreitol, 1.0% sheared salmon sperm DNA, 1.0% yeast tRNA, 1 × Denhardt's solution]. The sections were subsequently washed in 50% formamide at 52 °C for 5 and 20 min, RNAse A (100 µg/mL; Sigma) for 30 min at 37 °C, and 50% formamide for 5 min at 52 °C. Sections were then dehydrated in ethanol and were coated with Kodak NTB3 nuclear emulsion diluted 1 : 1 with distilled water containing 300 mm ammonium acetate. Sections were stored for 5–7 days at room temperature in light-tight boxes, developed in Kodak D-19 for 3.5 min at 14 °C, lightly counterstained with hematoxylin and eosin, and mounted with Eukitt mounting medium. Sections adjacent to the sections processed for in situ hybridization histochemistry were processed with a Nissl counterstain.

Quantification of emulsion autoradiographs

First, the numbers of neuronal profiles labeled with c-fos on emulsion radioautographs or on Nissl-stained sections were measured by manual counting on digitized photomicrographs of the STN on the side of the 6-OHDA lesion. Photomicrographs were taken by light microscopy with a 20 × objective on a Nikon Eclipse 600 microscope connected to a monochrome SPOT Insight digital camera (Diagnostic Instruments Sterling Heights, MI, USA). Second, the area covered by reduced silver grains over individual neurons in the STN on the side ipsi- and contralateral to the 6-OHDA lesion was quantified for each rat. The analyses were conducted on emulsion-coated slides by computerized image analysis using the NIH Image 1.55 software. The emulsion radioautographs were scanned under bright-field illumination with a 60 × objective on an E600 Nikon microscope. The microscopic fields were viewed on a monitor using a Sony CCD video-camera connected to a McIntosh computer. Using the NIH software, a circle of constant diameter was superimposed over each sampled neuronal profile and the area covered by reduced silver grains within the limits of the circle was calculated and expressed as a number of pixels per cell. The reliability of the system was assessed prior to this study by showing a strong correlation between computerized and manual silver grain counting. c-Fos mRNA labeling was measured on a sample of 40–50 neurons per coronal section by moving non-overlapping frames over the STN. With the observer blind to the experimental groups, all neurons demonstrating labeling three times greater than background were measured until 40–50 neurons were analysed. Two slides from each animal were used for the quantitative analysis (i.e. 80–100 neurons/side/animal). The final value of mRNA labeling for each rat was the average value from the two sections. Differences between the intact and lesioned side were compared with a non-parametric Mann–Whitney test (significance P < 0.05).

6-OHDA lesions and behavioral effects of L-DOPA

A single de novo injection of L-DOPA induced axial and limb dyskinesias in about a quarter of all 6-OHDA-lesioned rats. This observation was made in the group of rats injected only once with L-DOPA (lesion-acute group) or the group of rats chronically injected with L-DOPA (lesion-chronic group). However, at Day 11 of a chronic daily L-DOPA administration, most 6-OHDA-lesioned rats (14 out of 17) demonstrated dyskinetic movements in response to an L-DOPA challenge. Thereafter, the severity of dyskinesias increased between Day 11 and 30 of a daily administration of L-DOPA (Fig. 1). However, the interindividual variability in the severity of dyskinesia scores was maintained throughout the injection schedule. This was illustrated by a highly significant correlation between individual dyskinesia scores measured at Day 11 and 30 of a daily administration of L-DOPA (Fig. 2). Among the 17 rats in the lesion-chronic L-DOPA group, two never developed dyskinesias whereas one rat had locomotor dyskinesia only. None of the sham-operated rats (sham-chronic group) that received chronic injections of L-DOPA developed dyskinesias.

Observation of sections labeled for TH immunohistochemistry revealed that all but one rat with a unilateral 6-OHDA injection had a marked loss of labeling in the ipsi- when compared with contralateral SNc (Fig. 3B). A smaller difference was seen in the adjacent VTA (Fig. 3B). Quantitative analysis confirmed a significant difference in TH immunoreactivity between the ipsi- and contralateral SNc (average relative OD value for the contra- and ipsilateral side: 0.532 ± 0.078 vs. 0.029 ± 0.011; P < 0.0001), but the difference in the VTA did not reach statistical significance (0.504 ± 0.079 vs. 0.359 ± 0.058; P = 0.079). The extent of loss of dopamine neurons was confirmed by ³H-mazindol-binding radioautography in the striatum (Fig. 3A). Quantitative analysis indicated a significant difference in ³H-mazindol binding in the striatum between the side ipsi- and contralateral to the lesion (average relative OD value for the contra- and ipsilateral side: 0.208 ± 0.003 vs. 0.088 ± 0.005; P < 0.0001). The only rat that had minimal loss of TH immunoreactivity also had a small loss of ³H-mazindol binding in the striatum. There was no significant correlation between TH immunoreactivity in the SNc or ³H-mazindol binding in the striatum on the side ipsi- or contralateral to the 6-OHDA lesion and the severity of dyskinesias (Fig. 4).

c-Fos mRNA levels in the STN

Observation of emulsion radioautographs by bright-field microscopy indicated that c-fos mRNA labeling was distributed over neuronal profiles in the STN. The average number of neuronal profiles labeled with c-fos mRNA in the STN was significantly lower than the number of neuronal profiles measured on adjacent Nissl-stained sections (mean number of profiles per mm² ± SEM: c-fos, 21.12 ± 1.95; Nissl, 90.88 ± 2.73; P < 0.0001). In the group of 6-OHDA-lesioned rats chronically injected with vehicle (lesion-vehicle group), c-fos mRNA levels in the STN were slightly above background and did not differ between the ipsi- and contralateral side (mean ± SEM: 19.42 ± 1.25 vs. 20.78 ± 1.26; P = 0.5) (Fig. 5). In the group of 6-OHDA-lesioned rats injected only once with L-DOPA (lesion-acute group), c-fos mRNA levels were significantly higher in the ipsi- compared with contralateral STN (mean ± SEM: 116.4 ± 67.27 vs. 19.41 ± 5.77; P = 0.029) (Fig. 5). However, this increase appeared to be mainly due to one rat that had dyskinesia. Indeed, c-fos mRNA levels were not significantly different between the ipsi- and contralateral STN in rats that did not exhibit dyskinesias (mean ± SEM: 42.51 ± 6.66 vs. 20.33 ± 8.06; P = 0.1) (Fig. 5). In the group of 6-OHDA-lesioned rats that received chronic injections of L-DOPA (lesion-chronic group), c-fos mRNA levels were significantly higher in the ipsi- compared with the contralateral STN (mean ± SEM: 83.77 ± 6.80 vs. 30.71 ± 1.41; P ≤ 0.0001) (Fig. 5). In sham-operated rats chronically injected with L-DOPA (sham-chronic group), there was no significant difference in c-fos mRNA levels between the ipsi- and contralateral STN (mean ± SEM: 17.33 ± 2.37 vs. 11.64 ± 3.69; P = 0.3) (Fig. 5).

The observation of emulsion radioautographs indicated that c-fos mRNA levels in the STN of 6-OHDA-lesioned rats chronically injected with L-DOPA were variable between rats. Rats with severe dyskinesias appeared to have stronger c-fos mRNA labeling than rats with little or no dyskinesia (Fig. 6A–C). Correlation analyses were carried out between dyskinesia scores and c-fos mRNA levels in the STN on the side of the lesion in the group of rats chronically injected with L-DOPA (lesion-chronic group). Consistent with the observations made on emulsion radioautographs, there was a significant correlation between c-fos mRNA levels and dyskinesia scores measured on the last day (Day 30) of L-DOPA administration (Fig. 7A). Separate correlation analyses were conducted for each dyskinetic component. Results indicated significant correlations between c-fos mRNA levels and axial and limb dyskinesia, but not between c-fos mRNA levels and locomotor or orolingual dyskinesia (Fig. 7B–D). Consistent with the observation that individual dyskinesia scores at Day 11 correlated with individual dyskinesia scores at Day 30, the severity of axial and limb dyskinesia scores measured at Day 11 was correlated with c-fos mRNA levels in the STN (Fig. 7E and F).