In vivo characterization of somatodendritic dopamine release in the substantia nigra of 6-hydroxydopamine-lesioned rats

Authors

  • Sophie Sarre,

    1. Department of Pharmaceutical Chemistry and Drug Analysis, Research group Experimental Pharmacology, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author
  • Hong Yuan,

    1. Department of Pharmaceutical Chemistry and Drug Analysis, Research group Experimental Pharmacology, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author
  • Nadine Jonkers,

    1. Department of Pharmaceutical Chemistry and Drug Analysis, Research group Experimental Pharmacology, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author
  • An Van Hemelrijck,

    1. Department of Pharmaceutical Chemistry and Drug Analysis, Research group Experimental Pharmacology, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author
  • Guy Ebinger,

    1. Department of Pharmaceutical Chemistry and Drug Analysis, Research group Experimental Pharmacology, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author
  • Yvette Michotte

    1. Department of Pharmaceutical Chemistry and Drug Analysis, Research group Experimental Pharmacology, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author

Address correspondence and reprint requests to Professor Y. Michotte, Department of Pharmaceutical Chemistry and Drug Analysis, Vrije Universiteit Brussel, Laarbeeklaan 103, B1090 Brussel, Belgium. E-mail: ymichot@fasc.vub.ac.be

Abstract

We investigated the effect of an injection of 6-hydroxydopamine (6-OHDA) into the rat medial forebrain bundle (MFB) on the degeneration and the function of the dopaminergic cell bodies in the substantia nigra (SN) 3 and 5 weeks after lesioning. After injection of 6-OHDA into the MFB a complete loss of dopamine content was apparent in the striatum 3 weeks after lesioning. In the SN the amount of tyrosine hydroxylase-immunoreactive dopamine cells decreased gradually, with a near-complete lesion (> 90%) obtained only after 5 weeks, indicating that neurodegeneration of the nigral cells was still ongoing when total dopamine denervation of the striatum had already been achieved. Baseline dialysate and extracellular dopamine levels in the SN, as determined by in vivo microdialysis, were not altered by the lesion. A combination of compensatory changes of the remaining neurones and dopamine originating from the ventral tegmental area may maintain extracellular dopamine at near-normal levels. In both intact and lesioned rats, the somatodendritic release was about 60% tetrodotoxin (TTX) dependent. Possibly two pools contribute to the basal dopamine levels in the SN: a fast sodium channel-dependent portion and a TTX-insensitive one originating from diffusion of dopamine. Amphetamine-evoked dopamine release and release after injection of the selective dopamine reuptake blocker GBR 12909 were attenuated after a near-complete denervation of the SN (5 weeks after lesioning). So, despite a 90% dopamine cell loss in the SN 5 weeks after an MFB lesion, extracellular dopamine levels in the SN are kept at near-normal levels. However, the response to a pharmacological challenge is severely disrupted.

Abbreviations used
ABC

avidin–biotin complex

DAext

extracellular dopamine

Ed

extraction fraction

LC

liquid chromatography

LOD

limit of detection

MFB

medial forebrain bundle

MTN

medial terminal nucleus

6-OHDA

6-hydroxydopamine

PBS

phosphate-buffered saline

SN

substantia nigra

TH

tyrosine hydroxylase

TH-IR

tyrosine hydroxylase immunoreactive

TTX

tetrodotoxin

VTA

ventral tegmental area

WW

wet weight

Dopaminergic neurones of the substantia nigra (SN) pars compacta send axon projections to the dorsal striatum via the medial forebrain bundle (MFB) (Fallon et al. 1978). These cells release dopamine at the somatodendritic level as well as at the terminal region.

Somatodendritric dopamine release exerts several functions. Its physiological relevance in MFB-lesioned rats was the subject of this in vivo study. Similar to striatal dopamine release, it is reported to be important in the control of movement (Crocker 1997), possibly by its effects on striatonigral D1 receptors (Robertson and Robertson 1989; Yurek and Hipkins 1993). The release of dopamine at the level of the SN can be considered determinative in the control of movement because the reticulate part of this structure together with the globus pallidus form the main output from the basal ganglia (Albin et al. 1989). Furthermore, some improvement in motor function has been seen in animal models of Parkinson's disease when dopaminergic transplants are implanted in the SN, even though the nigrostriatal pathway is not reconstructed (Yurek 1997; Collier et al. 2002). Recently, Bergquist et al. (2003), using the rotarod test combined with dual microdialysis, demonstrated that physiologically released dopamine in the SN pars reticulata has a role in overall motor function.

Somatodendritic dopamine targets D2 receptors on its own cell bodies, as well as D1 receptors located on striatonigral GABAergic afferents. There are also known interactions between nigral dopamine and glutamatergic transmission (reviewed by Morari et al. 1998), GABAergic transmission (Timmerman and Westerink 1995; Cobb and Abercrombie 2002) and serotonergic transmission (Thorréet al. 1997, 1998a, 1998b).

A common characteristic of somatodendritic release is that it is calcium sensitive (Cuello and Iversen 1978; Westerink et al. 1994). However, there have been reports of a Ca2+-independent component of nigral dopamine release (Bergquist et al. 1998; Hoffman and Gerhardt 1999; Chen and Rice 2001). This phenomenon is explained either by mobilization of intracellular calcium or by release via the dopamine transporter (Falkenburger et al. 2001; for review, see Adam-Vizi 1992). Somatodendritic release also seems to be depolarization dependent, because it is induced by high potassium concentrations (Robertson et al. 1991; Hoffman et al. 1997; Rice et al. 1997). Data from microdialysis studies suggest that it is also sensitive to the fast sodium channel blocker tetrodotoxin (TTX) as well as to reserpine (Heeringa and Abercrombie 1995; Sarre et al. 1998). However, little is known about the nature of the dopamine release after a 6-hydroxydopamine (6-OHDA) lesion of the MFB. This seems to be an uninvestigated but interesting area of research, especially as several authors have noticed that even after a > 95% depletion of the striatum, basal dialysate dopamine levels (Jonkers et al. 2000; Bergquist et al. 2003) in the SN were similar to those observed in normal rats. Hoffman et al. (1997) observed reduced dialysate dopamine and dihydroxyphenyl acetic acid (DOPAC) levels in the SN, but potassium-evoked dopamine release was unchanged.

The time of experimentation after lesioning may be critical as it is known that dopaminergic terminals are more sensitive to 6-OHDA than dopamine cell bodies (Malmfors and Sachs 1968; Jonsson and Sachs 1970). In general, measurements in the SN of 6-OHDA-lesioned rats are performed across a large time window ranging from 1 to 5 weeks after lesioning (Robertson and Robertson 1989; Orosz and Bennett 1992; Lee et al. 1996; Hoffman et al. 1997; Jonkers et al. 2000; Marti et al. 2000). At these time points, the degree of striatal dopamine depletion usually exceeds 90%. However, in the SN of these rats neither the degree of dopamine depletion nor the functionality of the remaining neurones have been considered. Because of the ability of somatodendritic dopamine release to influence motor behavior, it is important to establish the functional status of the lesioned dopamine cell populations within the SN at different time points.

We therefore investigated the effect of an injection of 6-OHDA into the rat MFB on the degeneration and the functionality of the dopaminergic cell bodies in the SN 3 and 5 weeks after lesioning. At both time points, we examined the dopamine content in the striatum. Tyrosine hydroxylase (TH) immunostaining combined with Nissl staining was used to count the dopamine cells in the SN and the ventral tegmental area (VTA) after lesioning with 6-OHDA. By means of in vivo microdialysis in freely moving rats, we investigated the functionality of the remaining cells by local perfusion with TTX (1 µm) and by a challenge with d-amphetamine (2 mg/kg i.p.) or the selective dopamine reuptake blocker GBR 12909 (20 mg/kg i.p.).

Experimental procedures

Chemicals

d-Amphetamine sulfate, dopamine.HCl, decane sulfonic acid, 6-hydroxydopamine.HBr, GBR 12909.2HCl, TTX and phosphate-buffered saline (PBS) were from Sigma (St Louis, MO, USA); sodium chloride, potassium chloride, calcium chloride hexahydrate, acetonitrile, orthophosphoric acid, sodium acetate trihydrate, disodium EDTA, sodium disulfite and hydrochloric acid were from Merck (Darmstadt, Germany); ascorbic acid was from Roche (Brussels, Belgium); dihydroxybenzylamine.HCl was from Janssen (Beerse, Belgium); and hydrogen peroxide, xylene and acetic acid were from Carlo Erba Reagentia (Val de Reuil, France). Vectastain rabbit IgG avidin–biotin complex (ABC) kit and Vector SK-4100 DAB kit were from Vector Laboratories (Burlingame, CA, USA) and DPX mountant was from from BDH Laboratory Suppliers (Poole, UK).

All aqueous solutions were prepared in water purified by a Seralpur Pro 90 CN system (Merck Belgolabo, Overijse, Belgium) and filtered through a membrane filter with pore size 0.2 µm.

6-OHDA lesion of the MFB and experimental groups

Animal experiments were carried out according to the national guidelines on animal experimentation and were approved by the Ethical Committee for Animal Experiments of the Faculty of Medicine of the University. All efforts were made to minimize animal suffering and the minimal number of animals necessary to produce reliable scientific data was used.

Male albino Wistar rats weighing 180–200 g were anaesthetized with a mixture of ketamine : diazepam (50 mg/kg : 5 mg/kg i.p.) and placed on a stereotaxic frame. The skull was exposed and a burr hole was drilled to introduce a syringe for injection of a 6-OHDA solution (containing 4 µg 6-OHDA per µL in 0.01% ascorbic acid, pH 5). The solution was injected into the left MFB according to the atlas of König and Klippel (1963); for rats of 180–200 g, coordinates relative to bregma were L − 1.5, A − 3.2 and V + 8.7. A total volume of 4 µL of the 6-OHDA solution was injected at a flow rate of 1 µL/min. The syringe was left in place for 2 min and was then slowly removed over a period of 1–2 min. The skin was sutured, the animals received 4 mg/kg ketoprofen i.p. as an analgesic and were allowed to recover before returning to the animal housing facilities. Either 3 or 5 weeks after the lesion of the MFB, the rats either underwent surgery for in vivo microdialysis experiments (n = 40) or were killed with an overdose of sodium pentobarbital. In the latter case, in one group of rats (n = 12), TH immunostaining and Nissl staining of the rat brain was performed for cell counting in the SN and the VTA. In the other group (n = 20), dopamine content of the striatum was determined.

We also created a combined MFB–VTA lesion (n = 8) in an attempt to deplete all dopamine neurones in the A9/A10 region. The approach was similar to that described for the MFB lesion, but a second injection of 6-OHDA (16 µg per 4 µL) was made in the VTA at following coordinates relative to bregma: L − 0.6, A − 5.1 and V + 7.3. Four animals were used for cell counting in the SN–VTA after TH and Nissl staining. In the other four rats, dialysate dopamine levels were determined in baseline conditions.

Experiments were also carried out in normal (intact) rats: 22 underwent microdialysis, six were used for cell counting in the SN–VTA and six for determination of dopamine content in the striatum.

TH immunostaining and Nissl staining of the brain

The rat was deeply anaesthetized with sodium pentobarbital (100 mg/kg i.p.) and perfused transcardially with saline for 5 min followed by 4% formalin for 10 min. The brains were removed and placed in vials containing 4% formalin for 2–3 days. The brains were cut using a rodent brain matrix in order to obtain a 4-mm segment containing the SN. This was processed and embedded in paraffin wax. Five-micrometre coronal sections were taken using a microtome (Leitz 1400 sledge microtome (Wetzlar, Germany)). The sections were left to dry overnight in an oven at 60°C.

The sections were deparaffinized and rehydrated. Endogenous peroxidase was quenched with 0.3% H2O2 for 30 min. The sections were washed with distilled water (three times each for 5 min) and incubated in citrate buffer 0.1 m pH 5.8 for 20 min at 60–80°C. They were then washed in distilled water followed by PBS for 5 min. Non-specific binding was blocked with 1.5% normal goat serum (Vectastain rabbit IgG ABC kit) for 60 min. This was followed by incubation with rabbit polyclonal anti-TH antibody (AB152; Chemicon, Temecula, CA, USA) at room temperature. After washing three times each for 5 min in PBS, the sections were incubated with the biotinylated secondary antibody for 60 min (Vectastain rabbit IgG ABC kit) followed by three 5-min washes with PBS. The horseradish peroxidase conjugate (Vectastain rabbit IgG ABC kit) was applied for 60 min followed by three rinses in PBS each for 5 min. 3,3′-Diaminobenzidine (Vector SK-4100), mixed with distilled water, buffer pH 7.5, H2O2 and Ni+ solution, was applied until staining was optimal as determined by light microscopy. The sections were then washed with tap water, dehydrated, cleared with xylene and coverslipped using DPX mountant.

For each TH-stained section, an adjacent section was stained with cresyl violet. These were used for structure identification as well as for cell counting. Indeed, TH cell counts are only a marker of the dopamine phenotype and a reduction in this measure does not necessarily imply cell death.

After TH and Nissl staining, images of the sections were digitalized with a camera (Sony DXC, Sony Belgium), connected to the light microscope and the computer (eight-bit grey images). Counting of the dopaminergic cell bodies in the SN–VTA was carried out on six sections per animal. A cell was counted when intact and round with a clear nucleus and/or cytoplasm. The number of SN–VTA neurones was expressed as the average of the counts obtained from six representative sections. The full extent of the structure in each section was examined in the lesioned and non-lesioned sides.

Only sections in which the medial and lateral parts of the SN–VTA were clearly separated by the medial terminal nucleus (MTN) of the accessory optic tract level (AP − 5.3 mm; Paxinos and Watson 1986) were selected for analysis of TH-immunoreactive (TH-IR) cell number (Fig. 1). This approach has been used by others to ensure that comparable rostrocaudal levels of the SN are sampled between animals (Lee et al. 1996; Kirik et al. 1998).

Figure 1.

Micrograph of TH-immunostained SN and VTA. (a) Normal rats, (b) 3 weeks after a unilateral 6-OHDA lesion of the MFB; (c) 5 weeks after a unilateral 6-OHDA lesion of the MFB; (d) 5 weeks following an MFB–VTA lesion.

Microdialysis experiments

Surgery

Microdialysis experiments were performed on intact rats and on 6-OHDA-lesioned rats either 3 weeks or 5 weeks after lesioning. The rats, weighing ± 250 g, were anaesthetized with a mixture of ketamine : diazepam (50 mg/kg : 5 mg/kg i.p.) and placed on a stereotaxic frame with bregma placed 1.0 mm higher than lambda. The skull was exposed and a burr hole was drilled to implant a guide cannula (CMA Microdialysis, Stockholm, Sweden). Guide cannulas were positioned 2 mm above the left SN according to the atlas of Paxinos and Watson (1986). For rats weighing < 270 g, coordinates relative to bregma were L − 1.4, A − 5.0, V + 6.5; for rats weighing > 270 g, coordinates relative to bregma were L − 2.0, A − 5.2, V + 7.0. After surgery, the rats received 4 mg/kg ketoprofen i.p. as an analgesic.

Brain microdialysis

Immediately after surgery, the rats were placed in cages and microdialysis probes with a membrane length of 2 mm were introduced via the cannula (CMA 10 microdialysis probes). The probes were perfused with modified Ringer's solution containing 147 mm NaCl, 1.1 mm CaCl2.6H2O and 4 mm KCl at a constant flow rate of 2 µL/min using a CMA 100 microdialysis pump. Animals were allowed to recover from surgery overnight.

Experiments

Dialysate collection was started the day after surgery (equilibration time ∼ 20 h). Samples were collected every 20 min. Before any pharmacological manipulation was performed, six dialysate samples were collected. The mean of these neurotransmitter dialysate concentrations was taken as basal value at time zero. For i.p. injection, 2.73 mg d-amphetamine sulfate to obtain 2 mg d-amphetamine was dissolved in 1 mL saline and 1 mL/kg was administered immediately after the baseline samples had been obtained (time 20 min). For the dopamine reuptake blocker, 23.2 mg GBR 12909.2HCl to obtain 20 mg GBR 12909 was dissolved in 2 mL saline and 2 mL/kg was administered to the rat, after taking baseline samples (time 20 min). TTX (1 µm) dissolved in Ringer's solution was administered locally by perfusion through the microdialysis probe. After collection of the baseline samples, TTX was perfused for 2 h from collection 7 (time 20 min) to 12 (time 120 min). From collection 13 onwards, the rat was perfused again with Ringer's solution.

No net flux microdialysis experiments were carried out in a separate group of animals to estimate the extracellular dopamine concentration (DAext) in the SN of intact and lesioned rats after 3 and 5 weeks. After collecting four baseline samples (Cin = 0 nm), every 2 h and in an ad random order, the perfusion fluid was switched to one containing a different concentration of dopamine (Cin = 0.5, 1 and 1.5 nm).

After the dialysis experiment, the rat was sedated with sodium pentobarbital and oriental ink was injected through the guide cannula at the level of the SN. The rat was killed with an overdose of the anaesthetic, decapitated and the brain was removed. The whole brain was stored in buffered formalin (4%) for 24 h, then coronal sections 50 µm thick were stained with cresylviolet. Microdialysis data from animals with a wrong probe position were not taken into account.

Determination of dialysate levels of dopamine

Determination of dialysate dopamine levels was performed by microbore liquid chromatography (LC) with a column internal diameter of 1.0 mm and automatic injection of 10-µL samples as described previously (Sarre et al. 1997). In summary, the assay of dopamine was based on ion-pair reversed-phase (C8) LC, coupled to amperometric detection (Decade, Antec, Leiden, The Netherlands). The mobile phase consisted of 28 mL acetonitrile and 200 mL of the following buffer: 0.1 m sodium acetate trihydrate, 20 mm citric acid monohydrate, 2 mm decane sulfonic acid and 0.5 mm sodium EDTA adjusted to pH 5.5. The oxidation potential was set at + 450 mV. The limit of detection (LOD) of the assay was 0.04 nm. Except for two animals in the MFB–VTA-lesioned group (see Results section), baseline dialysate levels of dopamine could be measured in all animals. When the dialysate dopamine levels were equal to or decreased to below the LOD, the concentration of dopamine in the sample was set at the LOD.

Determination of tissue levels of dopamine

To establish the extent of dopamine depletion in striatum, the method described by Izurieta-Sanchez et al. 1998) was used. Tissue dopamine content was expressed as micrograms of dopamine per g wet weight (WW) of tissue. The limit of detection of the LC system corresponds to a residual tissue dopamine content of less than 0.05 ng per g wet weight.

Statistical analysis

One-way anova followed by Tamhane post-hoc test was used to compare the dopamine content in the striatum between intact rats and rats 3 and 5 weeks after lesioning. Similarly, the numbers of cells in the SN and the VTA were compared between intact rats and rats 3 and 5 weeks after lesioning. One-way anova was used to compare the baseline dopamine levels in intact and lesioned animals. For the no net flux experiments, a linear equation was constructed from dialysate dopamine concentrations (Cout) collected during dopamine perfusion (Shippenberg et al. 2000). The net change in dopamine (Cin – Cout) was plotted against Cin. The DAext in the SN, and the extraction fracion of the probe (Ed) were calculated from the obtained equations and were also compared between intact rats and rats 3 and 5 weeks after lesioning using one-way anova.

To determine the statistical significance of differences in dopamine levels after administration of the drugs compared with baseline values, a one-way anova for repeated measures and student's paired t-test comparing the levels at baseline and after 40 min (amphetamine experiments), 120 min (TTX experiments) or 140 min (GBR 12909 experiments) were used. Two-tailed Mann–Whitney U-test was used to compare concentrations for different treatments at the same time point.

For all statistical analyses α = 0.05.

Results

Dopamine content in the striatum: effect of a 6-OHDA injection into the MFB (Table 1)

Table 1.  Dopamine content in the striatum of intact and MFB-lesioned rats
 Dopamine (µg per g WW)
  1. Results are expressed as mean ± SEM. *p < 0.05 versus intact (Tamhane test following one-way anova).

Intact (n = 6)12.37 ± 0.17
3 weeks (n = 11)0.20 ± 0.09*
5 weeks (n = 9)0.43 ± 0.16*

In the striatum of intact rats the mean ± SEM amount of dopamine found in the left hemisphere was 12.37 ± 0.17 µg per g WW (n = 6). This was not different from that in the right hemisphere (12.69 ± 0.19 µg per g WW; n = 6). Three weeks after lesioning, the denervation of the striatum was nearly complete; the remaining dopamine content was 1% of that in the intact hemisphere. No further decline in dopamine content was observed 2 weeks later. The lesion had no effect on the striatal dopamine content in the right, unlesioned hemisphere (data not shown).

Number of TH-IR cells in the SN and the VTA: effect of the MFB lesion as a function of time (Table 2 and Fig. 1)

Table 2.  Number of TH-IR and Nissl-stained cells in SN and VTA of intact rats and of rats following an MFB-lesion or a combined MFB–VTA lesion
 SNVTA
THNisslTHNissl
  1. Data are expressed as mean ± SEM. Cell counts obtained from the sections at the level of the MTN of the left hemisphere. *p < 0.05 versus intact rat, p < 0.05 versus 3 weeks MFB-lesioned rat, p < 0.05 versus 5 weeks MFB-lesioned rat (Tamhane test following one-way anova).

Intact rat (n = 6)90.97 ± 1.88101.08 ± 3.3752.61 ± 1.2356.17 ± 1.55
3 weeks (n = 6)16.11 ± 0.98*26.08 ± 1.66*24.17 ± 0.96*28.50 ± 1.07 *
5 weeks (n = 6) (MFB lesion)0.72 ± 1.16*11.94 ± 1.43*12.39 ± 0.72*22.64 ± 0.72*
5 weeks (n = 4) (MFB–VTA lesion)1.16 ± 0.46*10.50 ± 0.89*4.80 ± 0.96*7.33 ± 0.89*

In the SN, there was a significant decrease in the amount of TH-IR cells on the lesioned side as a function of time (anova: F2,107 = 1548.1, p < 0.001). After 3 weeks 18% of the neurones remained. This decreased further to less than 1% by 5 weeks after lesioning. A significant reduction in cell number was also observed after Nissl staining (anova: F2,107 = 424.9, p < 0.001). After 3 weeks 26% of the neurones remained. This decreased further to 12% by 5 weeks after lesioning. In the combined MFB–VTA-lesioned rats, the amount of TH-IR and Nissl-stained cells 5 weeks after lesioning was not significantly different from that in MFB-lesioned rats at the same time point. As expected, on the intact side the number of TH-IR or Nissl-stained cells was constant as a function of time (data not shown).

In the VTA there was a significant decrease in the number of TH-IR cells on the lesioned side as a function of time (anova: F2,107 = 434.3, p < 0.001). This decrease was less pronounced than that observed in the SN. After 3 weeks 46% of the TH-IR neurones remained. This decreased further to 24% by 5 weeks after lesioning. A significant reduction in cell number was also observed by Nissl staining (anova: F2,107 = 188.2 p < 0.044). After 3 weeks 51% of the neurones remained, decreasing to 40% by 5 weeks after lesioning. In the rats with combined MFB–VTA lesions, TH and Nissl staining revealed that about 10% of the dopaminergic cells remained in the VTA 5 weeks after lesioning. As expected, on the intact side the number of TH-IR and Nissl-stained cells was constant as a function of time (data not shown).

Extracellular dopamine levels in the SN: effect of the MFB lesion as a function of time (Fig. 2)

Figure 2.

No net flux plot of basal dopamine levels in SN of intact rats (n = 7) and of rats 3 weeks (n = 6) and 5 weeks (n = 5) after MFB lesioning. Plots show the mean ± SEM gain or loss of dopamine (Cin − Cout) as a function of Cin (0, 0.5, 1 and 1.5 nm) and the average linear regression of the data in the different groups. The Cin at which Cin – Cout = 0 equals the real DAext in the SN. The slope of the regression corresponds to the extraction fraction of the probe (Ed). The y-intercept corresponds to the basal dopamine dialysate level. The mean ± SEM for DAext and Ed were calculated for each group. There was no significant effect of the lesion on these variables.

In intact rats, the mean ± SEM dialysate dopamine levels in the SN were 0.19 ± 0.03 nm (n = 15). Surprisingly, lesioning of the nigrostriatal tract at the level of the MFB had no significant effect on the baseline release of dopamine in the SN. Three weeks after lesioning, dialysate dopamine was 0.21 ± 0.03 nm (n = 14), which was not significantly different from levels found in intact animals, nor from those determined after 5 weeks: 0.19 ± 0.02 nm (n = 15).

In the combined MFB–VTA-lesioned rats, a mean dialysate dopamine level of 0.14 ± 0.06 nm (n = 4) was observed in the SN. This value was not significantly different from those obtained in the other groups. However, in two of the four animals in this group, the dialysate dopamine level was below the LOD (0.04 nm).

No net flux experiments confirmed that there was no significant effect of the lesion on the DAext in the SN nor on the Ed of the microdialysis probe (Fig. 2). Quite a large variation was observed at 3 weeks after lesioning.

Effect of TTX on basal dialysate levels of dopamine in the SN: effect of the MFB lesion as a function of time (Fig. 3)

Figure 3.

Effect of local perfusion of 1 µm TTX into the SN on extracellular dopamine levels of the SN of intact rats (open bars, n = 4) and in rats 3 weeks (dotted bars, n = 5) and 5 weeks (black bars, n = 6) after injection of 6-OHDA into the MFB. Data are mean ± SEM. The baseline value at time 0 is the mean of six values. *p < 0.05 versus baseline value (one-way anova for repeated measures followed by Student's paired t-test). LOD1 indicates the LOD for the dopamine assay from Heeringa and Abercrombie (1995); LOD2 indicates that achieved in the present study (signal to noise ratio 3 : 1).

Figure 3 demonstrates the effect of local infusion of 1 µm TTX on dialysate dopamine levels in the SN in intact rats and in rats 3 and 5 weeks following a 6-OHDA lesion of the MFB. The 6-OHDA lesion had no effect on the TTX sensitivity of dopamine release in the SN. Nigral dopamine release was mainly TTX dependent in the three groups studied. A significant decrease in dialysate levels to about 40% of the initial value was observed in all three groups. In nine of the 15 rats, values decreased to below the LOD of 0.04 nm. Reintroduction of Ringer's solution brought dopamine levels back to levels before TTX perfusion.

Effect of d-amphetamine on basal dialysate levels of dopamine in the SN: effect of the MFB lesion as a function of time (Fig. 4)

Figure 4.

Effect of d-amphetamine (AMPH) 2 mg/kg i.p. on extracellular dopamine levels in the SN. Data are mean ± SEM. The baseline value at time 0 is the mean of six values. *p < 0.05 versus baseline value (one-way anova for repeated measures followed by Student's paired t-test). §p < 0.05 for intact rats versus 5 weeks after lesioning, #p < 0.05 for 3 versus 5 weeks after lesioning (Mann–Whitney U-test). Arrow indicates administration of drug.

Figure 4 compares the influence of systemically administered d-amphetamine (2 mg/kg i.p.) on dialysate dopamine levels in the SN of intact rats and in rats 3 or 5 weeks after 6-OHDA lesioning of the MFB. In intact rats, dialysate dopamine levels in the SN increased significantly after a d-amphetamine challenge from mean ± SEM 0.19 ± 0.05 to 1.16 ± 0.38 nm (anova: F10,40 = 5.41, p < 0.001, n = 5). Three weeks after the lesion, 2 mg/kg d-amphetamine i.p. induced a significant increase in dialysate dopamine levels from 0.19 ± 0.04 to 0.68 ± 0.13 nm (anova: F10,30 = 15.38, p < 0.001, n = 4). Five weeks after the lesion, the effect of 2 mg/kg d-amphetamine was attenuated. Dialysate dopamine levels in the SN increased from 0.18 ± 0.05 to 0.32 ± 0.05 nm (n = 5). However, this effect did not reach statistical significance. The peak effect (time 40 min) of amphetamine on dialysate dopamine 5 weeks after the lesion was significantly different from the peak effect in intact rats and that 3 weeks after lesioning (p < 0.05). No significant difference in effect was observed between the effects of d-amphetamine in intact rats and 3 weeks after lesioning.

Effect of the dopamine reuptake blocker GBR 12909 on basal dialysate levels of dopamine in the SN: effect of the MFB lesion as a function of time (Fig. 5)

Figure 5.

Effect of the selective dopamine reuptake blocker GBR 12909 (GBR) 20 mg/kg i.p. on extracellular dopamine levels in the SN. Data are mean ± SEM. The baseline value at time 0 is the mean of six values. *p < 0.05 versus baseline value (one-way anova for repeated measures followed by Student's paired t-test). $p < 0.05 for intact rats versus 3 weeks after lesioning, §p < 0.05 for intact rats versus 5 weeks after lesioning. #p < 0.05 for 3 weeks versus 5 weeks after lesioning.

Figure 5 compares the influence of the systemic administration of the selective dopamine reuptake blocker GBR 12909 (20 mg/kg i.p.) on dialysate dopamine levels in the SN of intact rats and in rats 3 or 5 weeks after 6-OHDA lesioning of the MFB. In intact rats, dialysate dopamine levels in the SN increased significantly after GBR 12909 to reach a plateau after 60 min from mean ± SEM 0.21 ± 0.06 to 1.22 ± 0.34 nm at 140 min (anova: F10,50 = 6.95, p < 0.001, n = 6). Three weeks after the lesion, the effect of GBR 12909 was less pronounced. There was a significant increase in dialysate dopamine levels from 0.24 ± 0.06 to 0.74 ± 0.12 nm at 140 min (anova: F10,40 = 9.99, p < 0.001, n = 5). Five weeks after lesioning, the effect of GBR 12909 was even smaller. Dialysate dopamine levels in the SN increased significantly from 0.27 ± 0.04 to 0.36 ± 0.07 nm (anova: F10,30 = 8.56, p < 0.001, n = 4). Some 3 and 5 weeks after lesioning, the dopamine levels measured 140 min after GBR 12909 administration were significantly different from those in intact rats (p < 0.05). There was also a significant difference between the effects of GBR 12909 in rats 3 and 5 weeks after lesioning.

Discussion

In this study, we compared the degeneration of dopaminergic neurones in the SN and the striatum after injection of 6-OHDA into the MFB. A near-complete degeneration of the striatum was observed 3 weeks after lesioning. Two weeks later there was no further change. TH immunostaining and Nissl staining showed that the degeneration at the level of the SN was more gradual, only reaching > 90% after 5 weeks. This gradual degeneration was reflected in an attenuated response of the remaining cells in vivo to pharmacological manipulation with d-amphetamine and the dopamine reuptake blocker GBR 12909 at both time points. Basal dopamine release as well as its TTX sensitivity were not affected by the lesion.

This study confirms that dopaminergic cell bodies in the SN–VTA degenerate more gradually than the nerve terminals in the striatum. The gradual retrograde degeneration of the cell bodies is in line with data from others (Kirik et al. 1998) and has also been seen after intrastriatal injection of 6-OHDA, although the final degree of denervation is less pronounced in the latter case (Sauer and Oertel 1994; Lee et al. 1996; Kirik et al. 1998; Grant and Clarke 2002).

Our TH-immunostaining data indicate that 99% nigral degeneration is achieved by 5 weeks after lesioning. However in the VTA, 25% of the TH-positive cells still remain. Kirik et al. (1998) observed a similar loss of cells in the SN and VTA (95% and 80%) 8 weeks after lesioning using 13.5 µg 6-OHDA. Establishing loss of dopaminergic neurones requires concomitant loss of a dopamine marker (e.g. dopamine uptake or TH immunoreactivity) and loss of a general neuronal marker, such as staining for Nissl (Sauer and Oertel 1994; Przedborski et al. 1995). Our data show a near-complete loss of TH-IR cells 5 weeks after the MFB lesion, but with about 10% Nissl-positive cells remaining in the SN. This disproportionate loss may occur because some neurones in the SN pars compacta and the VTA are non-dopaminergic (Ljungdahl et al. 1975; Van der Kooy et al. 1981). Alternatively, some portion of the neurones may lose their dopaminergic markers without dying (Przedborski et al. 1995). However, the fact that extracellular dopamine was still measurable suggests that dopaminergic cells may still be present 5 weeks after creating the lesion.

There was no effect of the MFB lesion on dialysate dopamine levels in the SN, both 3 and 5 weeks after lesioning. This is in agreement with a recent study by Bergquist et al. (2003) who reported similar baseline dialysate dopamine levels to those observed in intact rats, 6 weeks after lesioning, despite a 99% lesion of the striatum. These authors used nigral dialysate DOPAC levels as an indicator of dopamine cell loss in the SN, and observed a reduction to 12% of that seen in intact rats. The value of dialysate DOPAC levels for estimating neuronal loss rather than dopamine levels has been demonstrated before in the striatum (Castañeda et al. 1990; Wachtel and Abercrombie 1994; Sarre et al. 1996). However, in this study this was not pursued as our analytical assay was aimed at determining very low concentrations of dopamine in the lesioned SN. The dopamine metabolites were eluted in the void, thereby enhancing the LOD for the dopamine assay.

No net flux experiments also showed that there was no effect of the lesion on DAext in the SN. In contrast to previous experiments carried out in the nucleus accumbens of 6-OHDA-lesioned rats (Parsons et al. 1991), we did not see a decrease in the Ed of the probe. Several authors have suggested that this parameter provides an index of dopamine uptake in vivo (Justice 1993; Parsons and Justice 1994; for review: Shippenberg and Thompson 1997). In this study, we clearly showed reduced dopamine reuptake (GBR 12909 experiments) resulting from the MFB lesion. This was, however, not accompanied by a reduction in Ed, supporting the suggestion that diffusion, rather than uptake, is important in determining extracellular dopamine concentrations (Cragg et al. 2001; see further).

In the striatum, dialysate dopamine levels are maintained as long as the degree of neurodegeneration is 80% or less. To account for this effect, several compensatory mechanisms have been described, such as increased firing rate and release by remaining dopamine neurones, decreased dopamine breakdown and increased dopamine diffusability (Stricker and Zigmond 1976; Robertson and Robertson 1989; Zigmond et al. 1990; Hornykiewicz 1993). As a result of these compensatory mechanisms, dialysate dopamine levels in the striatum decrease only moderately despite an 80–95% loss of dopaminergic neurones. When the loss of dopaminergic cells exceeds the limit of 95%, the compensatory mechanisms are no longer sufficient and dialysate dopamine levels decrease dramatically (Robinson and Whishaw 1988; Castañeda et al. 1990). It is possible that compensatory mechanisms exist in the SN to maintain extracellular dopamine after dopaminergic denervation, as has been described for the striatum. This hypothesis is strengthened by reports showing that the remaining dopaminergic neurones in the SN increase their rate of discharge (Hollerman et al. 1986; Hollerman and Grace 1988; Herkenham et al. 1991). The fact that in our microdialysis experiments extracellular dopamine was maintained at near-normal levels 5 weeks after lesioning, despite a lesion size of at least 90% (Nissl staining data), suggests that, as in the striatum, the compensatory mechanisms remain sufficient until a near-complete lesion is obtained.

Alternatively, it is possible that the remaining dopamine cells in the VTA, as shown by the TH and Nissl counts, contribute to the dopamine concentrations measured in the dialysate, even at 5 weeks after lesioning. Although we were not able to determine the extent of dopamine loss in the SN and the VTA in rats that underwent microdialysis, the fact that in two of the four rats with a combined MFB–VTA lesion the dopamine levels were not detectable suggests that the VTA may contribute to nigral dopamine levels measured with microdialysis after an MFB lesion. If this is the case, one might question whether the microdialysis technique selectively samples dopamine from the SN or whether several dopamine sources (including the VTA) are being sampled simultaneously. Because histological verification of the probe implantation clearly showed that the probe was never positioned at the level of the VTA in our study, we doubt that exocytotic tegmental dopamine contributed to the measured dopamine levels. However, TTX-insensitive volume transmission (Agnati et al. 1995) of dopamine from the VTA to the SN could account for the maintenance of the extracellular levels of dopamine after 6-OHDA lesioning of the MFB. Indeed, it is known that dopamine relies on extracellular diffusion to reach receptor sites within the SN (Groves and Lindner 1983), non-synaptic release from dopaminergic soma has been demonstrated (Jaffe et al. 1998) and electrophysiological data suggest, in contrast to the striatum, that diffusion, rather than uptake is the most important determinant of the dopamine time course in the SN (Cragg et al. 2001). In addition, Van Horne et al. (1992) described a better diffusion of exogenously applied dopamine in 6-OHDA-lesioned rats compared with intact rats.

Finally, contribution to dopamine levels of noradrenaline projections cannot be excluded. Indeed, although not the subject of this study, we were able to measure noradrenaline in the baseline dialysates of both intact and lesioned rats (data not shown). Taken together, it is reasonable to assume that the combination of compensatory mechanisms and the contribution of different dopamine sources result in the sustained dopamine levels.

Local administration of TTX is often used in vivo to verify the voltage dependency of the neurotransmitter released. Infusion of TTX into the SN has been shown to decrease dialysate dopamine to undetectable levels (Robertson et al. 1991; Santiago et al. 1992; Heeringa and Abercrombie 1995; Sarre et al. 1998), suggesting that under normal conditions somatodendritic dopamine is released via the mechanism of exocytosis.

For the first time, we show TTX sensitivity of the dialysate dopamine levels of the SN after lesioning of the MFB. Because dopamine levels lower than the LOD were considered to be equal to the LOD, our data suggest that at least 60% of the measured dopamine concentration is sodium-channel dependent. In contrast with the above-mentioned former studies, we were still able to measure dopamine levels during TTX perfusion in six of the 15 animals studied. This is probably because of the superior sensitivity (see Fig. 3) of our analytical assay. This implies that a maximum of 40% of the somatodendritic release may be of non-vesicular origin, supporting the suggestion that volume transmission might contribute to the nigral dopamine levels. Partial TTX independence of somatodendritic release has been observed before in microdialysis studies of the A10 region (Kalivas and Duffy 1991). More recently, Falkenburger et al. (2001) proposed that dendrodendritic release occurs through reversal of the dopamine transporter. It seems possible that two different pools contribute to basal dopamine levels in the SN: a fast sodium channel-dependent portion and a TTX-insensitive one originating from diffusion of dopamine.

Dialysate dopamine levels increased significantly in reponse to a challenge with d-amphetamine (2 mg/kg i.p.) 3 weeks after lesioning. The peak level of dialysate dopamine did not differ from that induced by d-amphetamine in intact rats. Two weeks later, amphetamine had no significant effect on nigral dopamine levels. These findings are in agreement with those of Hoffman et al. (1997), who showed that amphetamine-induced (250 µm for 40 min) dopamine release in lesioned SN was attenuated by the lesion. However, potassium-evoked release was not affected.

Similar effects have been described in the striatum of 6-OHDA-lesioned rats. As long as the lesion size did not exceed 95%, the amphetamine-induced dopamine release in dopamine-depleted striatum was comparable to that in intact striatum. When the dopamine depletion was > 95%, dialysate dopamine levels no longer increased after systemic administration of amphetamine (Robinson and Whishaw 1988; Castañeda et al. 1990). These authors suggested that presynaptic compensatory changes in the remaining dopaminergic neurones exist to normalize extracellular dopamine concentrations and that this contributes to recovery of function. Once the lesion exceeds 95% this mechanism is lost.

The attenuated response to amphetamine is accompanied by a reduced effect of the dopamine reuptake blocker GBR 12909 on nigral dopamine release as a function of time. Because loss of dopamine transporter (Joyce 1991) and [3H]dopamine uptake (Altar et al. 1987) have been shown accurately to represent the degree of dopamine loss, these data indicate a loss of dopaminergic cells and a reduced clearance of dopamine (Hoffman and Gerhardt 1998) in the SN. Similar to observations in the striatum (Miller and Abercrombie 1999), after a severe lesion, the effects of GBR 12909 are practically abolished.

Our data therefore suggest that the effect of d-amphetamine on dialysate dopamine levels of dopamine-depleted animals is comparable in striatum and SN. Three weeks after the lesion, the number of dopaminergic cells in the SN is reduced, but not abolished. These cells may show compensatory changes enabling a response to amphetamine. Indeed, there is probably less dopamine release than in intact rats but there is also less reuptake, resulting in a similar effect in intact rats 3 weeks after lesioning. Two weeks later, near-normal dopamine levels are measured, but the effect of GBR 12909 is strongly attenuated. The amphetamine response is also attenuated suggesting that the remaining dopaminergic cells in the SN and/or the VTA at that time are not able to fully respond to this pharmacological challange.

In summary, injection of 6-OHDA into the MFB caused a complete loss of dopamine content in the striatum 3 weeks after lesioning. Two weeks later there was no further change. In the SN, the number of dopaminergic cells decreased more gradually, with a near-complete lesion (> 90%) obtained only after 5 weeks, indicating that neurodegeneration of the nigral cells was still ongoing when total degeneration of the striatum had already been achieved. Baseline extracellular dopamine levels in the SN were not altered by the lesion. A combination of compensatory changes of the remaining neurones and dopamine originating from the VTA may maintain extracellular dopamine at near-normal levels. In both intact and lesioned rats, the somatodendritic release was about 60% TTX dependent. Possibly two pools contribute to the basal dopamine levels in the SN: a fast sodium channel-dependent portion and a TTX-insensitive one originating from diffusion of dopamine. The amphetamine- and GBR 12909-evoked dopamine release were attenuated after a near-complete denervation of the SN (5 weeks after lesioning). So, despite a dopaminergic cell loss of 90% in the SN 5 weeks after an MFB lesion, extracellular dopamine levels in the SN are kept at near-normal levels. However, the response to a pharmacological challenge is severely disrupted. These data add to the emerging insights that somatodendritic dopamine release has important physiological functions.

Acknowledgements

The authors wish to acknowledge the excellent technical assistance of Mr G. De Smet, Mrs C. De Rijck, Mrs R. Berckmans and Mrs R.-M. Geens. The scientific input and suggestions from Dr I. Smolders and R. Clinckers are greatly appreciated. We also thank Dr Michael O'Neill from Eli Lilly & Co Ltd (Lilly Research Centre, UK) for his help in developing the TH-immunostaining technique, and Dr A. Michotte, Dr M. Marichal and Dr L. Bauwens for the use of their laboratory facilities. N. Jonkers has a research grant (no. 961310) from the Vlaams Instituut voor de Bevordering van het Wetenschappelijk-Technologisch onderzoek in de Industrie (IWT). This research was also supported by grant G.0133.99 from the National Fund for Scientific Research (FWO).

Ancillary