Introduction

Amphetamine (AMPH) is the prototypal psychoestimulant, and AMPH derivatives are widely used as drugs of abuse. It is classically considered that AMPH acts by promoting dose- and time-dependent increases in extracellular dopamine (DA) in regions of the brain rich in DA nerve terminals (Kuczenski and Segal 1989). It has been shown that amphetamine impairs DA re-uptake (Dubocovich and Zaniser 1985), increases DA synthesis (Uretsky and Snodgrass 1977), and inhibits MAO activity (Green and El Hait 1978). However, several studies have shown dissociation between dopamine levels and behavioral responses (Kuczenski and Segal 1989, 1990), suggesting that DA release is necessary but by itself not sufficient to produce the characteristic AMPH-induced responses, and that other neurochemical systems may indeed be more important than has been generally accepted. The pre-frontocortical glutamatergic, the serotonergic (5-HT) and, particularly, the noradrenergic (NA) systems appear to be involved.

It has been shown that AMPH alters noradrenergic (NA) transmission (Florin et al. 1994; Ramirez and Wang 1986) and that manipulation of the NA system influences AMPH-induced behavior (Archer et al. 1986). Interestingly, it has been shown that stimulation of cortical α1-adrenergic receptors by NA is involved in the locomotor activating effects of AMPH (Blanc et al. 1994; Darracq et al. 1998). Finally, extensive behavioral data support the involvement of 5-HT in the locomotor effects of AMPH, although controversial results have been reported (Neill et al. 1972; Lipska et al. 1992). However, the role and possible interactions between these neurochemical systems in AMPH-induced responses have not been totally clarified. In addition, anatomical evidence in support of neurochemical and behavioral data is minimal.

The striatum, including dorsal striatum and ventral striatum/nucleus accumbens, is of particular interest to the study of these questions, and particularly the possible interactions between systems. The striatum is densely innervated by nigral DA fibers, cortical glutamatergic afferents, and 5-HT fibers, but not significantly innervated by NA afferents (see Aston-Jones et al. 1995 for review). In addition, it is known that characteristic AMPH-induced behavioral responses such as locomotor activity or stereotypy are mediated by striatal neurons (Pijnenburg et al. 1976; Sharp et al. 1987), and AMPH-induced responses in striatal neurons can be observed and quantified using Fos immunohistochemistry (Graybiel et al. 1990; Labandeira-García et al. 1996; Morgan and Curran 1991).

In the study reported here, we have investigated the mechanism involved in the effects of AMPH on the striatal neurons, and particularly the involvement of the NA and 5-HT systems and their possible interaction, using a double behavioral and immunohistochemical approach. We have studied the changes in locomotor activity and the expression of Fos in striatal neurons induced by systemic administration of D-amphetamine in normal rats and in rats subjected to treatment with the α1-adrenergic receptor antagonist Prazosin and/or acute or chronic lesion of the serotonergic system, or DA denervation.

Materials and methods

Experimental design

Adult female Sprague-Dawley rats (Letica, Barcelona, Spain) each weighing about 200 g at the beginning of the experiment were used. All experiments were carried out in accordance with the "Principles of laboratory animal care" (NIH publication Nº 86-23, revised 1985), as well as in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC) and the Spanish laws. Rats were divided into five groups (A–E). Rats in group A were injected intraperitoneally (i.p.) with Prazosin (0.5 mg/kg, n=10) or saline (n=10), and 30 min later with D-Amphetamine (Sigma; 1 mg/kg, n=10, group A1; or 5 mg/kg, n=10, group A2). Rats in group B were subjected to 5-HT lesions with DL-p-chlorophenylalanine (pCPA, see below), and then injected with Prazosin (0.5 mg/kg, n=10) or saline (n=10), and 30 min later with D-Amphetamine (1 mg/kg, n=10, group B1; or 5 mg/kg, n=10, group B2). Rats in group C were subjected to serotonergic denervation with 5,7-dihydroxytryptamine (5,7-DHT, see below) and at least 1 month later injected with Prazosin (0.5 mg/kg, n=10) or saline (n=10), and 30 min later with D-Amphetamine (1 mg/kg, n=10, group C1; or 5 mg/kg, n=10; group C2). Rats in group D were subjected to unilateral lesion of the DA system with 6-hydroxydopamine (6-OHDA) and at least 1 month later injected with Prazosin (0.5 mg/kg, n=10) or saline (n=10), and 30 min later with D-Amphetamine (1 mg/kg, n=10, group D1; or 5 mg/kg, n=10, group D2). Rats in group E (i.e., controls; n=5) were injected (i.p.) with saline and 30 min later with a second injection of saline.

Lesion of the serotonergic and the dopaminergic systems

All surgery was performed under equithesin anesthesia (3 ml/kg i.p.). Lesion of the serotonergic (5-HT) system was performed by injection of DL-p-chlorophenylalanine methyl ester hydrochloride (pCPA; Sigma; acute lesion, group-B rats) or 5,7-dihydroxytryptamine (5,7-DHT; Sigma; chronic lesion, group-C rats). The pCPA (300 mg/kg i.p.) was administered 24 and 1 h before injection of AMPH. The 5,7-DHT was stereotaxically injected into the medial raphe and into the dorsal raphe nuclei (AP=−7.8; L=0; V=8.7 and 6.7, respectively; tooth bar at −5). Each injection consisted of 6 μg of 5,7-DHT (weight of free base) dissolved in 3 μl of sterile saline containing 0.2% ascorbic acid. The solution was injected using a 5-μl Hamilton syringe coupled to a motorized injector (Stoelting) at 0.5 μl/min, and the canula was left in situ for 5 min between the two injections and after the second injection. Thirty min prior to surgery, rats received desipramine to prevent damage to the noradrenergic system. Lesions were confirmed in Cresyl Violet sections and in sections immunostained for 5-HT (see below).

Unilateral lesions of the DA system were performed by injection into the right medial forebrain bundle of 12 μg of 6-OHDA HBr (Sigma) in 4 μl of sterile saline containing 0.2% ascorbic acid. The stereotaxic coordinates were 3.7 mm posterior to bregma, 1.6 mm lateral to midline, 8.8 mm ventral to the skull at the midline, in the flat skull position (Paxinos and Watson 1986). The solution was injected using a 5-μl Hamilton syringe coupled to a motorized injector (Stoelting) at 0.5 μl/min, and the canula was left in situ for 2 min after injection. Thirty min prior to surgery, rats received desipramine (Sigma, 25 mg/kg i.p.) to prevent uptake of 6-OHDA by noradrenergic terminals. The efficacy of the lesion was evaluated using a rotometer (see below): rats showing more than 200 net contraversive turns per hour after injection of 0.05 mg/kg of apomorphine (i.e., maximally lesioned rats; Hudson et al. 1993) were used, and the lesion confirmed by subsequent immunohistochemistry (see below).

Behavioral testing

Locomotor activity was automatically monitorized with the aid of a Videomex-X motion analyzer (Columbus Instruments, U.S.A.), a video-based apparatus which monitors the video image in real time (30 frames per second) for each of up to six animals. The DTPM (distance traveled and pattern of movements) software option was used. Drug-induced rotation was tested in a bank of eight automated rotometer bowls (Rota-count 8, Columbus Instruments, U.S.A.), which monitor full (360°) body turns in either direction. For each rat, the net rotation asymmetry score was calculated by subtracting the total number of full turns to the left (i.e., contralateral to the lesion) from the total number of full turns to the right (i.e., ipsilateral to the lesion) over the test period.

The rats were acclimatized to the Videomex or the Rota-count for at least 15 min before drug treatment, and locomotor activity or turning behavior was monitored for 90 min after injection of D-amphetamine (or saline, group-E rats). The results are presented as means±SEM, and statistical differences were tested using three-way ANOVA (lesion×Prazosin×amphetamine dose) followed by post hoc Tukey tests (P<0.05).

Immunohistochemistry

After behavioral testing, the animals were killed by chloral hydrate overdose and then perfused first with 0.9% saline and then with cold 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. The brains were removed and subsequently washed and cryoprotected in the same buffer containing 20% sucrose, and finally cut into 40 μm sections on a freezing microtome.

Series of free-floating sections were processed for Fos, 5-HT and tyrosine hydroxylase (TH) immunohistochemistry as follows. The sections were incubated for 1 h in 10% normal serum (rabbit serum for Fos and swine serum for 5-HT and TH) with 0.25% Triton-X-100 in 0.02 M potassium phosphate-buffered saline containing 1% bovine serum albumin (KPBS-BSA), and then incubated overnight at room temperature with the corresponding primary antibody: sheep polyclonal anti-Fos antiserum directed against the N′-terminal region of the c-Fos protein molecule (1:1000 in KPBS-BSA containing 1% normal rabbit serum, 0.25% Triton-X-100 and 0.1% sodium azide; antiserum OA-11-824A, Genosys Biotechnologies, UK), rabbit polyclonal antiserum to 5-HT (1:1500 in KPBS-BSA containing 2% normal swine serum and 0.25% Triton-X-100; Incstar, USA), or rabbit polyclonal antiserum to TH (1:500 in KPBS-BSA containing 2% normal swine serum and 0.25% Triton-X-100; Pel-Freez, USA). The sections were subsequently incubated first for 60 min with the corresponding biotinylated secondary antibody (Vector, USA; diluted 1:200) and then for 90 min with an avidin-biotin-peroxidase complex (ABC, Vector, USA; diluted 1:100 in KPBS, containing 0.25% Triton-X-100 for Fos). Finally, the labeling was visualized by treatment with 0.04% hydrogen peroxidase and 0.05% 3-3′ diaminobenzidine (DAB, Sigma).

Quantification of Fos

Fos immunoreactive nuclei were counted with the aid of NIH-Image 1.55 image analysis software (Wayne Rasband, MIMH) on a Macintosh personal computer coupled to a videocamera (CCD-72, MTI) connected to a Nikon Optiphot 2 microscope with a 20× Nikon plan-Apo objective. The thresholding options of the image analysis program (size of the particle and optical density) were set so that only cells with unequivocally positive nuclei were counted, and the background was ignored. We measured a sample area with fixed size (0.36×0.46 mm) and the same location in all striatal sections, and no part of the sample area was excluded prior to counting. The area to be analyzed was encircled, and the Fos-positive nuclei counted automatically to obtain the number of nuclei per square millimeter. The nuclei were counted blind to the treatment of the animals. At least three sections through the central striatum and three sections through the nucleus accumbens of each rat were taken, and the density of nuclei per square millimeter in the dorsomedial striatum, dorsolateral striatum, and nucleus accumbens estimated. Means were compared by three-way ANOVA (lesion×Prazosin×Amphetamine dose) followed by post hoc Tukey tests (P<0.05).

Results

Normal (non-lesioned) rats

Normal rats (group A) injected with saline and then with AMPH (1 or 5 mg/kg) showed significantly higher locomotor activity than controls (i.e., rats injected with saline alone, group E), as shown in Fig. 1. The TH and 5-HT immunohistochemistry revealed dense striatal innervation by DA and 5-HT terminals, respectively. The Fos immunohistochemistry did not reveal detectable striatal expression of Fos in control rats (group E). However, rats injected with AMPH (1 or 5 mg/kg) showed numerous Fos-immunoreactive (Fos-ir) nuclei in the medial striatum, lateral striatum, and nucleus accumbens. The Fos-ir nuclei were usually more intensely stained in rats injected with 5 mg/kg of AMPH than in rats injected with 1 mg/kg of AMPH. The density of Fos-ir nuclei was higher in the medial striatum than in the nucleus accumbens and the lateral striatum (Figs. 2 and 3).

Fig. 1.
figure 1

Total distance traveled (meters over the 90-min session; mean±SEM) in control rats (saline+saline, group E), and in non-lesioned rats (group A), or rats lesioned with pCPA (group B) or 5,7-DHT (group C) that were pretreated with saline or Prazosin and then injected with 1 mg/kg (AMPH-1) or 5 mg/kg (AMPH-5) of amphetamine. Means with different letters (a, b or c) differ significantly. Three-way ANOVA and post-hoc Tukey test; P<0.05

Full size image
Fig. 2.
figure 2

Density of Fos-immunoreactive nuclei/mm2 (mean±SEM) in control rats (saline+saline, group E), and in non-lesioned rats (group A), or rats lesioned with pCPA (group B) or 5,7-DHT (group C) that were pretreated with saline or Prazosin and then injected with 1 mg/kg (AMPH-1) or 5 mg/kg (AMPH-5) of amphetamine. Means with different letters (a, b or c) differ significantly. Three-way ANOVA and post-hoc Tukey test; P<0.05

Full size image
Fig. 3A–D.
figure 3

Photomicrographs showing Fos immunoreactivity induced by 5 (A, B) or 1 (C, D) mg/kg of amphetamine in the dorsomedial region of the central striatum (A, C) and in the nucleus accumbens (B, D). Each dot is a Fos-positive neuron. Scale bar=200 μm. ac anterior commisure

Full size image

Rats pretreated with Prazosin (0.5 mg/kg) and then with 1 mg/kg of AMPH showed significantly lower locomotor activity than rats pretreated with saline (i.e., saline+1 mg/kg of AMPH; P<0.001). The locomotor activity of rats pretreated with Prazosin and then with 5 mg/kg of AMPH was not significantly lower than those pretreated with saline (Fig. 1). Similarly, pretreatment with Prazosin produced a significant reduction in the striatal expression of Fos induced by 1 mg/kg of AMPH. However, pretreatment with Prazosin did not produce statistically significant reduction in striatal expression of Fos in rats injected with 5 mg/kg of AMPH (Figs. 2 and 4A, C).

Fig. 4A–D.
figure 4

Striatal expression of Fos induced by 1 mg/kg of amphetamine in non-lesioned rats (A, C) and in rats lesioned with 5,7-DHT (B, D). Non-lesioned rats pretreated with Prazosin (C) showed significantly less expression of Fos than those pretreated with saline (A). Rats lesioned with 5,7-DHT (B) showed significantly less expression of Fos than the corresponding non-denervated rats (A). Pretreatment with Prazosin (D) did not produce any further reduction in striatal Fos expression in rats with serotonergic denervation. Each dot is a Fos-positive neuron. Scale bar=500 μm. cc corpus callosum

Full size image

Rats subjected to serotonergic lesion

Bilateral serotonergic denervation in groups B and C was confirmed by immunohistochemistry, and only data from rats showing striata with a total or almost total lack of 5-HT immunoreactive fibers were used. Administration of pCPA (i.e., acute serotonergic lesion) suppressed AMPH-induced increase in locomotor activity. No additional changes in AMPH-induced locomotor activity were observed in pCPA-lesioned rats that were treated with Prazosin. Administration of pCPA suppressed the striatal expression of Fos induced by 1 mg/kg of AMPH, and significantly reduced that induced by 5 mg/kg of AMPH. In rats lesioned with pCPA, pretreatment with Prazosin did not produce any further reduction in Fos-ir nuclei (Figs. 1 and 2).

Rats subjected to chronic serotonergic lesion with 5,7-DHT and then treated with amphetamine (1 or 5 mg/kg) showed significantly less AMPH-induced locomotor hyperactivity than the corresponding non-denervated rats. However, pretreatment with Prazosin did not induce any further reduction in locomotor activity (Fig. 1). Rats lesioned with 5,7-DHT showed significantly less striatal expression of Fos induced by 1 or 5 mg/kg of AMPH than the corresponding non-denervated rats. However, pretreatment with Prazosin did not produce any further reduction in AMPH-induced striatal Fos expression in rats with serotonergic denervation (Figs. 2 and 4A, B, D). The results showed statistically significant interaction between the presence/absence of serotonergic innervation and Prazosin effects, in both the striatal expression of Fos (F (2,47) =7.469, P=0.002) and the locomotor activity (F (2,52) =5.522, P=0.007). Therefore, Prazosin significantly reduced the expression of Fos and the locomotor activity, but only in the presence of serotonergic innervation.

Rats subjected to unilateral dopaminergic denervation

Rats subjected to maximal DA denervation of the right striatum showed intense rotational behavior towards the denervated side after injection of 5 mg/kg (799±65 net turns) or 1 mg/kg of AMPH (489±71 net turns). Pretreatment with Prazosin induced a significant reduction in the rotation induced by 5 or 1 mg/kg of AMPH (416±73 and 196±70 net turns, respectively).

The TH-immunohistochemistry revealed the absence of TH immunoreactivity (i.e., nigrostriatal DA terminals) in the right striatum. The AMPH-induced Fos expression was not detectable in the DA denervated striatum. Furthermore, the expression of Fos in the contralateral striatum was significantly lower than in striata of normal (i.e., non-denervated) rats (P<0.001 in all striatal regions), and was further reduced after pretreatment with Prazosin. The striatal expression of Fos elicited by 1 mg/kg of AMPH was 98±24 nuclei/mm2 in the medial striatum, 8±2 in the lateral striatum, and 61±6 in the nucleus accumbens, and was significantly lower in rats pretreated with Prazosin (40±7, 3±1, and 21±6, respectively).

Discussion

In the present study, administration of AMPH induced an increase in locomotor activity and the expression of Fos in the striatal neurons. Such effects have previously been attributed to AMPH-induced increase in synaptic concentration of DA, due to its increased release and decreased reuptake in the striatal terminals (Dubocovich and Zahniser 1985; Uretsky and Snodgrass 1977). Accordingly, in the present and previous experiments (Cenci et al. 1992; Labandeira-García et al. 1996; López-Martin et al. 1999) AMPH did not induce detectable expression of Fos in the DA denervated striatum, and elicited intense rotation towards the denervated side. However, there are several reports of discrepancy between striatal DA levels and behavioral events, and it has also been observed that locomotor hyperactivity induced by systemic injection of a low dose of AMPH is associated with modest increases in striatal DA levels, while local striatal perfusions of DA or AMPH do not activate any behavioral responses (Darracq et al. 1998). Similarly, DA agonists do not have AMPH-like effects (Eilam et al. 1991; Williams and Woolverton 1990). This suggests that striatal DA release is necessary but by itself not sufficient to produce the characteristic AMPH-induced responses. Our results show that the NA and 5-HT systems play an important role in the responses induced by AMPH in the striatal neurons.

Pretreatment with the α1-adrenergic-receptor antagonist Prazosin reduced the motor responses (i.e., locomotor activity and rotational behavior) and striatal expression of Fos induced by AMPH (1 mg/kg), indicating that the NA system plays an important role in these responses. These results are supported by several neurochemical studies which have shown that AMPH alters the activity of NA neurons and increases the extracellular NA levels in different regions of the brain (Archer et al. 1986; Florin et al. 1994; Ramirez and Wang 1986). Pretreatment with Prazosin led to lower responses (i.e., locomotor hyperactivity and striatal Fos expression) to 5 mg/kg of AMPH, although the reduction in the response was not statistically significant. It is possible that additional doses of Prazosin are required to obtain a significant reduction in the response, and it is also possible that high doses of AMPH may elicit additional mechanisms that compensate for the effects of Prazosin administration.

As there is no significant NA innervation of the striatum, the effects of the NA system on the striatal neurons must be mediated by one (or more) of the main striatal afferent systems: the serotonergic system, the corticostriatal glutamatergic system, and the DA system. The present behavioral and immunohistochemical data show that the 5-HT system plays a significant role in the responses induced by AMPH in the striatal neurons. Acute lesion of the 5-HT system by pCPA suppressed the locomotor response and striatal Fos expression induced by 1 mg/kg of AMPH and reduced the responses induced by 5 mg/kg of AMPH. Similarly, chronic lesion of the 5-HT system with 5,7-DHT reduced the locomotor hyperactivity and striatal Fos expression induced by either 1 or 5 mg/kg of AMPH. There was a more intense effect of the acute 5-HT lesion (i.e., pCPA lesion), which is consistent with the results of previous studies on interactions of the three main afferent striatal systems. It was observed that following chronic lesion of one of these afferent systems, other systems develop compensatory changes that reduce the effects of the lesion on the striatal function (Guerra et al. 1998; Lindefors and Ungerstedt 1990; Liste et al. 1995, 1997; Wüllner et al. 1994). It is particularly interesting to remark that in the present study, administration of Prazosin to rats with 5-HT denervation did not produce any further reduction of the AMPH-induced striatal Fos expression or locomotor hyperactivity compared with that observed in rats with 5-HT denervation only. This suggests not only that the NA system plays an important role in the AMPH-induced responses, but also that the 5-HT system is necessary for mediating the effects of AMPH-induced NA stimulation on the striatal neurons. Similarly, it has been suggested that an interaction between the NA system and the prefronto-cortical glutamatergic neurons to regulate the locomotor activated effects of AMPH (Darracq et al. 1998). It has been observed that AMPH induces release of NA in the prefrontal cortex (Florin et al. 1994) and that stimulation of cortical α1-adrenergic receptors is necessary for the locomotor activating effects of AMPH, which might either modify glutamic acid release in the striatum or act indirectly through the midbrain DA neurons to increase DA release (Blanc et al. 1994; Darracq et al. 1998).

Since NA (Grenhoff and Svensson 1993; Lategan et al. 1990; Mavridis et al. 1991), 5-HT (Blandina et al. 1989; De Deurwaerdere et al. 1996), and cortical glutamatergic (Wheeler et al. 1995) neurons, appear to exert excitatory influence on DA release, AMPH-induced changes in locomotor activity and striatal expression of Fos may be attributed to an increase in DA release, and the effect of Prazosin and/or 5-HT lesions may be merely a consequence of a reduction of DA release in striatal DA terminals. However, as indicated above, the striatal DA increase is necessary but by itself not sufficient to induce the characteristic responses to AMPH, which suggests that the effects of Prazosin and/or 5-HT lesions observed in the present study are a consequence of changes in 5-HT release by striatal 5-HT terminals. The interaction between the NA and 5-HT systems to regulate AMPH-induced responses in the striatal neurons is supported by previous studies showing that the firing of dorsal raphe 5-HT neurons and the release of 5-HT is subjected to NA facilitatory influence mediated by α1-adrenergic receptors (Baraban and Aghajanian 1980; Hjorth et al. 1995), and that administration of Prazosin decreased 5-HT levels in the striatum (Rouquier et al. 1994). In addition, stimulation of 5-HT receptors induces Fos expression in striatal neurons (Bhat and Baraban 1993; Guerra et al. 1998). Interestingly, we have recently observed similar effects (i.e., a similar reduction in AMPH-induced striatal Fos expression after Prazosin administration or 5-HT lesion) in DA denervated striata (i.e., subjected to maximal 6-OHDA lesions) grafted with fetal DA neurons (Muñoz et al. 2003). Since there are no DA neurons in the midbrain, the effects are necessarily mediated by 5-HT terminals within the striatum, which interact with the DA terminals of the grafted neurons (see also Mounir et al. 1994; Pierret et al. 1998) to regulate the AMPH-induced responses in the striatal neurons.

In conclusion, the present results suggest that striatal DA release is essential to elicit characteristic AMPH-induced responses. However, the involvement of the NA and 5-HT systems (and possibly the corticostriatal system; Darracq et al. 1998) in these responses is also important. Furthermore, the 5-HT system is necessary for mediating the effects of AMPH-induced NA stimulation on the striatal neurons. Concurrent stimulation of DA, 5-HT and glutamatergic receptors may be necessary to regulate AMPH-induced responses in the striatal neurons.