Bupropion is an antidepressant, whose efficacy has been suggested in animals (Cooper et al., 1980; Steru et al., 1987; Martin et al., 1990) and demonstrated in humans (Fabre and McLendon, 1978; Zung, 1983). In clinical studies, its antidepressant efficacy appears similar to that of selective serotonin uptake inhibitors or tricyclic antidepressants (Chouinard, 1983; Kavoussi et al., 1997). Moreover, bupropion has been more recently proposed as an aid to smoking cessation (Hurt et al., 1997).

The mechanism of its antidepressant action seems to depend on inhibition of noradrenaline uptake displayed by its major metabolite, 4-hydroxybupropion (Martin et al., 1990; Ascher et al., 1995), although the involvement of an increase in the limbic dopaminergic transmission cannot be discounted, as the drug also exhibits inhibition of dopamine (DA) uptake (Cooper et al., 1980; Nomikos et al., 1992; Learned-Coughlin et al., 2003). Furthermore, it has been shown that bupropion is without any effect on 5-hydroxytryptaminergic transmission (Ascher et al., 1995) and has no significant affinity for various types of receptors such as α or β adrenoceptors, 5-HT, DA or acetylcholine receptors (Ferris and Beamman, 1983).

In vitro, bupropion is a weak inhibitor of DA uptake as shown in synaptosomes (Sanchez and Hyttel, 1999), but it does not release [³H]DA via human dopamine transporters (DATs) transfected into COS7-cells (Eshleman et al., 1994). There are a few indications that in vivo, the effects of bupropion differ from those of dexamphetamine, such as the prevention by tetrodotoxin of the increase in extracellular DA (using microdialysis), following bupropion (Nomikos et al., 1989) but not that following dexamphetamine (Westerink et al., 1987b). However, bupropion is chemically related to dexamphetamine, which is well known to release DA. Like dexamphetamine, bupropion stimulates locomotor activity in rodents (Soroko et al., 1977) and reduces eating (Zarrindast and Hosseini-Nia, 1988). It produces only mild stereotypies in rats (Nomikos et al., 1989; Zarrindast et al., 1996). Thus, bupropion could exhibit, in vivo, DA-releasing effects and/or only inhibitory effects on DA uptake. Such a distinction is not insignificant in terms of safety for patients as amphetamine-like drugs are known to develop addictive and toxic effects, more marked than those developed by DA uptake inhibitors.

The aim of this study was to determine in vivo whether, in terms of DA transmission, bupropion exhibited, like dexamphetamine, DA-releasing properties or whether it is only a DA uptake inhibitor. For this purpose we compared in vivo, bupropion and consequently its metabolites with dexamphetamine, in different behavioural tests in which bupropion had not yet been tested or on tests performed in different conditions as these already described. Thus, we studied their effects on locomotion in animals pretreated with various agents modifying dopaminergic transmission such as reserpine, sodium hydroxy-4-butyrate and haloperidol. Bupropion and dexamphetamine were also compared in neurochemical experiments such as microdialysis and the determination of 3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA) contents in the striatum and nucleus accumbens of haloperidol-pretreated rats. The microdialysis experiments were carried out in the striatum of chloral hydrate anaesthetized and acutely implanted rats. Although far from animal physiology, this protocol was chosen to decrease the firing rate of DA neurons (Hamilton et al., 1992), allowing a better discrimination between a DA-releasing-effect independent from the electric activity of DA neurons, characteristic of amphetamine-like drugs, and a DA uptake inhibitory action, whose effect on extracellular DA concentration strongly depends on this electric activity.

We conclude that bupropion, although chemically related to dexamphetamine, a well-established releaser of DA, behaved in vivo only as a DA uptake inhibitor and was devoid of DA-releasing effects.

Figure 1 Effect of increasing doses of dexamphetamine or bupropion on locomotor activity. Mice were injected s.c. with saline (0) or increasing doses of dexamphetamine (a) or bupropion (b), 10min before their introduction into the actimeter. The horizontal (more ...)

Figure 2 Effect of dexamphetamine and bupropion on the akinesia induced by sodium hydroxy-4-butyrate (Gamma-OH). Mice were pretreated i.p. with saline or sodium hydroxy-4-butyrate (375mgkg−1) 10min before solvent (s.c.), dexamphetamine (more ...)

Figure 3 Effect of haloperidol on dexamphetamine- or bupropion-induced motor stimulant effect. Mice were pretreated i.p. with saline or haloperidol (50, 100 or 200μgkg−1); 20min later, mice were treated s.c. with dexamphetamine (more ...)

Figure 4 Effect of bupropion or dexamphetamine or their association on reserpine-induced akinesia. Mice were pretreated s.c. with reserpine 4mgkg−1, 5h before s.c. administration of saline or bupropion (5, 10 or 20mgkg (more ...)

Figure 5 Effect of dexamphetamine, GBR12783 or bupropion on extracellular DA concentration in the striatum of chloral hydrate-anesthetized rats. A microdialysis probe inserted into the right striatum and perfused with Ringer's solution at a flow rate of 1 (more ...)

In the striatum, bupropion (10, 20, 40 and 80mgkg⁻¹, s.c.) did not modify the DOPAC/DA or the HVA/DA ratios, when given alone (Figure 6a). Given to haloperidol pretreated animals, bupropion potentiated dose-dependently the effects of haloperidol on these two ratios (Figure 6a). Dexamphetamine (1, 2.5 and 7.5mgkg⁻¹, s.c.) given alone also did not modify the DOPAC/DA and HVA/DA ratios. However, in contrast to bupropion, it reduced the increase of these ratios elicited by haloperidol (Figure 6b).

Figure 6 Effect of dexamphetamine and bupropion on the DOPAC/DA and the HVA/DA ratios in striatum and nucleus accumbens of haloperidol-pretreated rats. Rats were pretreated i.p. with saline (white columns) or haloperidol (450μgkg−1 (more ...)

In the nucleus accumbens, neither bupropion nor dexamphetamine modified, by themselves, the DOPAC/DA and the HVA/DA ratios. In this structure, bupropion from the dose of 40mgkg⁻¹, as in the striatum, potentiated the haloperidol increasing effect on the DOPAC/DA ratio and at the higher dose of 80mgkg⁻¹, on the HVA/DA ratio (Figure 6c). Dexamphetamine still opposed the increasing effect of haloperidol but, in this tissue, only for the DOPAC/DA ratio (Figure 6d).

The locomotor stimulant effect of bupropion has already been described in rodents (Soroko et al., 1977). In the present experiments, we observed that this effect of bupropion in mice, like that of dexamphetamine, was biphasic. Thus, the maximal stimulation of horizontal locomotion was obtained at 40mgkg⁻¹, whereas it disappeared at 80mgkg⁻¹. As high doses of bupropion and dexamphetamine had been shown to produce stereotypies in rats (Nomikos et al., 1989), these stereotypies might oppose the stimulant locomotor effect of both drugs at the highest tested dose. The locomotor stimulant effect of bupropion (40mgkg⁻¹) was clearly lower than that of dexamphetamine (3mgkg⁻¹). This difference may be compared with the increase in the extracellular DA concentration obtained in our microdialysis experiments. In a microdialysis study performed in the striatum of conscious rats, Nomikos et al. (1989) observed that bupropion, at doses ranging from 10 to 100mgkg⁻¹, increased the extracellular concentration of DA in a dose- and time-dependent manner, with a maximal increase of 543%. Because of this strong increase and to make a clear distinction between a possible DA-releasing effect (typical of amphetamine-like drugs, independent of neuronal firing rate) and a potential DA uptake inhibitory effect of bupropion, we chose to carry out our microdialysis experiments in rats anaesthetized with chloral hydrate and acutely implanted, in order to compare bupropion and dexamphetamine. Indeed, we observed that the increase in extracellular DA concentration obtained in the striatum of anaesthetized rats by a low dose of dexamphetamine (1mgkg⁻¹) was by far more marked than that induced either by the pure DA uptake inhibitor GBR12783 (5–15mgkg⁻¹) or by bupropion, even at the dose of 100mgkg⁻¹. This difference results from the chloral hydrate anaesthesia administered to our rats as chloral hydrate reduces the firing rate of dopaminergic nigro-striatal neurons and, consequently, the DA release by striatal terminals (Hamilton et al., 1992). This DA released from neurons is normally removed from the extracellular space by the DA uptake system. Bupropion, by inhibiting DA uptake, leads to an accumulation of DA, which is measured as an increase in extracellular DA concentration. This effect is diminished when DA release is decreased, consequent to the reduction by chloral hydrate of the firing rate of nigro-striatal dopaminergic. This explanation is compatible with the results of Nomikos et al. (1989, 1990), who found that the increase in extracellular DA concentration elicited by bupropion in the striatum of conscious rats was blocked by tetrodotoxin, establishing that this effect was dependent on action potentials, that is, on neuronal activity. In this way, our results, as well as those obtained by Nomikos et al. (1989, 1990), clearly demonstrate that bupropion and dexamphetamine affect DA transmission differently, as the effect of dexamphetamine, which releases DA independently of the neuronal firing rate, remained elevated in spite of the chloral hydrate anaesthesia (this study) and was unaffected by tetrodotoxin (Westerink et al., 1987b; Nomikos et al., 1990).

The akinesia induced by sodium hydroxy-4-butyrate is due to the inhibition of the tonic electric activity of nigro-striatal and meso-accumbic dopaminergic neurons (Engberg and Nissbrandt, 1993; Erhardt et al., 1998). Amphetamine-like agents that release DA independently of the firing rate of these neurons reverse this akinesia (Duterte-Boucher and Costentin, 1991). In the present study, bupropion, in contrast to dexamphetamine, failed to reverse this akinesia, indicating a lack of DA-releasing effect. This is also corroborated by the inability of bupropion to reverse the reserpine-induced akinesia, whereas dexamphetamine was effective in this respect. As a matter of fact, reserpine inhibits the vesicular monoamine transporter type 2, preventing the vesicular storage of monoamines (Erickson et al., 1992). Reserpine-induced akinesia results from a suppression of dopaminergic transmission, as dopaminergic neurons are no longer able to release DA. Therefore, under these conditions, the DA uptake inhibitors are unable to increase the extra-cellular, synaptic concentration of the amine. On the other hand, amphetamine-like compounds, by releasing the newly synthesized cytosolic pool of DA, restore the stimulation of post-synaptic dopaminergic receptors and thereby locomotor activity. Moreover, we observed that bupropion, co-administered with dexamphetamine, inhibits the dexamphetamine-induced reversion of the akinesia elicited by reserpine. This may be explained by the fact that bupropion acts by blocking the DAT, preventing both neuronal uptake of amphetamine and the subsequent release of DA mediated by the transporters acting in a reverse manner (Garcia de Mateos-Verchere et al., 1998; Zaczek et al., 1991; Simon et al., 1995; Jones et al., 1998a).

Haloperidol did not prevent the stimulant effect of bupropion on motor activity. We have previously observed such a lack of antagonism when haloperidol was used with the DA uptake inhibitors GBR12783 (Duterte-Boucher et al., 1990) or oxolinic acid (Garcia de Mateos-Verchere et al., 1998). The blockade by haloperidol of DA D2 autoreceptors in somato-dendritic locations increases the firing rate of nigro-striatal and meso-accumbic dopaminergic neurons (Bunney et al., 1973), and thereby enhances in terminals of DA neurons, DA synthesis and synaptic release from the vesicular stores as well as increasing DA turnover (Anden et al., 1971) and thus production of DOPAC and HVA (Westerink and Korf, 1975; Shore, 1976; Kuczenski, 1980). Consequently, blockade of DAT by bupropion and the simultaneous increase of DA release induced by haloperidol, dramatically increased the extracellular DA concentration. At the level of post-synaptic DA receptors, these very high concentrations of DA could compete successfully with haloperidol, thereby maintaining the excito-locomotor effect. On the contrary, the dexamphetamine-stimulant effect was easily reversed by haloperidol, as the DA release induced by dexamphetamine is independent of presynaptic regulation of the neuronal activity by DA autoreceptors. Thus, as dexamphetamine induces release of a definite quantity of DA, blockade of post-synaptic DA receptors by haloperidol dose-dependently reduces its stimulant locomotor effect.

We observed that haloperidol increased the DOPAC/DA ratio, as well as the HVA/DA ratio, both in the striatum and the nucleus accumbens of rats. Interestingly, when bupropion was administered in haloperidol-pretreated rats, we found that the increase in DOPAC/DA and HVA/DA was much greater than in animals treated with haloperidol alone. Similar results were obtained with the selective DA uptake inhibitors amfonelic acid and GBR12783 (Miller and Shore, 1982; Boulay et al., 1996). Westerink et al. (1987a) explain this synergy by an impulse-dependent mechanism. The blockade of DAT by a DA uptake inhibitor might induce a compensatory increase in DA synthesis and thereby an increase in DOPAC formation by intraneuronal monoamine oxidase (MAO). This hypothesis is supported by recent studies performed in mice lacking the DAT, which show a twofold increase in DA synthesis as compared with control mice (Jones et al., 1998b; Benoit-Marand et al., 2000). Under normal conditions, this increase in DA and DOPAC formation is counteracted by the activation of D2 autoreceptors, stimulated by the increased DA concentration in the synaptic cleft, subsequent to the inhibition of DA uptake. The co-administration of the D2 DA receptor antagonist haloperidol, with a DA uptake inhibitor, should strongly stimulate DA synthesis and DOPAC formation. Alternatively, it may be also suggested that the high concentration of DA, established in the synaptic cleft by combining haloperidol and a DA uptake inhibitor, could participate in the increased DOPAC formation in the surrounding astroglial cells. In these cells, provided with MAO and catechol-O-methyl transferase (COMT) activities (Naudon et al., 1992), DA uptake is carried out by the so-called ‘extraneuronal monoamine transporter' (Wu et al., 1998) which displays a very different pharmacological profile of inhibition as, for instance, it is insensitive to either cocaine (Eisenhofer, 2001) or the selective and potent neuronal DA uptake inhibitor GBR12935 (Takeda et al., 2002). Finally, this high DOPAC level leads to an increase in HVA formation, as HVA is formed predominantly from DOPAC by COMT (Westerink, 1979). On the other hand, when dexamphetamine was administered in haloperidol-pretreated rats, a partial reversal of the haloperidol-induced increase of DA metabolism was observed. Four effects of dexamphetamine could explain these results: (i) DA release from nigral dendritic varicosities, which competes with haloperidol on somatodendritic dopaminergic autoreceptors, decreasing the activation of the firing rate of dopaminergic neurons operated by haloperidol; (ii) blockade of the neuronal DAT, which prevents the reuptake of released DA into the neuron and protects the amine from inactivation by the intra-neuronal MAO; (iii) blockade of the extraneuronal monoamine transporter preventing the DA uptake into astroglial cells and its metabolism by astroglial MAO and COMT (Wu et al., 1998); and (iv) a direct inhibition of MAO (Green and El Hait, 1978). Taken together, these effects lead to a decrease in DOPAC/DA ratio and in HVA/DA ratio, relative to those observed in response to haloperidol alone. In this respect, bupropion and dexamphetamine act in exactly opposite ways.

In conclusion, all our experiments were performed in vivo and thereby involved both bupropion and its major metabolites. They indicate, in a coherent manner, whatever the variables measured, that, as reported previously in vitro by other authors, bupropion behaved as a DA uptake inhibitor and not like the DA releaser, dexamphetamine. Such a difference could have consequences for the toxic and addictive properties of this drug, as these properties are known to be more marked in the amphetamines.

We thank Dr Nathalie Dourmap, Elisabeth Chapelot and Anne Marie Cottard for their assistance with microdialysis and HPLC procedures.

Conflict of interest

The authors state no conflict of interest.