0 25 50 100 Corticosterone ((jg/mL)
Fig. 2. Compared with controls (sham-operated animals), suppression of corticosterone by adrenalectomy (ADX) decreases responding for cocaine during a self-administration task. The effects of ADX are reversed dose-dependently by corticosterone replacement (corticosterone added to the drinking solution of the animals; ADX + Cort). The response to cocaine is fully restored when animals receive a replacement treatment reproducing stress levels of corticosterone. Modified from Deroche et al. (1997b).
reduces motivation to self-administer cocaine (Deroche-Gamonet et al, 2003).
Similar reduction in drug responding was observed following pharmacological blockade of corticosterone secretion. Thus, treatment with the corticosterone synthesis inhibitor metyrapone reduces self-administration of cocaine, both during the acquisition and maintenance phase (Goeders and Guerin, 1996a, 1996b). Another corticosterone synthesis inhibitor, ketoconazole, has similar effects and reduces acquisition of self-administration when it decreases circulating levels of glucocorticoids (Goeders et al, 1998). These effects are not due to nonspecific decreases in motor behavior or motivation, as these treatments do not modify seeking behavior in food-related tasks (Piazza et al, 1994). As mentioned previously, however, a note of caution should be employed when analyzing the effects of corticosterone synthesis inhibitors such as metyrapone or ketoconazole; these drugs could have nonspecific effects and should be used mostly to confirm studies that used other methods of manipulating the HPA axis.
It is noteworthy to mention that, similar to the locomotor effects of psychostimulant drugs, drug-induced increases in hormone levels are not a critical factor for intravenous drug self-administration. Thus, self-administration of psychostimulants dramatically increases glucocorticoid secretion (Baumann et al, 1995; Broadbear et al, 1999a, 1999b; Mantsch et al, 2000; Galici et al, 2000; Goeders, 2002), but blockade of this increase does not modify responding for cocaine (Deroche et al, 1997b; Broadbear et al, 1999c). In addition, animals have been shown to self-administer cocaine at doses that do not modify circulating levels of glucocorticoid hormones (Broadbear et al, 1999b), which further corroborates the notion that drug-induced glucocorticoid secretion is not important to maintain drug self-administration behavior.
The fact that drug-induced glucocorticoid secretion is irrelevant for the maintenance of drug self-administration could explain the apparent discrepancies between studies on rats, showing that blockade of corticosterone secretion reduces psychostimulant self-administration, and those on humans and on non-human primates, showing that blockade of glucocorticoid secretion has no effects on cocaine self-administration (Broadbear et al, 1999c), or on the subjective effects of smoked cocaine (Ward et al, 1998, 1999). Indeed, in rodent studies, glucocorticoid levels are always reduced to values that are well below those seen in control conditions. Instead, in primate studies, glucocorticoid levels are never brought below those observed in control subjects: the pharmacological treatments are simply aimed in preventing drug-induced increases in hormone levels, which, as mentioned above, are irrelevant for self-administration behavior. These are important considerations suggesting that differences in the literature might be perceived erroneously as inter-species differences, whereas they simply reflect differences in the attained glucocorticoid levels: blockade of drug-induced hormone secretion has no effects on drug responding, whereas blockade of basal or stress-like levels decreases drug self-administration.
Data showing that suppression of corticosterone decreases self-administration behavior are in contrast with studies showing that adrenalectomy does not modify cocaine-induced place preference (CPP) (Suzuki et al, 1995), a response that is often considered as an index of the rewarding effects of drugs of abuse (Carr et al, 1989; Hoffman, 1989;
Tzschentke, 1998). It is possible that this discrepancy could depend on the fact that the two behaviors seem to depend on different neuronal substrates. Thus, the dopamine system largely mediates psychostimulant self-administration (Roberts et al, 1980) but it does not mediate the effects if CPP induced by cocaine injected intreaperitoneally (Spyraki et al, 1982). Interestingly, CPP induced by intravenous administration of cocaine is dopamine-dependent (Spyraki et al, 1982), but, to our knowledge, the effects of glucocorticoids on this behavior have not been studied. Another possible explanation for the discrepancy between the role of glucocorticoid in CPP versus self-administration is that CPP might not be the most suitable paradigm to study changes in the intensity of the reinforcing or rewarding effect of drugs (for review, see Bardo and Bevins, 2000). In fact, this test can evaluate changes in the threshold dose of psychostimulants required to produce conditioning, but once the response is induced, the intensity of its effects does not change significantly as a function of drug dose (Costello et al, 1989). Therefore, this paradigm is mostly suited to evaluate horizontal shifts in dose-response functions, but not vertical ones. On the other hand, drug self-administration shows dose-dependent responding, thus allowing analysis of both horizontal and vertical shifts in dose-response functions. Given the fact that adrenalectomy induces a vertical shift in the dose-response to cocaine self-administration (Deroche et al, 1997b), it is understandable that this manipulation has no effects of cocaine-induced place conditioning.
Concerning the role of stress levels of glucocorticoids on self-administration, it has been shown that stress enhances self-administration behavior (for review see Kreek and Koob, 1998; Lu and Shaham, 2004). Again, different stressors such as tail pinch (Piazza et al, 1990a), foot-shock (Goeders and Guerin, 1994), social isolation (Schenk et al, 1987a, b), social stress (Haney et al, 1995; Miczek and Mutschler, 1996; Tidey and Miczek, 1997) and food restriction (Carroll et al, 1979; Papasava and Singer, 1985; Papasava et al, 1986; Macenski and Meisch, 1999; Marinelli et al, 2002) all increase intravenous self-administration of amphetamine and cocaine. These effects have been observed for different doses of the drugs, during the acquisition phase, the retention one, as well as in progressive ratio schedules. As in the case of locomotor activity, the increase in self-administration induced by stress seems to depend on stress-induced corticosterone secretion. The effects of stress levels of corticosterone have only been examined following treatment with ketoconazole, a corticosterone synthesis inhibitor. Repeated treatment with ketoconazole decreases the rate of acquisition of cocaine self-administration as well as the proportion of rats meeting acquisition criterion following food restriction stress (Campbell and Carroll, 2001). The effects of corticosterone reduction are not related to changes in motivation or motor behavior, because operant responding for food is not decreased in groups whose corticosterone levels have been modified (Micco et al, 1979; Piazza et al, 1994). As mentioned previously, studies on nonspecific corticosterone synthesis inhibitors should be viewed cautiously, as their effects could also depend on other non-specific properties of these drugs.
The effects of stress levels of glucocorticoids on drug self-administration have also been analyzed by studying the effects of repeated administration of high, stress-like levels of glucocorticoids on drug responding. These studies have shown that repeated exposure to high levels of glucocorticoids reproduces the effects of stress on drug responding. Thus, rats repeatedly treated with corticosterone have been shown to acquire cocaine self-administration at a lower dose compared with vehicle-treated controls (Mantsch et al, 1998), and the intravenous injection of corticosterone prior to a self-administration session also increases drug responding in animals that would not readily acquire cocaine self-administration (Piazza et al, 1991a).
Again, as in the case of locomotor activity, it is very important to specify that most studies on drug self-administration have observed increases in drug responding following repeated or prolonged exposure to high, stress-like levels of glucocorticoids (Goeders and Guerin, 1996b; Deroche et al, 1997b; Mantsch et al, 1998), although see (Piazza et al, 1991a). It is therefore possible that long-term exposure to high levels of these hormones is required for the development of stress-induced increase in drug responding, whereas acute increase in glucocorticoids might not be sufficient. Indeed, a single exposure to high levels of glucocorticoids does not modify the subjective responses to amphetamine in humans (Wachtel et al., 2001).
Was this article helpful?