Tolerance and sensitization to the behavioral effects of cocaine in rats: relationship to benzodiazepine receptors

Role of GABA-active neurosteroids in the efficacy of metyrapone against cocaine addiction

Previous research has demonstrated a complicated role for stress and HPA axis activation in potentiating various cocaine-related behaviors in preclinical models of drug dependence. However, the investigation of several antiglucocorticoid therapies has yielded equivocal results in reducing cocaine-related behaviors, possibly because of varying mechanisms of actions. Specifically, research suggests that metyrapone (a corticosterone synthesis inhibitor) may reduce cocaine self-administration in rats via a nongenomic, extra-adrenal mechanism without altering plasma corticosterone. In the current experiments, male rats were trained to self-administer cocaine infusions and food pellets in a multiple, alternating schedule of reinforcement. Metyrapone pretreatment dose-dependently decreased cocaine self-administration as demonstrated previously. Pharmacological inhibition of neurosteroid production by finasteride had significant effects on cocaine self-administration, regardless of metyrapone pretreatment. However, metyrapone's effects on cocaine self-administration were significantly attenuated with bicuculline pretreatment, suggesting a role for GABA-active neurosteroids in cocaine-reinforced behaviors. In vitro binding data also confirmed that metyrapone does not selectively bind to GABA-related proteins. The results of these experiments support the hypothesis that metyrapone may increase neurosteroidogenesis to produce effects on cocaine-related behaviors.

The extra-adrenal effects of metyrapone and oxazepam on ongoing cocaine self-administration

Investigation of the role of stress in cocaine addiction has yielded an efficacious combination of metyrapone and oxazepam, hypothesized to decrease relapse to cocaine use by reducing stress-induced craving. However, recent data suggest an extra-adrenal role for metyrapone in mediating stress- and addiction-related behaviors. The interactions between the physiological stress response and cocaine self-administration were characterized in rodents utilizing surgical adrenalectomy and pharmacological treatment. Male Wistar rats were trained to self-administer cocaine (0.25mg/kg/infusion) and food pellets under a concurrent alternating fixed-ratio schedule of reinforcement. Surgical removal of the adrenal glands resulted in a significant decrease in plasma corticosterone and a consequent increase in ACTH, as expected. However, adrenalectomy did not significantly affect ongoing cocaine self-administration. Pretreatment with metyrapone, oxazepam and their combinations in intact rats resulted in a significant decrease in cocaine-reinforced responses. These same pharmacological treatments were still effective in reducing cocaine- and food-reinforced responding in adrenalectomized rats. The results of these experiments demonstrate that adrenally-derived steroids are not necessary to maintain cocaine-reinforced responding in cocaine-experienced rats. These results also demonstrate that metyrapone may produce effects outside of the adrenal gland, presumably in the central nervous system, to affect cocaine-related behaviors.

The combination of metyrapone and oxazepam for the treatment of cocaine and other drug addictions

Although scientists have been investigating the neurobiology of psychomotor stimulant reward for many decades, there is still no FDA-approved treatment for cocaine or methamphetamine abuse. Research in our laboratory has focused on the relationship between stress, the subsequent activation of the hypothalamic-pituitary-adrenal (HPA) axis, and psychomotor stimulant reinforcement for almost 30 years. This research has led to the development of a combination of low doses of the cortisol synthesis inhibitor, metyrapone, and the benzodiazepine, oxazepam, as a potential pharmacological treatment for cocaine and other substance use disorders. In fact, we have conducted a pilot clinical trial that demonstrated that this combination can reduce cocaine craving and cocaine use. Our initial hypothesis underlying this effect was that the combination of metyrapone and oxazepam reduced cocaine seeking and taking by decreasing activity within the HPA axis. Even so, doses of the metyrapone and oxazepam combination that consistently reduced cocaine taking and seeking did not reliably alter plasma corticosterone (or cortisol in the pilot clinical trial). Furthermore, subsequent research has demonstrated that this drug combination is effective in adrenalectomized rats, suggesting that these effects must be mediated above the level of the adrenal gland. Our evolving hypothesis is that the combination of metyrapone and oxazepam produces its effects by increasing the levels of neuroactive steroids, most notably tetrahydrodeoxycorticosterone, in the medial prefrontal cortex and amygdala. Additional research will be necessary to confirm this hypothesis and may lead to the development of improved and specific pharmacotherapies for the treatment of psychomotor stimulant use.

The dopamine antagonist cis-flupenthixol blocks the expression of the conditioned positive but not the negative effects of cocaine in rats

Human cocaine users report that the initial "high" produced by cocaine administration is followed by an anxiogenic "crash". Given that cocaine has such robust and opposing properties, it is likely that both positive and negative effects of cocaine contribute to an individual's motivation to administer the drug. Despite this likelihood, the neurobiology underlying cocaine's dual processes remains unclear. While much literature supports a role for dopamine (DA) in cocaine reward, it is uncertain if DA also contributes to the drug's negative effects. Our laboratory has extensively utilized a modified conditioned place test to explore cocaine's opponent processes. In this paradigm rats develop conditioned place preferences (CPPs) for an environment paired with the immediate/positive effects of cocaine, and conditioned place aversions (CPAs) for an environment paired with the delayed/negative effects present 15-min after i.v. injection. In the current study rats were conditioned to associate an environment with either the immediate or delayed effects of i.v. cocaine (1mg/kg/0.1ml) 3h after i.p. pre-treatment with either the DA D1/D2 receptor antagonist cis-flupenthixol (0.5mg/kg/ml) or saline vehicle. As expected, vehicle-treated control animals developed the normal pattern of CPPs for cocaine's immediate effects or CPAs for the delayed effects of cocaine. However, while DA receptor antagonism prevented the expression of cocaine CPPs it did not alter the expression of cocaine-induced CPAs. These data confirm a role for DA transmission in cocaine reward but suggest that different neural pathways mediate the drug's negative/anxiogenic properties.

Acute and chronic effects of cocaine on the spontaneous behavior of pigeons

The present experiment examined the effects of acute and daily cocaine on spontaneous behavior patterns of pigeons. After determining the acute effects of a range of doses, 9 pigeons were divided into three groups that received one of three doses of cocaine daily, either 1.0, 3.0, or 10.0 mg/kg cocaine. Measures were taken of spontaneous locomotion, pecking, preening, and emesis. Under daily administration, cocaine induced consistent and substantial enhancements of its locomotor effects in all 9 pigeons, consistent with the phenomenon of locomotor sensitization. The maximum locomotor output did not differ according to the size of the daily dose. Locomotion was not elevated following tests of the saline vehicle, suggesting the effect was due to cocaine, not to a change in baseline or reactivity to the injection procedure. Cocaine dose-dependently decreased preening when given acutely, and those effects were not altered by repeated cocaine administration. Pecking occurred at very low rates and was unresponsive to cocaine treatment. Cocaine-induced emesis showed a dose-dependent increase under initial tests with cocaine, and those effects were attenuated following daily exposure. In a final condition, cocaine was replaced with daily saline for 30 days to assess the persistence cocaine-related increases in locomotion. Approximately half of the pigeons continued to show enhanced effects even after 30 days without cocaine, so although persistence was obtained, it showed marked intersubject variability. The data indicate that the effects of repeated cocaine administration on the behavior of pigeons shows parallels with many effects commonly reported with rodents (i.e., increased locomotion following repeated treatment, decrease in preening or grooming, persistence following drug withdrawal).

Effects of buspirone on the immediate positive and delayed negative properties of intravenous cocaine as measured in the conditioned place preference test

In prior work, we have demonstrated that the behavioral effects of cocaine adhere to the predictions of the opponent-process theory of drug action. Animals develop conditioned place preferences for distinct locations paired with the immediate effects of IV cocaine, but learn to avoid places paired with the effects present 15-min post-injection. It was of interest to assess the putative role of 5-HT in producing the negative properties of cocaine since cocaine acts to inhibit the reuptake of serotonin (5-HT) and since such actions have been associated with anxiogenic consequences. Male rats were administered a reinforcing dose of cocaine (1.0 mg/kg IV) and then placed - either immediately or after a 15-min delay - into one side of a two-compartment (black-white) conditioned place preference (CPP) box for 5-min. On alternate days, the animals received IV saline injections and were placed in the opposite side of the CPP box. This continued for eight days after which animals had experienced 4 pairings of cocaine with one side (black or white) of the CPP apparatus, and 4 saline pairings with the opposite side. Other groups of rats were treated identically except that 30-min prior to placement into the apparatus, these animals received an IP injection of saline or buspirone (a partial 5-HT1A agonist) at a dose that we have shown to be anxiolytic (2.5 mg/kg IP). Control animals experienced either buspirone or saline pretreatments without cocaine. Our results confirm that animals increase the time spent on the side paired with the immediate effects of cocaine (compared to baseline), but tend to avoid the side paired with effects present 15-min post-injection. Buspirone had no effect on the immediate rewarding properties of cocaine, but completely reversed the negative properties present 15-min post-cocaine. These results are consistent with the view that attenuation of 5-HT neurotransmission (via the autoreceptor agonist properties of buspirone) can reverse the negative impact of IV cocaine.

Similar effects of cocaine and immobilization stress on the levels of heat-shock proteins and stress-activated protein kinases in the rat hippocampus, and on swimming behaviors: the contribution of dopamine and benzodiazepine receptors

Cocaine (COC) has been reported to cause effects similar to physiological stressors in the brain neuroendocrinal system, including heat-shock protein (HSP) expression, although these effects have not been elucidated in detail. In the present study, we examined the effects of repeated (4 days) treatments with cocaine hydrochloride (35 mg/kg, i.p.) and 10 min immobilization stress (IM) on the distribution of HSP (HSP27, HSP60, HSP70, HSC70) and stress-activated protein kinase (SAPK) (SAPKalpha, SAPKbeta, SAPKgamma) immunoreactive nerve cells (positive cells) in the rat hippocampus. The swimming behaviors of the rats in the forced swimming test were also examined. In both COC and IM groups, an early enhancement (5 h time point) of hippocampal HSP (HSP27, HSP60, HSP70, HSC70) and SAPK (SAPKbeta, SAPKgamma) positive cells was observed, whereas a recovery (SAPKs) or attenuation (HSP60 and HSC70) was observed at the 24 h time point. In both groups, a depression of the swimming behaviors (attenuation in the activity counts and time until immobility) below the control level was observed at the 5 h point, but a recovery was observed at the 24 h time point. At the 48 h time point, all parameters returned to the control level. These alterations in the levels of HSPs and SAPKs, and the swimming behaviors were similar to those observed in the stress (IM) group, and were characteristic in that all of these alterations were attenuated by the benzodiazepine inverse agonist, Ro 15-4513 (5 mg/kg, i.p.), and the dopamine D1 receptor antagonist, SCH23390 (0.5 mg/kg, i.p.), which was not observed in the groups treated with another stressor-like drug (bicuculline).

Gamma-vinyl GABA, an irreversible inhibitor of GABA transaminase, alters the acquisition and expression of cocaine-induced sensitization in male rats

We examined the effect of (+/-)-gamma-vinyl GABA (GVG, Vigabatrin), an irreversible inhibitor of the enzyme GABA transaminase, on the acquisition and expression of cocaine-induced sensitization in albino male Sprague-Dawley rats. Animals received a single injection of 1 ml/kg i.p. of 0.9% saline or 15 mg/kg i.p. of (-)-cocaine and locomotor activity was assessed using automated locomotor cages and stereotyped behaviors were scored using a 4-point rating scale (Day 1). Subsequently, animals were given 15 mg/kg i.p. of cocaine every 48 h in their home cage for 1 week (Days 3, 5, and 7) and then given no treatment for 1 week. A challenge injection of 15 mg/kg i.p. of cocaine, but not vehicle, produced a significant increase in locomotor activity and stereotyped behaviors on Day 15 compared to animals that received cocaine on Day 1. Administration of 75 mg/kg i.p. of GVG 2.5 h before the cocaine injections did not significantly alter the acquisition of cocaine-induced locomotor sensitization. However, 150 mg/kg i.p. of GVG significantly attenuated the acquisition of cocaine-induced locomotor sensitization. Administration of 150 mg/kg i.p. of GVG 2.5 h before the cocaine challenge injection on Day 15 significantly attenuated the expression of cocaine-induced locomotor sensitization. Acquisition and expression of cocaine-induced sensitization of stereotypy was also significantly attenuated by 150 mg/kg i.p. of GVG. Since sensitization may be one of the factors involved in relapse to drug use, the present results, in combination with previous findings that GVG blocks the rewarding and incentive motivating effects of cocaine, suggest that GVG might prove useful in the treatment of cocaine addiction.

Different requirements for cAMP response element binding protein in positive and negative reinforcing properties of drugs of abuse

Addiction is a complex process that relies on the ability of an organism to integrate positive and negative properties of drugs of abuse. Therefore, studying the reinforcing as well as aversive components of drugs of abuse in a single model system will enable us to understand the role of final common mediators, such as cAMP response element-binding protein (CREB), in the addiction process. To this end, we analyzed mice with a mutation in the alpha and Delta isoforms of the CREB gene. Previously we have shown that CREB(alphaDelta) mutant mice in a mixed genetic background show attenuated signs of physical dependence, as measured by the classic signs of withdrawal. We have generated a uniform genetically stable F1 hybrid (129SvEv/C57BL/6) mouse line harboring the CREB mutation. We have found the functional activity of CREB in these F1 hybrid mice to be dramatically reduced compared with their wild-type littermates. These mice maintain a reduced withdrawal phenotype after chronic morphine. We are now poised to examine a number of complex behavioral phenotypes related to addiction in a well defined CREB-deficient mouse model. We demonstrate that the aversive properties of morphine are still present in CREB mutant mice despite a reduction of physical withdrawal. On the other hand, these mice do not respond to the reinforcing properties of morphine in a conditioned place preference paradigm. In contrast, CREB mutant mice demonstrate an enhanced response to the reinforcing properties of cocaine compared with their wild-type controls in both conditioned place preference and sensitization behaviors. These data may provide the first paradigm for differential vulnerability to various drugs of abuse.

Chronic cocaine differentially affects diazepam's anxiolytic and anticonvulsant actions. Relationship to GABA(A) receptor subunit expression

Benzodiazepines are used to treat the anxiety associated with cocaine withdrawal, as well as cocaine-induced seizures. Since cocaine exposure was shown to affect BZ binding density, abuse liability, subjective hypnotic actions and seizure susceptibility, we assessed whether chronic cocaine alters diazepam's anxiolytic and anticonvulsant actions. Changes in GABA(A) receptor subunit protein expression were also assessed as they may relate to BZ activity at the receptor. Male Sprague-Dawley rats were injected with cocaine-HCl (15 mg/kg, i.p.) or saline once daily for 14 days. One day after the last injection, DZP (1 mg/kg i.p.) significantly increased time spent on and entries into open arms of an elevated plus maze in both saline- and cocaine-treated groups, yet the effect was greater in cocaine-treated rats. Eight days after cessation of treatment DZP did not have a significant anxiolytic effect in either group. To assess the effect of cocaine on DZP's anticonvulsant actions, PTZ was infused at a constant rate via the lateral tail vein and clonus onset was recorded in the presence and absence of DZP (5 mg/kg, i.p). DZP significantly elevated seizure threshold in both groups of rats. Chronic cocaine also had no effect on the beta-CCM seizure threshold. Quantitative immunohistochemistry of GABA(A) receptor subunit protein demonstrated significant regulation of alpha2 (-10%) and beta3 (+9%) subunits in the hippocampal dentate gyrus and CA1 regions, respectively. Small changes in GABAR subunit expression in specific brain areas may relate to DZP's enhanced anxiolytic effectiveness whereas it's anticonvulsant actions likely remain intact following cocaine administration.

A comparison of fixed and variable doses of cocaine in producing and augmenting tolerance to its effects on schedule-controlled behavior

Twelve pigeons were trained to peck a key under a fixed-ratio 20-response schedule of food presentation. Acute effects of cocaine (03-10.0 mg/kg), determined by administering the drug once per week, revealed dose-dependent decreases in frequency of key pecking. The pigeons were then divided into six pairs, matched with respect to acute dose-response curves. One of each pair received one of five different doses before each daily session (variable-dosing condition) and the other received a fixed dose equal to the arithmetic average of the doses experienced by its pair mate (fixed-dosing condition). Following 50 days of exposure, subjects in the variable-dosing condition were then switched to the fixed-dosing condition. Dose-response functions were then determined in both groups by substituting doses for the fixed daily dose, once per week. Rate-decreasing effects were attenuated similarly in both groups of subjects, both at the end of the variable-dosing regimen and during subsequent fixed dosing. Next, an attempt was made to increase the degree of tolerance. Specifically, pigeons in the variable-dosing condition were exposed repeatedly to a range of doses in which the largest dose was 1/8 to 1/4 log unit larger than in the original variable-dosing phase. Pigeons in the fixed-dosing group were exposed daily to the largest dose that did not eliminate key pecking by the end of the initial repeated-dosing regimen. Dose effects were determined after at least 35 days of exposure. If the dose-response function had shifted to the right, the largest dose for the variable-dosing subjects was increased by 1/8 to 1/4 log unit and the smallest dose in the sequence was eliminated, and another period of variable dosing commenced. For the fixed-dosing subjects, if the curve had shifted to the right, the fixed dose was increased by 1/8 to 1/4 log unit and the process repeated. Only very modest shifts of the dose-response function to the right were observed, and in several cases curves shifted left after exposure to larger doses. Overall the results suggest that a variable-dosing regimen holds promise as a technique for investigating the development of tolerance to the effects of cocaine, and that the magnitude of tolerance cannot be increased to any great degree by increasing the dose or doses repeatedly experienced. Additionally, it appears that experience with relatively large doses of cocaine may limit the degree to which tolerance can be developed, or decrease the magnitude of tolerance previously observed.

Specificity of effects of chronically administered diazepam on the responding of rats under two different spaced-responding schedules

Lever pressing of rats was maintained in different chambers during two different sessions each day. At 0900 h, responding was maintained under a two-component multiple schedule in which responses initiated an interval that had to elapse before delivery of food (time delay of 20 s and 40 s). In this schedule, a 'response-pause' sequence preceded reinforcers, and acutely administered diazepam only decreased responding. At 1400 h, responding by the same subjects was maintained under a different two-component multiple schedule, in which individual responses initiated an interval that had to be terminated by another response before delivery of food (DRL 20 s and 40 s). In this second schedule, a 'response-pause-response' sequence preceded reinforcers, and acutely administered diazepam increased responding. After studying the acute behavioral effects of diazepam during each separate 'timing' schedule, animals systematically received 1.7 mg/kg per day diazepam 2-5 min prior to their different schedule components, in order to study the influence of reinforcement contingency on the chronic effects of this drug. Diminution of the initial effects of diazepam during daily drug administration prior to DRL 20 s responding did not extend to DRL 40 s responding or to time-delay responding, and tolerance did not develop at all for time-delay responding. When diazepam was again administered after all the daily schedules for approximately 1 month, and then given before the individual DRL schedules, DRL responding was increased again as it had been prior to chronic drug administration. These results suggest that the behavioral effects of acutely administered diazepam are influenced by different 'timing' requirements, and that the behavioral effects of chronically administered diazepam are influenced by 'timing' requirements and by drug- and chamber-related stimuli.

Cocaine Addiction: Neurobiology and Related Current Research in Pharmacotherapy

In this article, recent research studies in the field of cocaine addiction are reviewed, with an eye toward emergent options for treatment innovation. Particular attention is paid to the neurobiology and specific neurotransmitter and receptor mechanisms involved in cocaine abuse, dependence, and other unique phenomena of addiction such as sensitization, craving, compulsive drug use, and withdrawal. The vicissitudes in the dopamine theory of brain reward mechanisms, dopaminergic effects of cocaine, and emerging roles of GABA, serotonin, glutamate, and nitric oxide in cocaine addiction and its sequelae are discussed. Neuroanatomic findings elicited with imaging studies using PET and functional MRI are summarized. These findings support the role of specific brain regions within the dopaminergic system such as the ventral tegmentum and nucleus accumbens in the induction of the cocaine "high" and craving, respectively. Research approaches to the problem of developing effective pharmacotherapeutic options to render cocaine ineffective and modalities under study, such as dopamine uptake inhibitors and immunotherapy, are also discussed in the context of a variety of practical problems faced by these experimental therapies. Pharmacotherapeutic strategies and new directions in this research, such as the adaptive changes of the opioid system in cocaine addiction, are reviewed. Potential areas for further study are brought forth for further debate and possible clinical evaluation.

Long-term alterations in benzodiazepine, muscarinic and alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor density following continuous cocaine administration

Animal models of clinical phenomena, such as stimulant-induced psychosis have focused primarily on persisting alterations that develop in brain after chronic stimulant administration. The present study utilized autoradiographic measures to examine changes in the density of benzodiazepine ([3H] flunitrazepam), muscarinic ([3H] quinuclidinyl benzilate), and non-NMDA glutamatergic (3H alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid: AMPA) receptor binding in rats 21 days following two exposures to cocaine administered continuously for 5 days via subcutaneous pellets. A marked, selective increase in [3H] flunitrazepam binding in both the lateral and medial habenula nucleus was observed. Reduced [3H] quinuclidinyl benzilate binding was observed in various brain areas, including large decreases in the anterior cingulate cortex and ventral thalamus. A reduction in [3H]AMPA binding was observed in the ventral striatum and was suggested in the nucleus accumbens. [3H] Flunitrazepam binding was also examined 12 hr following a single 5 day cocaine exposure to determine if the long-term habenular changes were evident at acute withdrawal. No alterations in [3H] flunitrazepam binding were observed in the habenula or any other structure analyzed at this time point. The relation of these results to persisting alterations in mesocorticolimbic pathways and previous findings of cocaine-induced degeneration in lateral habenula circuitry is discussed.

Cocaine preexposure fails to sensitize the acquisition of cocaine-induced taste aversions

In two separate experiments, rats were given either an intraperitoneal (IP) injection of 10 mg/kg cocaine once a day for 10 consecutive days (Experiment 1) or a single IP injection of 40 mg/kg of cocaine (Experiment 2) prior to receiving repeated pairings of a novel saccharin solution with cocaine (32 mg/ kg; subcutaneous; SC). Although vehicle-preexposed subjects given saccharin-cocaine pairings readily acquired an aversion to the cocaine-associated saccharin solution, subjects preexposed to cocaine (whether 10 times or only once) displayed a retarded acquisition of the aversion. That is, cocaine preexposure attenuated the acquisition of cocaine-induced taste aversions. There was no difference in the degree of attenuation between the two preexposure conditions. Thus, under conditions that are effective in inducing sensitization within other behavioral preparations there was no evidence of sensitized cocaine-induced taste aversions. The results from the present investigation are similar to reports from this laboratory and others demonstrating that preexposure to cocaine, as with a range of other psychoactive drugs, results in weaker taste aversions. The basis for the attenuating effects of cocaine preexposure was discussed in terms of an adaptation to the aversive effects of cocaine.

A neuroendocrine role in cocaine reinforcement

Cocaine stimulates the secretion of corticosterone and ACTH, probably through a CRF-related mechanism, indicating that the drug activates the HPA axis. Indeed, cocaine has been reported to produce anxiety and to precipitate episodes of panic attack during chronic use and withdrawal in humans and to induce anxiogenic behavior in animals. Cocaine also alters benzodiazepine receptor binding in discrete regions of the rat brain. Some of these changes in binding are obviously related to the convulsions and seizures which are often observed in an acute cocaine overdose. However, data from behavioral studies have suggested that some of these effects may be related directly to cocaine reinforcement since receptor changes also were observed when binding in the brains of rats that self-administered cocaine was compared with that from animals that had received identical yoked, but non-contingent infusions of the drug. In this regard, pretreatment with the benzodiazepine receptor agonists chlordiazepoxide and alprazolam decreased cocaine self-administration without decreasing food-reinforced responding, suggesting that these effects were specific for cocaine. Since this attenuation of self-administration was reversed by increasing the unit dose of cocaine, it is likely that these drugs were decreasing cocaine reinforcement. In contrast, exposure to stress increases vulnerability to self-administer psychostimulants. In these experiments, low-dose cocaine self-administration was related directly to stress-induced increases in plasma corticosterone, such that plasma corticosterone was always greater than 150 ng/ml for rats which subsequently self-administered cocaine at doses of 0.125 mg/kg/infusion or lower, suggesting a threshold for the hormone in cocaine reinforcement. In other experiments, bilateral adrenalectomy completely abolished the acquisition of intravenous cocaine self-administration in naive rats, while metyrapone decreased ongoing self-administration. In addition, ketoconazole pretreatment resulted in patterns of self-administration that were virtually indistinguishable from that observed during saline extinction, suggesting that plasma corticosterone is not only important, but may even be necessary for cocaine reinforcement. The mechanisms through which adrenocorticosteroids alter cocaine reinforcement remain to be determined, but there is increasing evidence that the mesocorticolimbic dopaminergic system is involved. In particular, the medial prefrontal cortex appears to be at least one brain region where dopamine and adrenocorticosteroids may interact to affect cocaine reinforcement.