Making a bad thing worse: adverse effects of stress on drug addiction (original) (raw)

Individuals with stress-related psychiatric disorders, such as anxiety and depression, often engage in some form of drug use. Furthermore, exposure to chronic stressful life events, such as physical or sexual abuse (39), is linked to an increase in nicotine, alcohol, and cocaine usage (40). Recently, a study demonstrated that the greater the physical abuse in childhood (i.e., the longer it lasted), the more likely the subject was to develop drug addiction later in life (41). In addition, stress exposure can increase current drug use and precipitate relapse back to drug-taking behaviors (2, 3). Although chronic stress can produce changes in weight loss as well autonomic and endocrine outputs, the significance of these alterations on vulnerability to addiction behaviors has not been systematically characterized. However, the correlative observations in humans that stress exposure can affect various stages of the addiction cycle are supported by evidence from animal studies. Furthermore, these animal studies have enhanced our ability to investigate the underlying mechanisms and molecular targets that are involved in the interaction between stress and addiction.

Acquisition of drug taking. Acquisition is defined as the initial, rewarding exposure to a drug of abuse with development into more chronic use. It has long been hypothesized that exposure to a stressful event or situation would increase the rate of acquisition to drug taking. In animal models, exposure to physiological as well as physical stressors, including social isolation, tail pinch, and footshock, can enhance initial amphetamine and cocaine self administration (42–44). Furthermore, repeated exposure to forced swim stress can augment the rewarding properties of cocaine (45). These studies implicate stress in modulating the initial rewarding effects of addictive drugs.

Corticosterone release via the HPA axis is vital to the acquisition of drug administration. Inhibiting corticosterone release by adrenalectomy or pharmacologic treatment blocks cocaine self administration in rats (46, 47). Furthermore, corticosterone release following drug administration in rats increases neuronal activity above the critical levels needed for self administration to occur (48). This additional neuronal activation by the HPA axis is particularly evident at lower doses of cocaine, such that doses not normally rewarding are now readily self administered. These results are consistent with a study that examined corticosterone levels in rats that exhibit different behavioral and endocrine responses to a novel environment (49). Rats that showed increased locomotor activity and high corticosterone levels upon exposure to a novel environment were termed high responders, whereas low responders exhibited decreased locomotor activity and lower corticosterone levels. Following this initial classification, animals were trained to self-administer cocaine. Low responders did not learn to self administer cocaine, whereas robust self administration was observed in rats assessed as high responders. Interestingly, daily corticosterone administration induced and maintained amphetamine self administration in the low-responding rats, effectively switching their behavior to that of high responders (49).

The ability of corticosterone to modulate cocaine reward can be mediated by glucocorticoid receptors (GRs) located on neurons throughout the mesolimbic dopamine reward pathway (50). Adrenalectomized animals exhibit a blunted dopamine response in the NAc following either drug exposure (51) or stress (7). Corticosterone replacement prevents the attenuation of this dopamine response. Furthermore, GR antagonists decrease extracellular dopamine levels in the NAc by 50% (52), similar to the decrease observed following an adrenalectomy (51). In addition, GR antagonists locally injected into the VTA decrease morphine-induced increases in locomotor activity (52), indicating that activation of GRs in the VTA can mediate dopamine-dependent behavioral outputs. Interestingly, in mice where the gene encoding the GR was deleted specifically in the CNS, a dose-dependent decrease in motivation to self administer cocaine was observed (53). These results suggest that the dopamine increase observed in rodents following either drug administration (51) or stress (7) is dependent, at least in part, on the release of corticosterone from the HPA axis and the subsequent activation of GR.

The role of CRF in the acquisition of drug reward has not been thoroughly investigated. CRF protein and mRNA levels are altered following acute administration of many addictive drugs (34). Studies using CRFR1 antagonists demonstrate their involvement in the initial behavioral and biochemical effects of cocaine. For example, pharmacological blockade of CRFR1 inhibits cocaine-induced dopamine release (54) as well as reductions in the rewarding properties of cocaine (54) and locomotor activating effects (54, 55). These studies point to a role of CRF in modulating the initial effects of addictive drugs, but more studies are needed to fully determine the role of CRF in the development of drug addiction.

Withdrawal. Abnormalities in stress circuitry continue following the cessation of drug taking, in both immediate and long-term withdrawal. Activation of the HPA axis, as evidenced by a marked increase in corticosterone levels, occurs following acute withdrawal from most drugs of abuse both in humans and in animal models (20). Interestingly, following this initial activation, basal corticosterone and cortisol levels return to normal in humans and rodents, respectively (20). However, during long-term withdrawal from psychostimulants and opiates, the HPA axis displays an augmented response upon exposure to a stressor. In former cocaine (56) and opiate addicts (57), increased levels of ACTH and cortisol were measured following administration of the chemical stressor metyrapone. Metyrapone blocks the synthesis of cortisol, disrupting the normal negative feedback of cortisol on the hypothalamus and thereby causing activation of the HPA stress pathway (20). Furthermore, in abstinent cocaine users, hyperresponsiveness to emotional and physical stress, as well an increased drug craving, is observed (58), which is consistent with an altered HPA axis. In rats, during acute withdrawal, corticosterone responses are augmented upon exposure to restraint stress (59). These data suggest that the stress response can be sensitized by drug exposure and subsequent withdrawal. In contrast, recent evidence has demonstrated an attenuated response to stress during nicotine withdrawal in animals. Corticosterone levels were substantially lower in rats exposed to restraint stress during nicotine withdrawal, although their basal corticosterone levels were similar (60). Chronic smokers demonstrate increased cortisol secretion (61, 62), and a reduction in cortisol after smoking cessation has been associated with increased withdrawal severity and relapse (63, 64). Together, these studies demonstrate alterations in the responsiveness of the HPA axis to a stressor during long-term withdrawal, which might play a role in the ability of stressors to reinstate drug seeking well after the drug is removed.

Alterations in CRF peptide and mRNA levels throughout the CNS are observed following acute withdrawal from several drugs of abuse, including cocaine and opiates, and these alterations vary by brain region as well as the drug administered. Interestingly, increases in CRF mRNA in the PVN correlate with increases in anxiety behaviors during ethanol, cocaine, and morphine withdrawal (65–67). In addition, blockade of the CRF system with antagonists or antibodies decreases the anxiety observed in this acute withdrawal phase (65–67). CRFR1 antagonists decreased the physical symptoms of morphine withdrawal in dependent rats (67). Together, these data suggest that the CRF system plays a role in the psychological as well as the physical symptoms of acute withdrawal from addictive drugs. However, the role of CRF or stress circuitry in long-term withdrawal has yet to be elucidated.

Reinstatement of drug seeking. Many theories of addiction hypothesize that stress is one of the primary causes of relapse in human addicts (2, 3). Using animal models, several laboratories have demonstrated that exposure to an acute stressor can effectively reinstate drug seeking of various drugs, including opiates, psychostimulants, alcohol, and nicotine (68–71). Stress facilitates relapse by activating central CRF brain circuits. Animals that have been trained to self administer drug and then have the drug removed reinitiate lever pressing following an intracerebroventricular CRF injection (72). A distinct circuitry involving CRF in the extended amygdala, an important structure for emotional and effective behavior, has been delineated in mediating stress-induced relapse. Structures comprising the extended amygdala overlap with those of the reward pathway, including the central nucleus of the amygdala, bed nucleus of the stria terminalis (BNST), and parts of the NAc (Figure 2) (73). The significance of this pathway in the addiction cycle is evident primarily in relapse or reinstatement. Inactivation of the CRF projection from the central amygdala to the BNST blocks stress-induced (e.g., by footshock) cocaine reinstatement (74, 75), and local injections of D-Phe, a nonspecific CRF receptor antagonist, into the BNST, but not the amygdala, attenuates footshock-induced reinstatement (75). Specifically, CRFR1s localized in the BNST, but not the amygdala or NAc, mediate stress-induced relapse into drug seeking (68). Interestingly, selective CRFR1 antagonists attenuate footshock-induced reinstatement of cocaine or opiate seeking (68, 76) but have no effect on drug-induced reinstatement (72, 77). These data demonstrate that stress stimulation of the CRF-containing pathway, originating in the amygdala and extending into the BNST, and subsequent activation of CRFR1 localized in the BNST, triggers drug seeking in previously addicted animals.

Recently, CRF has been detected in the VTA, the site of origin of the dopamine neurons of the reward pathway (78). In both cocaine-naive and cocaine-experienced rats, CRF is released into the VTA following an acute footshock; however, the source of this CRF is not known (78). In cocaine-experienced animals, glutamate and dopamine are released in the VTA in conjunction with CRF in response to a stressor. This release of glutamate and dopamine is dependent upon CRF and subsequent activation of its receptors, as local injections of CRF antagonists into the VTA attenuated the release of these 2 neurotransmitters (78, 79). In addition, administration of CRFR2 antagonists, but not CRFR1 antagonists, locally into the VTA blocked the ability of footshock to reinstate cocaine seeking in a self-administration paradigm (79). Taken together, these studies suggest a role for CRF in modulating dopamine cell activity, specifically following drug experience.

Although studies have clearly demonstrated a role for CRF in reinstatement of stress-induced drug seeking, very few have examined whether other molecular mechanisms are important in stress-induced reinstatement. The transcription factor CREB, implicated in both stress and addiction, was recently shown to be involved in stress-induced reinstatement. CREB-deficient mice do not exhibit stress-induced reinstatement of cocaine-conditioned place preference (70). However, these mice do exhibit reinstatement of drug seeking to a priming dose of cocaine (70). This deficit in stress- and not drug-induced reinstatement indicates a specific requirement for CREB in stress-induced behavioral responses to drugs of abuse. Of interest, a putative CREB target gene, brain-derived neurotrophic factor (BDNF), localized in the VTA and the NAc of the mesolimbic dopamine reward pathway, was increased following withdrawal from chronic cocaine (80). The increase in BDNF in these brain areas positively correlated with the response of the rats to drug-associated cues (80), and more recent studies demonstrate that BDNF might facilitate relapse to drug-seeking behavior (81). Additional experiments detailing the molecular mechanisms of stress-induced reinstatement are needed to fully understand this complex process.