Reinstatement of Drug Self-Administration
Reinstatement of drug seeking; Reinstatement of drug-taking behavior
In the learning literature, the term “reinstatement” refers to the resumption of a learned response that occurs when a subject is exposed noncontingently to the unconditioned stimulus. In the addiction field, “reinstatement of drug self-administration” in experimental animals (typically termed “reinstatement of drug seeking”) refers to the resumption of a previously drug-reinforced behavior by noncontingent exposure to drugs, different types of drug cues, or stressors after the extinction of the drug-reinforced behavior. The animal model that addiction researchers use to study the reinstatement of drug seeking is termed the “reinstatement model” (Shaham et al. 2003).
Impact of Psychoactive Drugs
A major problem in the treatment of drug addiction is the high rate of relapse to drugs of abuse following a prolonged abstinence period. Results from studies in humans suggest that in drug-free individuals, relapse to drug use during periods of forced or voluntary abstinence can be triggered by acute exposure to the self-administered drug, stimuli previously associated with drug taking, or stressors. Acute exposure to the self-administered drug or related drugs, drug-associated cues, or stress also reinstates drug seeking in laboratory rats and monkeys (Shaham et al. 2003). Because of the similarities between the human condition and the laboratory animal model, many investigators currently use the reinstatement model to study mechanisms underlying relapse to drug use. Conceptual issues related to the validity of the reinstatement model as an animal model of relapse to drug use are discussed by Epstein et al. (2006).
In the drug self-administration version of the reinstatement model, which is based on an operant-conditioning procedure, laboratory animals (rats, mice, and monkeys) are initially trained to self-administer drugs by lever pressing (or another operant response such as nose poking) for intravenous drug infusions (or for oral delivery of alcohol). Subsequently, the drug-reinforced behavior is extinguished by substituting the drug solutions with saline or by disconnecting the infusion pumps. After the extinction of the drug-reinforced behavior, the ability of acute noncontingent exposure to drugs (termed “drug priming”), drug-associated cues, or stress to reinstate lever responding is measured under extinction conditions (Shaham et al. 2003). The non-reinforced responding on the “active lever” – the lever that previously delivered the drug – is interpreted to reflect the reinstatement of drug seeking. The responding on the “inactive lever” – the lever that has not been associated with drug injections – is often interpreted to reflect nonspecific activity, but it may also reflect response generalization.
In laboratory animals, the reinstatement of drug seeking has been studied using the “between-session,” “within-session,” and “between-within-session” variations of the reinstatement model (Shalev et al. 2002). In the between-session procedure, training for drug self-administration, the extinction of the drug-reinforced behavior, and tests for reinstatement are conducted on different daily sessions. In the within-session procedure, all three phases are conducted within a daily session. This daily session consists of 1–2 h of drug self-administration that is followed by 3–4 h of extinction of the drug-reinforced behavior; after extinction of lever responding, a test for reinstatement is conducted. In the between-within procedure, the laboratory animals are trained daily for drug self-administration; subsequently, the extinction of the drug-reinforced lever responding and tests for reinstatement are conducted on the same day. In addition, researchers also assess the reinstatement of drug seeking in variations of the reinstatement model that are based on the runway and conditioned place preference procedures. Shalev et al. (2002) provide a detailed description of the different procedural variations of the reinstatement model and discuss the advantages and disadvantages of using them for assessing the reinstatement of drug seeking.
Neuropharmacology of Reinstatement of Drug Seeking
In this section, we summarize the extant literature on the neuropharmacology of the reinstatement of drug seeking induced by exposure to the self-administered drug or related drugs (drug priming), drug-associated cues, or stress. Due to space constraints, we often do not differentiate between results obtained from studies in which rats were trained to self-administer cocaine and other psychostimulants, heroin and morphine, or alcohol. A detailed description of the neuropharmacology of the reinstatement of drug seeking is provided by Shalev et al. (2002) and Bossert et al. (2005).
Results from studies on drug-priming-induced reinstatement of drug seeking indicate that this reinstatement is primarily mediated by dopaminergic and glutamatergic neurotransmission in the mesocorticolimbic system (Kalivas and McFarland 2003; Schmidt et al. 2005; Self 2004). This system consists of dopamine cell bodies in the ventral tegmental area (VTA) that project to limbic areas, including nucleus accumbens, amygdala, and bed nucleus of stria terminalis (BNST), and cortical areas, including medial prefrontal cortex (mPFC) and orbital frontal cortex (OFC).
Regarding dopamine, systemic injections of either D1-family dopamine receptor agonists or antagonists decrease cocaine-induced reinstatement; D1-family dopamine receptor antagonists also decrease heroin-priming-induced reinstatement. On the other hand, systemic injections of D2-family receptor antagonists decrease heroin- or cocaine-priming-induced reinstatement, while injections of D2-family receptor agonists reinstate drug seeking. The effect of cocaine priming on reinstatement is also attenuated by systemic injections of selective D3 receptor antagonists. Additionally, injections of D1-family or D2-family receptor antagonists into the dorsal mPFC, accumbens shell (but not core), and amygdala (both central and basolateral subregions) decrease cocaine-priming-induced reinstatement. Accumbens core injections of D1-family or D2-family receptor antagonists decrease heroin-priming-induced reinstatement. Finally, activation of VTA dopamine neurons by local injections of morphine reinstates heroin and cocaine seeking, while local inhibition of these neurons by a mixture of GABAA + GABAB receptor agonists (muscimol + baclofen) decreases cocaine-priming- and heroin-priming-induced reinstatement.
Regarding glutamate, systemic injections of metabotropic glutamate receptor 5 antagonists or group II metabotropic glutamate agonists (which decrease evoked glutamate release via a presynaptic autoreceptor mechanism) decrease cocaine-priming-induced reinstatement. Additionally, local injections of nonselective ionotropic glutamate receptor antagonists into the VTA, selective AMPA receptor antagonists into the accumbens, or reversible inactivation of the glutamate projection from the dorsal mPFC to the accumbens decrease heroin- or cocaine-priming-induced reinstatement. Finally, there is evidence that cocaine self-administration and subsequent withdrawal lead to long-lasting neuroadaptations in glutamatergic transmission in the mPFC-accumbens pathway and that the pharmacological reversal of these neuroadaptations prevents cocaine-priming-induced reinstatement.
Results from studies in which investigators used reversible inactivation manipulations (the sodium channel blocker tetrodotoxin or muscimol + baclofen) suggest a role of the ventral pallidum in cocaine- and heroin-priming-induced reinstatement. In the case of heroin-priming-induced reinstatement, there is also evidence for a role of the BNST (dorsal and ventral subregions), dorsal striatum, substantia nigra, and ventral mPFC. Additionally, there is evidence from pharmacological studies that other neurotransmitter systems, including endocannabinoids, serotonin, and GABA, contribute to drug-priming-induced reinstatement. However, the brain sites involved in these effects are unknown. Finally, there is evidence for a role of accumbens protein kinase A and calcium-calmodulin-dependent kinase II signaling pathways in cocaine-priming-induced reinstatement.
In humans, relapse-provoking drug-associated stimuli can be divided into two general categories: discrete drug cues (e.g., drug paraphernalia) that are associated with the acute rewarding effects of the drug and contextual drug cues (e.g., a specific environment such as a local bar) that predict drug availability. In laboratory animals, procedures to assess cue-induced reinstatement of drug seeking are classified into three types according to the type of the conditioned cue: discrete cue, discriminative cue, and contextual cue.
Discrete-cue-induced reinstatement: Results from pharmacological studies suggest a role of several neurotransmitters and receptors in discrete-cue-induced reinstatement (Feltenstein and See 2008). Systemic injections of D1-family receptor antagonists, selective D3 receptor antagonists, nicotinic cholinergic antagonists, several serotonergic agents, group II metabotropic receptor agonists, cannabinoid 1 (CB1) receptor antagonists, and mu opiate receptor antagonists decrease the discrete-cue-induced reinstatement of drug seeking. In the case of D1-family receptors, two brain sites are critical for their effects: the amygdala (both central and basolateral subregions) and the accumbens core (but not shell). The brain sites involved in the modulation of the discrete-cue-induced reinstatement by the other neurotransmitter systems are unknown. Additionally, results from studies in which reversible inactivation manipulations were used suggest a role of the dorsal mPFC and dorsal striatum in the discrete-cue-induced reinstatement. In heroin-experienced rats, there is evidence that discrete-cue-induced reinstatement also involves the ventral mPFC, the dorsal BNST, the ventral pallidum, and the substantia nigra.
Discriminative-cue-induced reinstatement: Results from pharmacological studies suggest a role of several neurotransmitter systems in discriminative-cue-induced reinstatement (Weiss 2005). Systemic injections of D1-family and D2-family receptor antagonists, selective D3 receptor agonists, group II metabotropic receptor agonists, metabotropic glutamate receptor type 1 or type 5 antagonists, CB1 receptor antagonists, sigma 1 receptor antagonists, 5-HT2B/2C receptor agonists, and mu opioid receptor antagonists decrease the discriminative-cue-induced reinstatement. The discriminative-cue-induced reinstatement is also attenuated by ventricular injections of the opioid peptide nociceptin/orphanin FQ (the endogenous ligand for the opioid-like orphan receptor). In the case of CB1 receptor antagonists, two brain sites are critical for their effects: the mPFC and the accumbens. The brain sites involved in the modulation of discriminative-cue-induced reinstatement by the other neurotransmitter systems are unknown. However, in the case of dopamine receptor antagonists, likely brain sites are the amygdala, the accumbens, and the mPFC. Thus, exposure to cocaine discriminative cues increases dopamine release in the accumbens and amygdala. Additionally, discriminative-cue-induced Fos (a neuronal activity marker) expression in the basolateral amygdala and prefrontal cortex is decreased by systemic injections of a D1-family receptor antagonist.
Context-induced reinstatement: Results from pharmacological studies suggest a role of several neurotransmitter systems in context-induced reinstatement (Crombag et al. 2008). Systemic injections of D1-family and D2-family receptor antagonists, group II metabotropic receptor agonists, CB1 receptor antagonists, and 5-HT2B/2C receptor agonists decrease context-induced reinstatement. In the case of D1-family receptors, a critical brain site is the accumbens shell. In the case of group II metabotropic receptors, the critical brain sites are the accumbens shell and VTA. The brain sites involved in the modulation of the context-induced reinstatement by the other neurotransmitter systems are unknown. Results from studies in which reversible inactivation methods were used suggest a role of the dorsal mPFC, the dorsal striatum, the basolateral amygdala, and the dorsal hippocampus in context-induced reinstatement. Results from studies on the effect of context-induced reinstatement testing on Fos expression in the brain suggest a role of the lateral hypothalamus in this reinstatement.
The phenomenon of stress-induced reinstatement of drug seeking was initially demonstrated in studies in which investigators used an intermittent footshock stressor. The effect of this stressor generalizes to certain other stressors, including swim stress, acute food deprivation, social defeat, and several pharmacological stressors: the stress neurohormone corticotropin-releasing factor (CRF), the anxiogenic drug yohimbine (an alpha-2 adrenoceptor antagonist), and the agonists of the kappa opioid receptor.
Evidence from studies in which pharmacological agents were used suggests a role of several neurotransmitter systems in stress-induced reinstatement. The stressor used in most of the neuropharmacological studies is intermittent footshock. Systemic injections of nonselective CRF antagonists, selective CRF1 receptor antagonists, alpha-2 adrenoceptor agonists (which decrease noradrenaline cell firing and release), hypocretin 1 receptor antagonist, nonselective dopamine receptor antagonist, selective serotonin reuptake blockers (SSRIs), and 5HT3 receptor antagonists decrease stress-induced reinstatement. The stress-induced reinstatement of alcohol (but not cocaine) seeking is attenuated by ventricular injections of nociceptin/orphanin FQ.
The effect of CRF antagonists on stress-induced reinstatement is independent on the activity of the hypothalamic-pituitary-adrenal axis (HPA) and the effect of these antagonists on the stress hormone corticosterone. The critical extrahypothalamic brain sites and projections for CRF’s role in footshock-induced reinstatement are the BNST, VTA, and a projection from the central amygdala to the BNST. Within the VTA, footshock causes local CRF release, which leads to increased glutamate transmission; this enhanced glutamate transmission is critical for stress-induced reinstatement, presumably via an activation of the mesocorticolimbic dopamine system. Support for this notion is provided by the findings that injections of D1-family receptor antagonists into the dorsal mPFC or OFC or preferential D3 receptor antagonists into the accumbens decrease stress-induced reinstatement. The critical brain sites and projections for noradrenaline’s role in footshock-induced reinstatement are the central amygdala and BNST and the noradrenergic projection from the lateral tegmental nuclei to these brain sites.
Finally, results from studies in which reversible inactivation methods were used confirm the findings on the role of the dorsal mPFC, the BNST, the central amygdala, the accumbens, and the VTA in stress-induced reinstatement and further suggest a role of the ventral pallidum in this reinstatement. Finally, as in the case of drug-priming-induced reinstatement, the glutamatergic projection from the mPFC to the accumbens plays an important role in stress-induced reinstatement.
The writing of this chapter was supported by the Intramural Research Program of the NIH, NIDA.