Portal de Periódicos da CAPES

Drug Addiction: The Yin and Yang of Hedonic Homeostasis

1996; Cell Press; Volume: 16; Issue: 5 Linguagem: Inglês

10.1016/s0896-6273(00)80109-9

1097-4199

George F. Koob,

Neural and Behavioral Psychology Studies

Drug addiction is usually defined as compulsion to take a drug with loss of control over drug intake. The term substance dependence is used to describe a syndrome basically equivalent to addiction and the diagnostic criteria used describe symptoms that lead to loss of control in drug intake. Substance use, substance abuse, and substance dependence are separate, definable entities in most formulations. An important challenge for neurobiological research is to understand how the transition occurs between controlled drug use and the loss of control that defines addiction or substance dependence and what molecular, cellular, and system processes contribute to the development of drug dependence. A neuroadaptive view of drug dependence was originally based on the phenomenon of tolerance and withdrawal, adaptive processes that have been hypothesized to be the body's attempt to counter the acute effects of the drug. Such adaptations have been explored at all levels of drug dependence research from the behavioral to the molecular (reviewed by8Koob G.F Bloom F.E Science. 1988; 242: 715-723Crossref PubMed Scopus (1722) Google Scholar). More recently, another adaptive process, the phenomenon of sensitization, has been conceptualized as a critical neuroadaptive process (18Robinson, T.E., Berridge, K.C. (1993). Brain Res. Rev. 18, 247–291.Google Scholar). Motivational View of Drug Dependence There are four major sources of reinforcement in drug dependence: positive reinforcement and negative reinforcement, conditioned positive reinforcement, and conditioned negative reinforcement (23Wikler A Arch. Gen. Psychiatry. 1973; 28: 611-616Crossref PubMed Scopus (473) Google Scholar). Clearly, positive reinforcing effects are critical for establishing self-administration behavior, which leads to the hypothesis that positive reinforcement is the key to drug dependence (24Wise R J. Abnorm. Psychol. 1988; 97: 118-132Crossref PubMed Scopus (428) Google Scholar). In contrast, neuroadaptation theories, such as opponent-process theory, postulate that the processes of affective habituation (hedonic tolerance) and affective withdrawal may be the driving force of addiction (21Solomon, R.L. (1977). In Psychopathology: Experimental Models, J.D. Maser and M.E.P. Seligman, eds. (San Francisco: W. H. Freeman and Company), pp. 124–145.Google Scholar). Clearly, this construct plays an important role in the maintenance of drug use after the development of dependence. Thus, while initial drug use may be motivated by the positive affective state produced by the drug, continued use leads to neuroadaptation to the presence of drug and to another source of reinforcement, the negative reinforcement associated with relieving negative affective consequences of drug termination. Indeed, the defining feature of drug dependence has been argued to be the establishment of a negative affective state (19Russell, M.A.H. (1976). In Drugs and Drug Dependence, G. Edwards, M.A.H. Russell, D. Hawks, and M. MacCafferty, eds. (Westmead, England: Saxon House/Lexington Books), pp. 182–187.Google Scholar). While a negative affective state is difficult to measure in animals, there is evidence of a compromised brain reward system during drug withdrawal in chronically exposed animals. Intracranial self-stimulation (ICSS) behavior has proven to be particularly sensitive to changes in the brain reward systems during the course of drug dependence. Acute administration of drugs that are abused in humans increases the reward value of ICSS, and drug withdrawal decreases the reward value of ICSS (reviewed by9Koob G.F Markou A Weiss F Schulteis G Semin. Neurosci. 1993; 5: 351-358Crossref Scopus (75) Google Scholar). The increases in reward threshold or decreases in reward have been observed following the withdrawal of opiates, stimulants, and alcohol and appear to be dose related in that the more drug that is administered over time, the larger the withdrawal response. If a deficit reward state defines addiction, it could be produced by several mechanisms. First, genetic or environmental factors could produce some neurobiological deficit that requires reversal, perhaps by drug initiation. Alternatively, chronic drug taking itself could produce a form of deficit state that required self-medication to reverse. This type of deficit state could be hypothesized to exist at the molecular, cellular, and system level, and the minireviews in this issue discuss numerous neuropharmacological mechanisms for such perturbations. However, drug dependence not only involves acquisition of drug taking and maintenance of drug taking, but also functions as a chronic relapsing disorder with reinstatement of drug taking after detoxification and abstinence. Both the positive and negative affective states can become associated with stimuli in the drug-taking environment or even internal cues through classical conditioning processes (23Wikler A Arch. Gen. Psychiatry. 1973; 28: 611-616Crossref PubMed Scopus (473) Google Scholar). Reexposure to these conditioned stimuli can provide the motivation for continued drug use and relapse after abstention. There is evidence in humans that the positive reinforcing effects of drugs such as heroin and cocaine, as measured by subjective reports of euphoria or "high," can become conditioned to previously neutral stimuli. Patients being treated for heroin addiction and allowed to self-administer either saline or heroin reported that both saline and heroin injections were pleasurable, particularly in the patient's usual injection environment. Alternatively, patients, even detoxified subjects, can report negative affective symptoms like those associated with drug abstinence when returning to environments similar to those associated with drug dependence (16O'Brien C.P Pharmacol. Rev. 1976; 27: 533-543Google Scholar). The neurobiological bases for the syndrome of protracted abstinence may involve subtle molecular and cellular changes that are the challenge of future research on the neurobiology of addiction. Key elements of some of the neuroadaptive processes that constitute protracted abstinence can be found in the minireviews of this issue. Tolerance has been hypothesized to develop via two possible mechanisms. Habituation occurs when the primary response to the drug diminishes with repeated presentation. Alternatively, the initial acute effect of the drug may be opposed or counteracted by homeostatic changes in systems that mediate the primary drug effects, i.e., a homeostatic adaptive process. These two models are not necessarily mutually exclusive (12Littleton, J.M., and Little, H.J. (1989). In Psychoactive Drugs: Tolerance and Sensitization, A.J. Goudie and M.W. Emmett-Oglesby, eds. (Clifton, New Jersey: Humana Press), pp. 461–518.Google Scholar). In the opponent-process theory, both tolerance and dependence are inextricably linked (21Solomon, R.L. (1977). In Psychopathology: Experimental Models, J.D. Maser and M.E.P. Seligman, eds. (San Francisco: W. H. Freeman and Company), pp. 124–145.Google Scholar). According to this theory, affective states, pleasant or aversive, are automatically opposed by centrally mediated mechanisms that reduce the intensity of these affective states. Thus, positive reinforcers such as drugs engage positive hedonic processes that are opposed by negative hedonic processes. The positive hedonic processes are hypothesized to be simple and stable and to follow administration of the drug closely in time. In contrast, the negative hedonic processes are of longer latency, slow to build up strength, and slow to decay. Within this framework, the intense pleasure of a drug "rush" or "high" is presumed to reflect a positive hedonic process, and the negative mood state associated with the drug wearing off or abstinence following a drug binge is presumed to reflect the opponent negative hedonic process. Sensitization has been defined as the long-lasting increment in response occurring upon repeated presentation of a stimulus that reliably elicits a response at its initial presentation. With drugs, sensitization is more likely to occur with intermittent exposure to a drug, in contrast with tolerance, which is more likely to occur with continuous exposure. In a recent conceptualization of the role of sensitization in drug dependence, drug craving was hypothesized to be progressively increased by repeated exposure to drugs of abuse (18Robinson, T.E., Berridge, K.C. (1993). Brain Res. Rev. 18, 247–291.Google Scholar); the transition to pathologically strong craving would then define compulsive use. In an attempt to bring homeostatic models to the cellular and molecular levels of analysis, adaptations can be distinguished between those occurring within a drug-sensitive reinforcement system and those occurring between interacting systems (8Koob G.F Bloom F.E Science. 1988; 242: 715-723Crossref PubMed Scopus (1722) Google Scholar). Similar neurobiological distinctions could also be made for the neuroadaptations associated with sensitization. In a within-system adaptation, repeated drug administration would elicit an opposing reaction within the same system in which the drug elicits its primary reinforcing actions. In a between-system adaptation, repeated drug administration would recruit a different neurochemical system, one not involved in the acute reinforcing effects of the drug but one that, when activated or engaged, would act in opposition to the primary reinforcing effects of the drug. Many examples of within-system and between-system adaptations can be found in the four minireviews to follow. Some of these neuroadaptations are specific to each drug class, while others may be common to all drugs of abuse. Even more intriguing is the possibility that the added system dimension of neurobiological circuitry may represent not only a common substrate for the reinforcing actions of drugs but a common substrate for the neuroadaptations known as substance dependence or addiction. Chronic morphine treatment has long been associated with increases in adenylate cyclase activity, an action opposite to that of acute administration (Nestler, 1996 [this issue of Neuron]). These effects have not only been observed in the locus coeruleus, a brain structure implicated in the physical signs of opiate withdrawal, but also in the region of the nucleus accumbens, a structure implicated in the motivational effects of opiate withdrawal. In addition, chronic opiate exposure also decreases levels of neurofilament protein in the ventral tegmental area, which contains the cell bodies of origin of the dopamine (DA) projection to the nucleus accumbens; administration of neurotrophic factors can reverse these effects. Injection of brain-derived neurotrophic factor into the brain prevented and reversed the structural changes observed in the ventral tegmental area DA neurons as well as the changes in cyclic AMP (cAMP) activity in the nucleus accumbens (15Nestler E.J Neuron. 1996; 16 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (233) Google Scholar). This opens the possibility that the long-term neuroadaptations associated with opiate dependence that have motivational consequences may involve changes in mesolimbic DA function that can be regulated by neurotrophic factors. Two major neurobiological actions of cocaine and amphetamine may represent neuroadaptations important for motivational effects of these drugs. Activation of D1 receptors stimulates a cascade of events that ultimately lead to cAMP response element–binding protein (CREB) phosphorylation and subsequent alterations in gene expression (Hyman, 1996 [this issue of Neuron]). These within-system adaptations not only could change the function of the DA system itself but may also trigger a second major neurobiological action, the increase in expression of protachykinin and prodynorphin mRNA. The subsequent activation of dynorphin systems could contribute to the dysphoric syndrome associated with cocaine dependence and also provide feedback to decrease DA release. Enhanced dynorphin actions could then be considered a between-system adaptation. The significance of the D1 receptor–cAMP–CREB pathway in adaptations to drugs of abuse is supported by the recent evidence of effective anti-cocaine actions from DA D1 antagonists and effective anti-cocaine priming effects of D1 agonists (20Self D.W Barnhart W.J Lehman D.A Nestler E.J Science. 1996; 271: 1586-1589Crossref PubMed Scopus (431) Google Scholar). The initial molecular site of action for nicotine is likely to be the nicotinic acetylcholine receptors, specifically those localized in the brain mesolimbic DA system. The capability of these receptors to exist in many different functional states has led to a combined within-system/between-system adaptation hypothesis to explain tolerance and dependence (Dani and Heinemann, 1996 [this issue of Neuron]). Acute nicotine briefly stimulates receptors that are normally in a closed resting state to an open state for conducting cations, but then they return to a desensitized state in which they are unresponsive to the agonist. In addition, long-term nicotine exposure causes an increase in the actual number of nicotinic acetylcholine receptors. Thus, the following scenario has been hypothesized by Dani and Heinemann. Nicotine stimulates the mesolimbic DA system, among other systems, via activation of nicotinic acetylcholine receptors to produce the reinforcing effects of nicotine. The inactivation of the receptors by desensitization leads to acute tolerance, which is compensated by an increase in receptors. During abstinence, the smoker's nicotine levels fall, and the excess nicotinic acetylcholine receptors begin to recover to a responsive state. However, this heightened responsitivity to ordinary synaptic input in the acetylcholinergic pathways would not be restricted to the receptors on the DA system but may engage nicotinic receptors on other, nonrewarding acetylcholinergic pathways that could contribute to the negative affective state associated with nicotine withdrawal. Smokers are, in effect, medicating themselves with nicotine to regulate the number of functional nicotinic acetylcholine receptors according to this hypothesis. A major advance in recent years has been the discovery of significant selectivity in the initial site of action of ethanol, particularly for low doses. Rather than acting through relatively unselective perturbations of membrane lipids, ethanol appears to act selectively on proteins involved in signal transmission and transduction (22Tabakoff B Hoffman P Neuron. 1996; 16 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar [this issue of Neuron]). Five neurotransmitter systems have been implicated in the acute reinforcing actions of ethanol: γ-aminobutyric acid (GABA), glutamate, DA, serotonin, and opioid peptides (11Koob, G.F., Rassnick, S., Heinrichs, S., and Weiss, F. (1994b). In Toward a Molecular Basis of Alcohol Use and Abuse, H. Jörnvall, ed. (Proceedings of Nobel Symposium on Alcohol), pp. 103–114.Google Scholar). Recent work has begun to identify elements of neuroadaptation in these neurotransmitter systems that result from chronic administration of ethanol. Pharmacological data suggest a role for GABA in ethanol dependence in that GABA agonists attenuate ethanol withdrawal and GABA antagonists exacerbate ethanol withdrawal. At a molecular level, there are data showing that the potentiation by ethanol of GABA-stimulated chloride flux is greatly reduced by chronic exposure to ethanol, and there are changes in the expression of particular subunits of the GABAA receptor with chronic ethanol exposure. In contrast with the potentiation of GABA signaling, there is significant evidence that acute ethanol at quite low doses decreases glutamate (NMDA) receptor function, possibly through an action on the glycine site of the receptor. Chronic treatment with ethanol appears to increase the number of binding sites for NMDA receptor ligands that again may reflect changes in the absolute or relative quantity of NMDA receptor subunit proteins. These effects on GABAA and NMDA receptor function may be examples of within-system adaptations, but the link to actual motivational effects of ethanol dependence remains to be determined. Two brain systems definitely linked to the motivational effects of ethanol are the DA system and the opioid peptide systems. Ethanol does appear to activate the mesolimbic DA system, possibly through an interaction with NMDA receptors, but perhaps more intriguing is the observation that DA function is decreased during ethanol withdrawal, and this decrease is reversed by ethanol administration (7Koob, G.F. (1996). In Pharmacological Effect on Ethanol on the Central Nervous System, R. Dietrich and V.G. Erwin, eds. (Boca Raton, Florida: CRC Press), pp. 1–12.Google Scholar). A role for opioid peptides in ethanol dependence derives from the preclinical observation of a decrease in ethanol self-administration with opioid antagonists and from the clinical observation that opioid antagonist administration can help prevent relapse in detoxified alcoholics (7Koob, G.F. (1996). In Pharmacological Effect on Ethanol on the Central Nervous System, R. Dietrich and V.G. Erwin, eds. (Boca Raton, Florida: CRC Press), pp. 1–12.Google Scholar). Serotonin has been implicated in the acute reinforcing effects of ethanol by the observation that serotonin reuptake inhibitors decrease ethanol intake in animals and humans (7Koob, G.F. (1996). In Pharmacological Effect on Ethanol on the Central Nervous System, R. Dietrich and V.G. Erwin, eds. (Boca Raton, Florida: CRC Press), pp. 1–12.Google Scholar). All of these changes would be considered within-system changes where the neurotransmitter system responsible for part of the acute reinforcing actions of ethanol is altered by chronic exposure to ethanol. One example of a potential between-system neuroadaptation with chronic ethanol is the activation of the neuropeptide corticotropin-releasing factor (CRF), a peptide implicated in behavioral responses to stressors. CRF antagonists injected into the amygdala reverse the anxiogenic-like effects of ethanol withdrawal, and ethanol withdrawal is characterized by increased release of CRF into the amygdala (10Koob G.F Heinrichs S.C Menzaghi F Pich E.M Britton K.T Semin. Neurosci. 1994; 7: 221-229Crossref Scopus (133) Google Scholar, 17Pich E.M Lorang M Yaganeh M DeFonseca F.F Koob G.F Weiss F J. Neurosci. 1995; 15: 5439-5447PubMed Google Scholar). The changes in the brain associated with the development of drug dependence outlined above have several common elements that may define a basic drug addiction circuitry. This circuitry can be defined at three levels of inquiry: molecular, neurochemical, or neuroanatomical. A system common to the acute and chronic actions of all drugs of abuse is the mesolimbic DA system. Activation of this system is critical for the reinforcing actions of stimulants, possibly including nicotine. The mesolimbic DA system also participates in opiate and alcohol reinforcement, although its integrity is not critical to support self-administration of those drugs. Drug withdrawal is accompanied by a decrease in DA function, and elements of protracted abstinence may involve long-term changes in the DA system. Neuroadaptation within this system clearly can develop by various means from the desensitization of nicotinic receptors (2Dani J.A Heinemann S Neuron. 1996; 16 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (506) Google Scholar) to the induction of signal transduction elements pre- and postsynaptically to changes in response to growth factors (15Nestler E.J Neuron. 1996; 16 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (233) Google Scholar). Evidence is strong for changes in opiate receptor signal transduction associated with chronic opiates (15Nestler E.J Neuron. 1996; 16 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (233) Google Scholar), but indirect evidence also exists for changes with cocaine and even chronic ethanol (14Nestler E.J J. Neurosci. 1992; 12: 2439-2450Crossref PubMed Google Scholar). Evidence for between-systems adaptations following chronic drugs of abuse can be found in studies exploring the role of several neuropeptides, notably dynorphin, CRF, and neuropeptide FF (NPFF). Dynorphin peptides appear to regulate the DA system via an action of κ opioid receptors on DA nerve terminals; and κ agonists produce aversive effects in rodents and humans (5Hyman S.E Neuron. 1996; 16 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). Thus, the induction of prodynorphin gene by chronic administration of cocaine or amphetamine could be an example of a neuroadaptation that is recruited by the chronic drug exposure but has opposing effects on brain reinforcement systems. Immunoreactivity for NPFF and its receptors has been reported in the CNS of the rat. Anti-opiate activities of NPFF have been hypothesized based on the effects of intracerebroventricular (ICV) injection of NPFF-related peptides (13Malin D.H Lake J.R Hammond M.V Fowler D.E Rogillio R.B Brown S.L Sims J.L Leecraft B.M Yang H.-Y.T Peptides. 1990; 11: 969-972Crossref PubMed Scopus (147) Google Scholar). NPFF attenuates morphine- and stress-induced analgesia, and an NPFF antagonist can increase both morphine- and stress-induced analgesia and reverse morphine tolerance. ICV administration of an NPFF antagonist also attenuates a naloxone-precipitated withdrawal syndrome in morphine-dependent rats. It remains to be determined whether these anti-opiate peptides have motivational significance and contribute to the negative affective state produced during drug abstinence. CRF is another neuropeptide that may be involved in the motivational aspects of drug dependence. Rats treated repeatedly with cocaine and ethanol show significant anxiogenic-like responses following cessation of chronic drug administration that are reversed with ICV administration of a CRF antagonist (10Koob G.F Heinrichs S.C Menzaghi F Pich E.M Britton K.T Semin. Neurosci. 1994; 7: 221-229Crossref Scopus (133) Google Scholar). However, a CRF antagonist was also active in reversing the aversive effects of opiate withdrawal (4Heinrichs S.C Menzaghi F Schulteis G Koob G.F Stinus L Behav. Pharmacol. 1995; 6: 74-80Crossref PubMed Scopus (182) Google Scholar). Thus, CRF activation may be a common element associated with the development of drug dependence and may be associated with motivational effects involving such subjective symptoms as increased stress and negative affect. It is tempting to speculate that the neuroadaptations outlined above actually involve specific brain areas that interface classical limbic (emotional) structures with the extrapyramidal motor system. The shell of the nucleus accumbens and the central nucleus of the amygdala share certain cytoarchitectural similarities that have led to the concept of the extended amygdala, a brain system in the basal forebrain that may be involved in emotional behavior and motivation (3Heimer, L., Alheid, G. (1991). In The Basal Forebrain: Anatomy to Function, T.C. Napier, P. Kalivas, and I. Hanin, eds. (New York: Plenum Press), pp. 1–42.Google Scholar). Particularly interesting is that the extended amygdala has significant efferents from limbic structures such as the basolateral amygdala, frontal cortices, and hippocampus and sends efferents not only to the medial part of the ventral pallidum but also a large projection to the lateral hypothalamus. Furthermore, the shell of the nucleus accumbens is particularly sensitive to the cocaine antagonist activity of DA D1 antagonists (1Caine S.B Heinrichs S.C Williams C.E Coffin V.L Koob G.F Brain Res. 1995; 692: 47-56Crossref PubMed Scopus (235) Google Scholar). The central nucleus of the amygdala is important for the GABA antagonist effects on ethanol self-administration (6Hyttia, P., and Koob, G.F. (1995). Eur. J. Pharmacol. 283, 151–159.Google Scholar) and the CRF interaction with ethanol (10Koob G.F Heinrichs S.C Menzaghi F Pich E.M Britton K.T Semin. Neurosci. 1994; 7: 221-229Crossref Scopus (133) Google Scholar, 17Pich E.M Lorang M Yaganeh M DeFonseca F.F Koob G.F Weiss F J. Neurosci. 1995; 15: 5439-5447PubMed Google Scholar). It is possible that the neurochemical components that comprise the neurocircuitry of the extended amygdala may be the long sought after reward systems that subserve not only the phenomenon of intracranial self-stimulation but also natural rewards. These reward systems may be usurped by chronic drug use, leading to dependence. Drug addiction is a chronic relapsing disorder, and relapse is associated with states of craving and protracted abstinence that are difficult to define but reflect some prolonged post-acute withdrawal perturbation or vulnerability to reinstatement of drug-taking behavior and ultimately compulsive use. There may be either a residual deficit state or a sensitization of the reward system to stimuli that predict drug effects or both. One could speculate that the combination would be particularly powerful as a motivator for reinstatement of drug use. Animal models of drug craving and relapse are being developed and refined and usually reflect secondary sources of reinforcement. The neural substrates for conditioned positive reinforcement may involve elements of afferent input of the extended amygdala such as the afferents from the basolateral amygdala and the mesolimbic DA system. The neural substrates for conditioned negative reinforcement are largely unknown. The challenge for future studies in the neurobiology of drug dependence will be to elucidate the neuroadaptive changes, such as changes in signal transduction function and gene expression, produced by chronic drug use in animal models of protracted abstinence and relapse. Presumably, the same molecular and cellular changes in the neurochemical systems and neurocircuitry responsible for the positive and negative reinforcement associated with chronic drug use will hold the key.