MECHANISMS OF ACTION OF ADDICTIVE STIMULI Incentive-sensitization and addiction

The question of addiction concerns the process by which drug-taking behavior, in certain individuals, evolves into compulsive patterns of drug-seeking and drug-taking behavior that take place at the expense of most other activities, and the inability to cease drug-taking, that is, the problem of relapse. In this paper we summarize one view of this process, the “incentive-sensitization” view, which we Ž rst proposed in 1993. Four major tenets of the incentive-sensitization view are discussed. These are: (1) potentially addictive drugs share the ability to alter brain organization; (2) the brain systems that are altered include those normally involved in the process of incentive motivation and reward; (3) the critical neuroadaptations for addiction render these brain reward systems hypersensitive (“sensitized”) to drugs and drug-associated stimuli; and (4) the brain systems that are sensitized do not mediate the pleasurable or euphoric effects of drugs (drug “liking”), but instead they mediate a subcomponent of reward we have termed incentive salience (drug “wanting”). Introduction Most contemporary explanations of addiction posit that addicts are motivated to take drugs primarily for one of two reasons, by “the desire to experience the positive hedonic effects of the drug ... and the desire to avoid aversive withdrawal symptoms ...” (Markou et al., 1993, p. 176). In other words, it is generally thought that addicts are motivated to take drugs either for the pleasure drugs produce (basically to achieve remembered pleasure), or to avoid the unpleasant consequences of withdrawal. We have argued, however, that the compulsive drug-seeking and drug-taking behavior that characterizes addiction often are not motivated by either the desire to obtain pleasure or by the desire to relieve withdrawal (see Robinson & Berridge, 1993 and 2000, for a critique of withdrawal avoidence and pleasure-seeking views of addiction). If this is true, then why do addicts compulsively seek drugs? We have attempted to address this question by proposing the concept of “incentive-sensitization” (Robinson & Berridge, 1993, 2000; Berridge & Robinson, 1995), which can be summarized in four points. (1) Potentially addictive drugs share the ability This paper is an abbreviated version of ‘The Psychology and Neurobiology of Addiction: An Incentive-Sensitization View’, which was published earlier in a special supplement of Addiction. (Vol. 95, Supplement 2, 2000). Correspondence to: Dr Terry E. Robinson, Department of Psychology (Biopsychology Program), The University ofMichigan, 525 E. University (East Hall), Ann Arbor, MI 48109,USA. Tel: (734) 763 4361; fax: (734) 763 7480; e-mail: [email protected] Submitted 28th October 1999; initial review completed 17th December 1999; Ž nal version accepted 2nd June 2000. ISSN 0965–2140 print/ISSN 1360–0443 online/01/010103–12 Ó Society for the Study of Addiction to Alcohol and Other Drugs Carfax Publishing, Taylor & Francis Limited DOI: 10.1080/09652140020016996 104 Terry E. Robinson & Kent C. Berridge to produce long-lasting changes in brain organization. (2) The brain systems that are changed include those normally involved in the process of incentive motivation and reward. (3) The critical neuroadaptations for addiction render these brain reward systems hypersensitive (“sensitized”) to drugs and drug-associated stimuli. (4) The brain systems that are sensitized do not mediate the pleasurable or euphoric effects of drugs (drug “liking”), but instead they mediate a subcomponent of reward we have termed incentive salience or “wanting” (Berridge, Venier & Robinson, 1989; Berridge & Valenstein, 1991; Robinson & Berridge, 1993; Berridge & Robinson, 1995, 1998; Berridge, 1996). We posit the psychological process of incentive salience to be speciŽ cally responsible for instrumental drug-seeking and drug-taking behavior (drug “wanting”). We have hypothesized that when sensitized, this incentive salience process produces compulsive patterns of drug-seeking behavior (Robinson & Berridge, 1993; Berridge & Robinson, 1995). Through associative learning the enhanced incentive value becomes focused speciŽ cally on drug-related stimuli, leading to increasingly compulsive patterns of drug-seeking and drug-taking behavior. Furthermore, the persistence of neural sensitization is hypothesized to leave addicts susceptible to relapse even long after the discontinuation of drug use. In the following we will review brie y some of the evidence for incentive-sensitization, and elaborate some of the major features of this view of addiction. Psychomotor sensitization Most studies showing that the repeated administration of drugs of abuse can produce sensitization (i.e. an increase in drug effect) involve measures of the psychomotor activating effects of drugs, such as their ability to enhance locomotor activity, rotational behavior or stereotyped motor patterns (Segal, Geyer & Schuckit, 1981; Robinson & Becker, 1986; Robinson & Berridge, 1993; Stewart & Badiani, 1993). Studies on the psychomotor activating effects of drugs are thought to be relevant to addiction because of the assumption that the neural substrate that mediates these effects is either the same as, or at least overlaps with, the neural substrate responsible for the rewarding effects of drugs (Wise & Bozarth, 1987). This neural substrate is, of course, the mesotelencephalic dopamine system, and especially dopamine projections to the nucleus accumbens and accumbens-related circuitry (often called the mesolimbic or mesocorticolimbic dopamine system). There is now a wealth of evidence showing that the repeated intermittent administration of a variety of drugs of abuse results in a progressive increase in their psychomotor activating effects, and although most studies of psychomotor sensitization involve the administration of psychomotor stimulants, such as amphetamine or cocaine, psychomotor sensitization has been reported with methylphenidate, fencamfamine, morphine, phencyclidine, MDMA, nicotine and ethanol (Robinson, 1993; Robinson & Berridge, 2000; for references). Psychomotor sensitization is a very complex and rich phenomenon. For example, it is dose-dependent (Kalivas et al., 1988; Browman, Badiani & Robinson, 1998a, 1998b), it is usually seen only when drugs are administered intermittently (Post, 1980, Robinson & Becker, 1986), it is often more evident long after the discontinuation of repeated drug treatment than shortly after the discontinuation of drug treatment (Antelman, 1988), and perhaps the most remarkable feature of sensitization is its persistence. Once sensitized, animals may remain hypersensitive to the psychomotor activating effects of drugs for months or years (Robinson & Becker 1986; Paulson, Camp & Robinson, 1991). Finally, sensitization is seen not only following experimenter-administered drug, but drug self-administration experience can also induce psychomotor sensitization (Hooks et al., 1994; Phillips & Di Ciano, 1996; Marinelli, Le Moal & Piazza,1998). Two other important features of sensitization deserve mention. One is that there is enormous individual variation in susceptibility to sensitization (Robinson, 1988). Some individuals show rapid and robust sensitization with a given dose of a drug, whereas others sensitize very little, if at all. There are many factors that contribute to individual variation in the susceptibility to sensitization, including genetic, hormonal and experiential factors (Shuster, Yu & Bates, 1977; Antelman et al., 1980; Robinson, 1988), although how they do so is largely unknown. Incentive-sensitization and addiction 105 Nevertheless, the incentive-sensitization theory posits that factors which render people susceptible to sensitization will also contribute to individual variation in susceptibility to addiction. Another important feature of psychomotor sensitization is that it is not an inevitable consequence of repeated exposure to drugs. Instead, the ability of drugs to induce or express sensitization is powerfully modulated by learning and the circumstances surrounding drug administration (Robinson et al., 1998). There are at least two ways that the circumstances surrounding drug administration modulate sensitization. The Ž rst is modulation of the expression of neural sensitization that has already been induced. Perhaps the best example of environmental modulation of expression is the phenomenon of context-speciŽ c sensitization. This refers to the observation that if animals are tested (i.e. receive a drug challenge) in an environment different from the one in which they received prior drug treatments, sensitization is often not expressed in behavior (Post et al., 1981; Pert, Post & Weiss, 1990; Anagnostaras & Robinson, 1996; Terelli & Terry, 1999). Despite this powerful conditioned stimulus control over the expression of behavioral sensitization there are at least two reasons to believe that in this situation neural sensitization has developed, even though animals do not express it in behavior. The Ž rst is that animals receiving drug treatments in an environment other than the test environment develop normal behavioral sensitization in their drug treatment environment; they simply do not express it in a different environment that has never been paired with drug administration (Anagnostaras & Robinson, 1996). Secondly, neural sensitization has been described under conditions that preclude the in uence of contextual stimuli on the neurobiological expression of the drug response, for example, in striatal tissue slices in vitro or in anesthetized animals (Robinson & Becker, 1982; Castañeda, Becker & Robinson, 1988; Henry & White, 1991; Nestby, Vanderschuren & De Vries, 1997; Kantor, Hewlett & Gnegy, 1999; Vanderschuren et al., 1999a). It appears, therefore, that repeated exposure to amphetamine may induce neural sensitization non-associatively, but whether the consequences of neural sensitization are expressed at a particular place or time is determined to a large extent by conditional stimuli (especially contextual stimuli) that have been associatively paired with drug administration (Anagnostaras & Robinson, 1996). Indeed, it needs to be remembered that the ability of sensitized neural systems to gain control over behavior is constantly modulated or gated by environmental (and probably interoceptive) stimuli that have been associated with drug administration. It may be that this interaction of neural sensitization with associative learning is responsible for the focus on drug-associated stimuli in addicts, whereby the acts and objects associated with drug-taking become especially powerful incentives themselves. Contextual modulation of the expression of sensitization may contribute to the critical role that context plays in precipitating relapse. That is, an implication for addiction is that the expression of sensitization to the incentive properties of drugrelated stimuli may be strongest in contexts that have been also distinctly related to drug-taking in the past. The ability of context to act as an occasion-setter and to modulate sensitization would interact with the ability of speciŽ c drugassociated conditioned stimuli to trigger craving as a classically conditioned response, combining to provide very strong contextual control over both craving and relapse (Robinson & Berridge, 1993; Anagnostaras & Robinson, 1996; Robinson & Berridge, 2000). The second way in which the circumstances surrounding drug administration may modulate sensitization is to in uence whether neural sensitization is induced in the Ž rst place (or at least the rate and extent of sensitization produced by a given dose of a drug). For example, there are now a number of reports that when low to moderate doses of amphetamine or cocaine are administered in the environment where an animal lives (i.e. at “home”) they are less effective in inducing psychomotor sensitization than if the same doses are given in a relatively distinct test environment (one that is novel to the animal until its Ž rst pairing with the drug; Badiani, Anagnostaras & Robinson, 1995; Badiani, Browman & Robinson, 1995; Crombag, Badiani & Robinson, 1996; Badiani, Camp & Robinson, 1997; Browman, Badiani & Robinson, 1998a; Robinson et al., 1998; Fraioli et al., 1999). Further studies have established that the effect of environmental context is not to completely preclude sensitization, but to shift the dose-effect curve for the induction of sensitization. When high enough doses of either cocaine or amphetamine are given sensitization is induced re106 Terry E. Robinson & Kent C. Berridge gardless of environmental condition (Browman et al., 1998a, 1998b). The ability of environmental context to modulate the induction of sensitization may be related to its ability to modulate the neural circuitry engaged by drugs. Badiani and colleagues (1998) reported, for example, that the ability of amphetamine to induce c-fos mRNA in the striatum is modulated powerfully by the environmental context in which amphetamine is administered. Indeed, it appears that that environmental context can modulate which cell populations in the striatum are engaged by amphetamine. When given at home amphetamine induced c-fos only in striatal neurons also positive for dopamine D1 receptor mRNA (not in cells positive for D2 receptor mRNA). However, when given in association with environmental novelty amphetamine induced c-fos in both D1 and D2 mRNA-positive neurons (Badiani et al., 1999). In summary, sensitization is not an inevitable consequence of exposure to potentially addictive drugs. That is, it is not a simple pharmacological phenomenon. Both the expression and the induction of sensitization can be powerfully modulated by non-pharmacological factors, including environmental factors associated with drug administration. The in uence of environmental factors on sensitization has important implications not only for understanding the phenomenon, but for thinking about therapeutic approaches in treating addiction. Sensitization and drug reward The studies reviewed above on sensitization to the psychomotor activating effects of drugs indicate that addictive drugs induce neural sensitization; but by themselves they provide only indirect evidence that sensitization occurs to the incentive motivational or rewarding effects of drugs (Wise & Bozarth, 1987). More direct evidence that the neural substrate that is sensitized is involved in mediating drug reward comes from two other sources. The Ž rst are studies showing that not only do the psychomotor stimulant effects of drugs sensitize, but so do their rewarding effects (Schenk & Partridge, 1997). There are a number of reports that prior exposure to a variety of potentially addictive drugs enhances the later acquisition of both a drug self-administration habit (Woolverton, Goldberg & Ginos, 1984; Piazza et al., 1989, 1990; Horger, Shelton & Schenk, 1990; Horger, Giles & Schenk, 1992; Valadez & Schenk, 1994; Pierre & Vezina, 1997; Pierre & Vezina, 1998) or a conditioned place preference (Lett, 1989; Gaiardi et al., 1991; Shippenberg & Heidbreder, 1995; Shippenberg, Heidbreder & Lefevour, 1996; Shippenberg, Lefevour & Heidbreder, 1996). Previous sensitization to amphetamine also increases the “breakpoint” for amphetamine self-administration when rats are tested using a progressive ratio schedule (Mendrek, Blaha & Phillips, 1998; Lorrain, Arnold & Vezina, 2000), and the enhanced responding for a conditioned reward produced by intra-accumbens amphetamine is potentiated by cocaine sensitization (Taylor & Horger, 1999). Furthermore, sensitization to amphetamine facilitates behavior guided by Pavlovian learning (Harmer et al., 1997; Harmer & Phillips, 1998, 1999a, 1999b). Finally, in recent studies Deroche, Le Moal & Piazza (1999) have found that experience with self-administered cocaine later enhances the motivation to seek cocaine in, for example, a runway apparatus, and De Vries and colleagues in the Netherlands have reported that the ability of different drugs to reinstate (prime) drug-seeking behavior is related positively to whether they also show cross-psychomotor sensitization (De Vries et al., 1997, 1998, 1999; Vanderschuren et al., 1999b). The second line of evidence that the neural substrate sensitized by drugs of abuse is involved in mediating drug reward, comes from studies on the neurobiology of sensitization. There is not space here to review this large literature, but sufŽ ce to say there is now considerable evidence that behavioral sensitization is associated with neuroadaptations in dopamine and accumbensrelated circuitry (Robinson & Becker, 1986; Kalivas & Stewart, 1991; Robinson & Berridge, 1993; Stewart & Badiani, 1993; Pierce & Kalivas, 1997; White & Kalivas, 1998; Wolf, 1998). This is important because it is well established that these neural systems play an important role in mediating the rewarding effects of drugs and other incentives (Wise & Bozarth, 1987; Koob & Bloom, 1988; Smith, 1995). Thus, if sensitization-related neuroadaptations are found in this mesocorticolimbic circuitry this is strong evidence that at least one neural system known to be critical for mediating drug reward undergoes “neural sensitization”. Both preand post-synaptic neuroplastic adaptations have been described in the Incentive-sensitization and addiction 107 dopamine/accumbens system of sensitized animals. An example of a presynaptic adaptation is a persistent increase in the ability of a variety of drugs to increase the over ow of dopamine in the nucleus accumbens and striatum of sensitized animals, in vitro and in vivo (Robinson & Becker, 1982, 1986; Kalivas & Stewart, 1991; Robinson & Berridge, 1993; Nestby et al., 1997; Pierce & Kalivas, 1997; Kantor et al., 1999; Vanderschuren et al., 1999a). Examples of postsynaptic adaptations include an increase in the sensitivity of dopamine D1 receptors (Henry & White, 1991; White & Kalivas, 1998) and a decrease in the sensitivity of glutamate receptors (White et al., 1995) in the nucleus accumbens of sensitized animals (see Clark & Overton, 1998; Wolf, 1998 for a review of the role of excitatory amino acids in sensitization). More recently it has been reported that both amphetamine and cocaine sensitization are also accompanied by persistent structural modiŽ cations in the morphology of output neurons in both the nucleus accumbens and prefrontal cortex (Robinson & Kolb, 1997, 1999). Repeated treatment with amphetamine or cocaine increases the length of dendrites on medium spiny neurons in the nucleus accumbens and on pyramidal neurons in the prefrontal cortex. This is accompanied by an increase in spine density on the distal dendrites of these cells. On medium spiny neurons there is an especially large increase in the number of branched spines; that is, spines with multiple heads. Furthermore, cocaine self-administration experience has similar effects (Robinson et al., 1999). These data suggest that sensitization may involve changes in patterns of synaptic connectivity in brain reward systems, changes that may be similar to those seen in other neural systems in association with other forms of experiencedependent plasticity (Robinson & Kolb, 1997, 1999). To reiterate the basic thesis of the incentivesensitization view of addiction, it was originally proposed (Robinson & Berridge, 1993) that addictive drugs share the ability to produce persistent neuroadaptations in brain regions involved in the process of incentive motivation and reward, adaptations that render these regions hypersensitive (“sensitized”). It should be clear from the above that there is now a wealth of evidence to support this claim. The incentive-sensitization view also posits that it is largely because of sensitization of a neural substrate that mediates drug reward that with repeated drug use drugs gradually become more and more attractive (i.e. they acquire greater and greater incentive value), and become increasingly able to control behavior. Studies on sensitization of drug reward and the neurobiology of sensitization support this claim. Furthermore, we have suggested that sensitization enhances the probability of relapse, even long after the discontinuation of drug use, and animal studies on the relationship between psychomotor sensitization and reinstatement support this claim. Of course, the hypothesis that incentive-sensitization mediates addiction in humans is more speculative, and is predicated on the assumption that repeated exposure to drugs of abuse can induce neural sensitization in humans. It is one thing to demonstrate incentivesensitization in animals models, but—as critics of our theory occasionally point out—quite another to demonstrate its occurrence in addicts. Sensitization in humans As might be expected from the difŽ culty in studying this issue in humans, there has been very little research on the topic of whether sensitization actually occurs in the brains of human addicts. Until recently, the only direct evidence that repeated exposure to psychostimulant drugs can produce sensitization in humans came from studies on the phenomenology of amphetamine and cocaine psychosis (Post & Contel, 1983; Segal & Schuckit, 1983; Sato et al., 1983; Sato, 1986; Angrist, 1994). There is a considerable clinical literature which suggests that repeated exposure to amphetamine or cocaine results in a progressive increase in their psychotomimetic effects (Angrist, 1994), and that this enhanced sensitivity may persist long after the discontinuation of drug use (Utena, 1966; Sato et al., 1983; Sato, 1986). Related effects have been described in non-human primates (Castner & GoldmanRakic, 1999). More direct evidence for sensitization to the psychomotor effects of amphetamine in humans has been lacking until only very recently, but there are now two reports of psychomotor sensitization in humans. Strakowski et al. (1996) Ž rst reported the results of a double-blind, placebocontrolled study in drug-naive volunteers given two treatments (48 hours apart) with 0.25 mg/kg d-amphetamine. They found that the second treatment with amphetamine elicited a 108 Terry E. Robinson & Kent C. Berridge signiŽ cantly greater increase than the Ž rst in four behavioral measures: activity/energy, mood, rate and amount of speech and eye-blink rate. In a second study Strakowski & Sax (1998) replicated and extended these Ž ndings to see if three treatments with amphetamine would produce a progressive increase in drug effect, as is usually seen in animal experiments. Two measures increased progressively with repeated amphetamine treatment: activity/energy and eye-blink rate. Indeed, for eye-blink rate there was no effect of the Ž rst treatment with amphetamine, relative to placebo, but an increase in eye-blink rate emerged with subsequent drug treatments even though the dose was the same. Finally, evidence supporting the concept of incentive-sensitization in humans, relevant speciŽ cally to drug taking, comes from the interesting tentative observation of Bartlett et al. (1997) that cocaine users who developed sensitization to the psychotomimetic effects of the drug have an elevated incidence of relapse, as indicated by more frequent rehospitalizations. In summary, although there is little research in humans and it is fraught with technical limitations, the available evidence suggests that repeated exposure to psychostimulant drugs can sensitize some drug effects in humans. Further studies on behavioral sensitization in humans will be critical in testing the notion of incentivesensitization, but it is worth injecting a note of caution in interpreting negative behavioral studies. It is not obvious a priori which behavioral measures in humans will provide the most sensitive indicators of a sensitization process. This is a difŽ cult issue even in animal studies. For example, it is often difŽ cult to quantify behavioral sensitization using measures of locomotor activity, unless exactly the right dose and treatment conditions are used (Crombag et al., 1999). Also, even in rats, some behaviors show robust sensitization, such as rotational behavior, repetitive snifŽ ng and repetitive head movements, whereas other seemingly related stereotyped behaviors do not, such as oral movements (Robinson & Becker, 1986; Crombag et al., 1999). Finally, one needs to keep in mind that for the most part indices of behavioral sensitization are important only as secondary measures because they provide indicators of underlying neuroadaptive processes (neural sensitization), and it is neural sensitization that we posit to be crucial to addiction. The incentive-sensitization hypothesis makes strong predictions regarding neural sensitization and drug-seeking in human addicts, but not necessarily about what which speciŽ c observable other behaviors might best re ect neural sensitization. The critical prediction made by the incentive-sensitization view of addiction is this: the brains of human addicts who compulsively crave drugs will contain a neural substrate that has been rendered sensitized by drugs. A role of that neural substrate will be to mediate the incentive salience of drug rewards. Further, individuals will differ in their susceptibility for sensitization of that neural substrate, and those who sensitize most readily will be most at risk for addiction. These predictions are testable, and so the incentive sensitization theory of addiction can be conŽ rmed or disproved on the basis of