D1 dopamine receptor: A putative neurochemical and behavioral link to cocaine action


  • Michele Hummel,

    Corresponding author
    1. Department of Pharmacology, Temple University School of Medicine, Philadelphia, Pennsylvania
    • Department of Pharmacology, Temple University School of Medicine, Philadelphia, PA 19140.
    Search for more papers by this author
  • Ellen M. Unterwald

    1. Department of Pharmacology, Temple University School of Medicine, Philadelphia, Pennsylvania
    2. Laboratory of the Biology of Addictive Diseases, The Rockefeller University, New York, New York
    Search for more papers by this author


Cocaine is one of the most abused psychostimulants known to man and as such, researchers have been steadfast in their attempts to understand the neurobiological mechanisms responsible for its abuse. Cocaine undoubtedly wreaks havoc on a number of mammalian neuronal neurotransmitter systems, and it is maintained that this dysregulatory effect supports cocaine abuse. Cocaine's mechanism of action has been well described. Cocaine binds differentially to the dopamine, serotonin, and norepinephrine transport proteins and directly prevents the re-uptake of dopamine, serotonin, and norepinephrine into pre-synaptic neurons (Heikkila et al., 1975, Biochem Pharmacol 24(8):847–852; Reith et al., 1986, Biochem Pharmacol 35(7):1123–1129; Ritz et al., 1987, Science 237:1219–1223). Inhibition of re-uptake subsequently elevates the synaptic concentrations of each of these neurotransmitters. In addition to this direct effect, cocaine also produces a number of indirect actions, which alter other neuromodulatory systems (i.e., opioidergic, glutamatergic, and GABAergic systems). Many of these effects are just beginning to be elucidated, but nonetheless contribute to this agent's diverse pharmacological profile. Interestingly, it is the indirect actions of this mellifluous molecule, which mediate most of its sought and unsought effects. The intricacy with which cocaine produces neuronal alterations beyond its direct effects on neurotransmitter re-uptake appear to be most relevant to cocaine abuse, and hence the phenomenon of addiction. In light of cocaine's multifarious effects on numerous neuronal systems, its effect on dopaminergic neurotransmission has attracted the most attention, particularly because of the implicated role of dopamine in brain reward. Pharmacologically, the psychostimulant effects of cocaine appear to be mediated by its ability to enhance dopaminergic activity within the mesocorticolimbic circuit (Roberts et al., 1977, Pharmacol Biochem Behav 6(6):615–620). Additionally, it is the intensity with which cocaine produces alterations in dopaminergic circuitry that have enabled this drug to prevail as one of the most addictive substances known to man. This review will summarize findings relevant to cocaine-induced alterations in dopamine-mediated signal transduction. Specifically, it will concentrate on the D1 dopamine receptor and intracellular signaling mediated by this receptor subtype. It will describe cocaine-induced cellular and behavioral alterations relevant to this pathway and how these changes potentially effect gene transcription and protein expression. This article too will review a common behavioral manifestation associated with repeated cocaine exposure, sensitization, and why the D1 dopamine receptor and its associated signaling pathway have been implicated in this phenomenon. Lastly, this article will discuss how targeting the D1 dopamine receptor and its signaling pathway may offer some insight into understanding cocaine addiction, a somewhat elusive brain disease. J. Cell. Physiol. 191: 17–27, 2002. © 2002 Wiley-Liss, Inc.


Dopamine is one the best described neurotransmitters within the mammalian nervous system. Among all the central nervous system (CNS) neurotransmitters and neuromodulators described, dopamine has occupied a central position in brain reward theory, and hence, drug abuse research. It has been shown to be an essential contributor not only of drug reward, but also of brain stimulation reward and reward from natural reinforcers like food (Wise and Rompre, 1989). Dopamine's alleged role in drug reward mechanisms has prompted the dopamine hypothesis addiction (Wise and Bozarth, 1987), and has essentially intensified research linking dopamine to substance use and abuse. This link has been supported by the observation that virtually all addictive substances increase extracellular dopamine and ultimately facilitate dopaminergic neurotransmission within the CNS (DiChiara and Imperato, 1988; Koob, 1992). Cocaine is not without exception to this observation. A number of neuropharmacological studies have established an important role for dopamine in the reinforcing effects of cocaine (Woolverton and Johnson, 1992). Cocaine enhances dopamine-mediated neurotransmission by elevating synaptic levels of dopamine within the mesocorticolimbic pathway, the major neuroanatomical circuit mediating drug reward (Heikkila et al., 1975; Koob, 1992). This pathway originates from dopaminergic-containing cell bodies in the ventral tegmental area (VTA) and projects to a number of forebrain areas including the frontal cortex, nucleus accumbens, caudate putamen, olfactory tubercle, hippocampus, and amygdala (Fallon and Moore, 1978). It is maintained that most abused substances modulate this circuit in one way or another, as this circuit is the final common pathway for reward and reinforcement. It is suggested, however, that the nucleus accumbens (i.e., ventral striatum) is the most common substrate within this pathway identifiable with the reward and reinforcement of addictive substances (Koob, 1992).

The striatum (nucleus accumbens and caudate putamen) is the major neuroanatomical target for midbrain dopaminergic neurons. Compared with other brain regions, the striatum is markedly homogeneous with most neurons displaying a similar morphology (DeLong, 2000). Greater than 90% of the striatal neuronal population are medium sized GABAergic cells possessing numerous dendritic spines with many synaptic contacts (Freund et al., 1984). Striatal spiny neurons project from the striatum forming either the direct or indirect basal ganglia output circuit. Half of these striatal neurons project via the direct pathway to the internal segment of the globus pallidus and substantia nigra pars reticulata. Neurons of the indirect pathway project to the internal segment of the globus pallidus via the external segment of the globus pallidus and the subthalmic nucleus. Together these projections are subsequently relayed through the thalamus to a number of cortical areas (Albin et al., 1989).

Dopamine neurotransmission within these two efferent pathways is mediated by two types of dopamine receptors, D1 class and D2 class. There are two classes of dopamine receptors, the D1 and the D2. The D1 class includes both D1 and D5 receptors, while the D2 class includes D2, D3, and D4 receptors (Sibley and Monsma, 1992). In terms of this discussion, we will focus on D1 and D2 receptors in view of their high density in the striatum and their relevance to the pharmacology of cocaine. D1 and D2 dopamine receptors are differentially distributed within the striatum, whereby D1 dopamine receptors are predominantly expressed within striatonigral neurons comprising the direct pathway and D2 dopamine receptors are located primarily within striatopallidal neurons in the indirect pathway (Gerfen et al., 1990) (see Fig. 1). There are also D2 autoreceptors on dopamine nerve terminals themselves. D1 and D2 receptor distribution, however, is not exclusive. Some neurons express both types of receptors (Surmeier et al., 1996). Nonetheless, D1 and D2 receptors have been implicated in mediating many of the actions of cocaine. It is important to note that many of these receptor-mediated effects can be clearly dissociated while others cannot. This suggests a modulatory and/or interactive role for these two receptors in cocaine-potentiated dopamine neurotransmission. For the present discussion, only those events relevant to cocaine's effects on the D1 dopamine receptor and its associated signal transduction pathway will be described.

Figure 1.

Simplified anatomical localization of striatal circuits important in the pharmacology of cocaine. Multiple neurotransmitter pathways are depicted. Emphasis is placed on dopamine and the D1 striatonigral pathway as these systems mediate many of the neuropharmacological actions of cocaine. Abbreviations in illustration: GPe, globus pallidus-external; GPi, globus pallidus-internal; SNc, substantia nigra pars compacta; SNr, substantia nigra pars reticulata; sn, subthalamic nucleus; VTA, ventral tegmental area. (Adapted from Koob, 1992; Delong, 2000.)


The role of D1 dopamine receptors in cocaine-induced dopamine signaling has been demonstrated both cellularly and behaviorally. For example, a number of electrophysiological studies have identified alterations in dopamine neurotransmission at the receptor level within the mesoaccumbal neurons following chronic cocaine exposure. Henry et al. (1989) demonstrated using single cell electrophysiological recording that twice daily injections of cocaine in rats for 14 days increased the sensitivity of post-synaptic dopamine receptors in the nucleus accumbens to the inhibitory effects of dopamine. They also found that this treatment decreased the sensitivity of D2 autoreceptors in the VTA, thereby facilitating dopamine neurotransmission via increased neuronal firing (Henry et al., 1989). In an attempt to further characterize these changes, they later demonstrated using the same treatment paradigm that nucleus accumbens neurons are sensitive to the inhibitory effects of the D1 agonist, SKF 38393, while being insensitive to the D2 agonist, quinpirole. These data provide evidence that D1 dopamine receptors mediate cocaine action. Additionally, it has been suggested that this change in D1 dopamine receptor sensitivity may account for some behavioral changes observed in rodents following chronic cocaine exposure, particularly the phenomenon of behavioral sensitization (Henry et al., 1989; Henry and White, 1991).

In addition to this pharmacological evidence, functional evidence also exists for a role for the D1 dopamine receptor in cocaine-induced neuronal signaling. As aforementioned, dopamine inhibits the firing of nucleus accumbens neurons and cocaine potentiates this inhibition by indirectly activating D1 receptors. Xu et al. (1994) use a gene targeting approach in combination with an electrophysiological paradigm to provide functional support for D1 receptor involvement in cocaine-induced neuronal signaling. In the study, cocaine was microiontophoretically applied to neurons in the nucleus accumbens of D1 receptor null mice and their wild-type littermates. Glutamate-stimulated neuronal activity was then recorded. Cocaine inhibited the generation of action potentials in wild-type mice, whereas its effect was markedly reduced in the D1 receptor null mutants (Xu et al., 1994). This study clearly substantiates the D1 receptor as a neurophysiological target in the overall pharmacology of cocaine.

As described previously, the striatum contains many GABAergic medium spiny neurons that express D1 dopamine receptors. Some of these neurons project to the VTA forming inhibitory synaptic contacts (Johnson and North, 1992). Under normal physiological conditions, dopamine acts via D1 receptors on the terminals of these afferent GABA neurons to facilitate GABA-mediated inhibition of midbrain VTA neurons (Cameron and Williams, 1993). Chronic cocaine exposure, however, modulates this normal physiological state. Chronic cocaine attenuates this GABA-mediated inhibition, thereby, dis-inhibiting midbrain dopamine-containing cells. This dis-inhibition promotes neuronal firing. Bonci and Williams (1996) have demonstrated that a once daily injection of cocaine in guinea pigs for 14 days attenuates dopamine-regulated GABA release. They also demonstrated that this was a D1 receptor-mediated event. When they applied the D1 dopamine receptor agonists, SKF82958 and SKF38393, to brain slices obtained from these animals, they found a marked reduction in the ability of these neurons to generate inhibitory post-synaptic potentials. These electrophysiological studies demonstrate a role for the D1 dopamine receptor in cocaine's action.


For decades, scientists have attempted to understand the neurobiological mechanism(s) by which cocaine produces such intense reward and how its repeated use perpetutates both short- and long-term behavioral changes, which may ultimately culminate in cocaine addiction. Experimentally, researchers have been able to delineate many putative cellular substrates responsible for cocaine's behavioral effects. The relevance of the D1 dopamine receptor as a specific cellular target mediating the behavioral effects of cocaine exposure has been demonstrated. This does not go without further disclosing the importance of D2 dopamine receptors in this regard. However, our greatest advances in cocaine research seem to have come about with a better understanding of the D1 dopamine receptor and its associated signaling pathway. Some of the experimental paradigms directed at further elucidating a role for the D1 receptor in cocaine-induced behaviors include studies on the discriminative stimulus, reinforcing, and locomotor-activating effects of cocaine.

The discriminative stimulus effects of drugs of abuse in animals are maintained to model those subjective effects often experienced by humans; i.e., the discriminative stimulus effects in animals help predict internal mechanisms of reward and well-being, which perpetuate man's repeated administration of a given drug. Experimental studies have found that cocaine functions as a discriminative stimulus in animals (Woolverton and Trost, 1978). Additionally, studies have confirmed a prominent role for dopamine (McKenna and Ho, 1980) and a supportive role for D1 dopamine receptors in mediating the discriminative stimulus effect of cocaine. For example, a number of investigations have found that pretreatment with D1 receptor antagonists attenuate the discriminative stimulus effects of cocaine (Kleven et al., 1988, 1990a; Barrett and Appel, 1989; Callahan et al., 1991; Sinnott and Nader, 2001).

In addition to the work with dopamine receptor antagonists, a number of studies sought to further characterize the role of D1 dopamine receptors in the discriminative stimulus effect of cocaine using dopamine receptor agonists. These findings, however, are somewhat more equivocal and are not always in agreement with those obtained with the receptor antagonists. Therefore, in keeping with the context of this discussion, one must consider a number of experimental variables in such studies like animal species, type of discriminative stimulus paradigm (substitution or interaction), and training dose of cocaine employed. These factors may partially account for some of the anomalous findings in the literature. For example, in substitution studies in rats and rhesus monkeys trained to discriminate cocaine from saline, the stimulus effects of cocaine failed to generalize to the D1 dopamine receptor agonist, SKF 38393 (Barrett and Appel, 1989; Kleven et al., 1990a; Callahan et al., 1991). Although the data are disparate, these findings with D1 receptor agonists and antagonists may be a reflection of the sensitivity of this behavioral measure to changes in cocaine-induced behavior mediated by a receptor mechanism. For example, the intrinsic efficacy of a D1 receptor ligand may impact the manifestation of this cocaine-influenced behavior, whereby, high efficacy agents function as discriminative stimuli and low efficacy agents fail to elicit such a response (see discussion below on intrinsic efficacy). Additionally, these findings may suggest that stimulation of the D1 receptor alone is insufficient to mimic cocaine-like behavior, but is a necessary requirement for cocaine's effects in drug discrimination paradigms.

With regard to the D1 receptor, the intrinsic efficacy of a D1 receptor agonist is measured by the ability of the compound to stimulate cAMP production in rat striatal membrane preparations (Weed and Woolverton, 1995). A number of studies within the last several years have maintained that the intrinsic efficacy of a D1 receptor agonist may be an important determinant in the assessment of the role of this receptor in cocaine-mediated behaviors. More specifically, some investigators have suggested that high efficacy D1 receptor agonists produce behavioral effects that mimic those of cocaine, while low efficacy agonists produce effects similar to that of D1 receptor antagonists (Spealman et al., 1997; Sinnott and Nader, 2001). For example, in a substitution study in rhesus monkeys, the high efficacy agonist SKF 81297 substitutes for cocaine, while the low efficacy agonist SKF 38393 does not produce such a response (Sinnott and Nader, 2001). These results lend credibility to the aforementioned notion of agonist efficacy being an important variable in discriminative stimulus studies. Additionally, these results support work discussed earlier on SKF 38393 in this regard (Barrett and Appel, 1989; Kleven et al., 1990a; Callahan et al., 1991). However, given the notion that low efficacy agonists may produce behavioral responses similar to D1 antagonists, Sinnott and Nader (2001) show in an interaction study with SKF 38393 and SKF 81297 that pretreatment with either ligand does not alter cocaine's discriminative stimulus effect in rhesus monkeys. These data are in agreement with earlier work demonstrating that SKF 38393 does not affect cocaine's discriminative stimulus effect in interaction studies in squirrel monkeys (Katz et al., 1999). Both of these findings, however, contrast work done by Spealman et al. (1997) who showed that low efficacy D1 agonists (SKF 38393, SKF 756201) attenuate the discriminative stimulus effects of cocaine in squirrel monkeys, while high efficacy agonists like SKF 81297 accentuate this behavioral effect. Taken together, these data show that a number of experimental variables may affect the sensitivity of the discriminative stimulus response to cocaine, whereby D1 receptor agonist efficacy may be more important in one type of study vs. another (i.e., substitution vs. interaction). Also, although these data do not undeniably confirm the nature of the role of the D1 receptor in this specific cocaine-induced behavior, collectively these data do suggest a supportive role.

Like the discriminative stimulus paradigm, the self-administration paradigm too has been employed experimentally to discern the neurobiological basis of cocaine-induced behavior. Self-administration studies in animal subjects were first described in the early 60's and have since been routinely used to assess the reinforcing properties of many abused substances including cocaine. Self-administration studies have identified cocaine as an extremely powerful reinforcer. Additionally, a substantial body of evidence has linked cocaine's reinforcing properties to its ability to increase dopaminergic neurotransmission within the mesocorticolimbic circuit (Goeders and Smith, 1983). Studies with dopamine receptor agonists and antagonists have demonstrated involvement of both D1 and D2 dopamine receptor subtypes in the acute reinforcing effects of cocaine (Bergman et al., 1990). Interestingly, researchers once suggested that D2 receptor stimulation played a more prominent role in cocaine reinforcement. Now, researchers are not so emphatic about this interpretation. Some seem to implicate a more functional role for the D1 receptor in cocaine-manifested reward, as described below. In addition, the possible utility of D1 receptor agonist intervention in attenuating cocaine seeking behavior has been suggested (Self et al., 1996). The truth, however, probably lies somewhere in between, where the concerted action of cocaine on both receptor systems summates to underlie the complex behavioral repertoire characteristic of cocaine addiction.

A number of studies have examined the modulation of cocaine reinforcement in animal subjects by dopamine receptor antagonists. Reinforcement paradigms employing dopamine receptor antagonists have demonstrated increases, decreases, or no change in cocaine self-administration following dopamine receptor antagonist pretreatment (Maldonado et al., 1993; Caine and Koob, 1994; Campbell et al., 1999). Many of the disparate findings can be attributed to differences in experimental design such as schedule of reinforcement (Caine and Koob, 1994). Until recently, D1 antagonist studies were directed at understanding the role of D1 receptors in cocaine reinforcement at mesocorticolimbic dopamine terminal sites (Maldonado et al., 1993; McGregor and Roberts, 1993, 1995). Recent evidence, however, suggests that D1 receptors in the VTA may mediate cocaine reinforcement. In addition to increasing extracellular dopamine at axon terminal projection sites, cocaine also increases extracellular dendritic dopamine. Dendritic dopamine then interacts with D1 receptors located on GABAergic and glutamatergic afferents projecting to the VTA. Ranaldi and Wise (2001) examined the role of cocaine-induced dendritic released dopamine in this regard. In a reinforcement paradigm, where rats were trained to self-administer cocaine in a fixed or progressive ratio schedule, bilateral VTA injections of the D1 antagonist SCH 23390 were shown to decrease cocaine reward. These findings are the first to suggest a functional role for dendritically released dopamine and VTA D1 receptors in cocaine reward (Ranaldi and Wise, 2001).

The role of the D1 receptor in cocaine reinforcement was also investigated using dopamine receptor agonists. As discussed previously, agonist intrinsic efficacy may be an important determinant in assessing dopamine receptor involvement in cocaine-mediate behaviors. This concept has been further validated in a study, where the reinforcing effects of six D1 receptor agonists (SKF 81297, SKF 82958, R (+) 6-Br-APB, SKF 38393, SKF 77434, S (−) 6-Br-APB) of varying intrinsic efficacy were examined. In a self-administration study in rhesus monkeys maintained by baseline cocaine, three of the six monkeys self-administered the high efficacy D1 agonists SKF 81297, SKF 82958, R (+) 6-Br-APB, whereas no monkey self-administered the low efficacy agents SKF 38393, SKF 77434, and S (−) 6-Br-APB (Weed and Woolverton, 1995). These findings substantiate previous work indicating D1 dopamine receptor involvement in cocaine reinforcement (Weed et al., 1993). Additionally, they suggest that intrinsic efficacy may affect a ligand's ability to reinforce behavior. These findings are in partial agreement with the finding that rats will self-administer the D1 agonists SKF 82958 and SKF 77434 (Self and Stein, 1992). As indicated previously, species differences can affect experimental outcome. Also, in vitro measures of agonist efficacy differ from species to species. Therefore, a ligand that functions as a high efficacy agonist in one species may function as a low efficacy agonist in another species or vice versa (Weed and Woolverton, 1995). These data suggest such variablity. Nonetheless, both studies implicate the D1 receptor in cocaine reinforcement.

Cocaine is an extremely powerful reinforcer, which perpetuates drug-seeking behavior in individuals who have chronically been exposed to it. The ability of cocaine to sustain drug-seeking behavior in drug-exposed individuals is maintained to be indicative of an addictive state. A modified self-administration paradigm known as reinstatement is regarded as a measure of drug-seeking behavior. The role of dopamine receptors in reinstatement to cocaine self-administration has been investigated. In one study, rats were trained to self-administer cocaine, then subjected to a 4-h reinstatement procedure, whereby cocaine responding was extinguished via saline substitution. The role of D1 and D2 receptor involvement in reinstatement to cocaine self-administration was subsequently evaluated by priming rats with either the D2-like agonists, 7-OH-DPAT or quinpirole, or the D1-like agent SKF 82958. Both of the D2 agonists significantly increased drug-associated responding, whereas the D1 agonist produced no such effect. Furthermore, pre-treatment with the D1 receptor agonist attenuated the ability of a cocaine prime to reinstate drug-seeking behavior (Self et al., 1996). These results suggest an important role for D1 and D2 receptors in cocaine-induced behaviors, whereby cocaine-induced dopamine stimulation of the D2 receptor may represent a mechanism to sustain drug-seeking behavior. Conversely, stimulation of the D1 receptor may inhibit such behavior.

Independent of cocaine's ability to perpetuate drug-seeking behavior or craving, drug-related environmental stimuli may serve as cues which elicit such behavioral responses. A recent investigation examined the association between cocaine's pharmacological actions and distinct stimuli within the environment. The investigation also sought to delineate the role of the D1 receptor as a neuropharmacological substrate mediating such behavior. In a reinstatement paradigm similar to the one described above, rats were trained to associate two discrete discriminative stimuli with either cocaine availability or saline availability. The cocaine-associated discriminative stimulus elicited a significant increase in responding or drug-seeking behavior, whereas the saline-associated discriminative stimulus produced no such response. Pretreatment with either of the D1 receptor antagonists, SCH 39166 and SCH 23390, attenuated the behavioral responses produced by the cocaine-associated discriminative stimulus. Additionally, further investigation found that these behaviors persisted 4 months following extinction (Ciccocioppo et al., 2001). Together, these data demonstrate that discriminative stimuli associated with cocaine availability can evoke drug-seeking behavior in rats and that cocaine-associated cues are undaunting. Furthermore, these results suggest a role for a D1 receptor-dependent mechanism in cocaine-seeking behavior elicited by distinct environmental cues.

Dopamine is important in modulating both motivational and behavioral reactivity (Delfs et al., 1990; Salamone, 1996) and as described, cocaine potentiates dopaminergic neurotransmission within the mesocorticolimbic dopamine circuit. This circuit has been directly implicated not only in reinforcement and reward, but also in the control of locomotor activity (Pijnenburg et al., 1976). Therefore, it is maintained that the ability of cocaine to accentuate locomotor behavior in animals is dependent upon its mechanism of indirectly potentiating dopaminergic function within mesolimbic areas such as the nucleus accumbens (Kelly et al., 1975; Kelly and Iversen, 1976; Delfs et al., 1990). Furthermore, the physiological actions of dopamine are mediated by D1 and D2 dopamine receptors, which have been implicated as specific targets in cocaine-induced locomotion. A number of studies have shown that D1 and D2 dopamine receptors are differentially involved in cocaine-induced changes in locomotor activity. Some maintain that the D1 receptor may be somewhat more important in this regard (McCreary and Marsden, 1993; Xu et al., 1994; Ushijima et al., 1995). For example, mice pre-treated with the D1 receptor antagonist, SCH 23390, showed an attenuated locomotor response to cocaine (Cabib et al., 1991). These findings were then later supported by studies in rats. Pre-treatment with SCH 23390 blocked cocaine-induced increases in locomotor activity in rats that were repeatedly administered cocaine, whereas the D2 receptor antagonists, haloperidol, and raclopride had no such effect (McCreary and Marsden, 1993; Ushijima et al., 1995). Interestingly, pre-treatment with the D2 receptor agonists, quinpirole and bromocriptine caused cocaine-induced locomotor activity to be replaced with stereotyped behaviors such as sniffing, licking, and gnawing (Ushijima et al., 1995). Taken together, these data demonstrate a functional role for the D1 receptor in cocaine-induced locomotion. They also suggest differing roles for the two receptors. The D1 receptor may be more important in cocaine-induced locomotor behavior, whereas the D2 receptor's role may be more relevant to cocaine-induced stereotyped behaviors.


Behavioral link

The repeated, intermittent administration of cocaine can result in the progressive and enduring enhancement of a number of behavioral and neurochemical responses to later drug challenge. This phenomenon is termed sensitization and is documented for many substances of abuse, not just cocaine (Vanderschuren and Kalivas, 2000). Moreover, sensitization is thought to underlie certain aspects of drug abuse, which may ultimately model certain aspects of human addictive behavior, such as drug craving (Robinson and Berridge, 2001). Although this phenomenon is well described in the literature, the neurobiological basis for its occurrence remains only partially understood. A substantial body of evidence suggests that the mesolimbic dopamine system is an important neural substrate critical for the development of sensitization (for reviews, see Kalivas and Stewart, 1991; Vanderschuren and Kalivas, 2000; Robinson and Berridge, 2001), and more specifically that cocaine-induced sensitization results from a number of different neuronal alterations, particularly long-term changes in dopamine D1 receptor-mediated neurotransmission. Given the intensity of the effect of cocaine on dopamine-mediated neurotransmission described, thus far, it is not surprising that this transmitter's receptor system has been implicated in such a role. Evidence now suggests that multiple neurotransmitter systems play a role in both the development and expression of sensitization. However, this subsection will briefly comment on the functional relevance of the D1 receptor and its associated signaling system to this drug-induced phenomenon.

Cocaine-induced sensitization in animal models is often described by behavioral measures, which include enhanced locomotor activity, stereotypy, and rotational behavior. Sensitization is very much dependent on variables such as drug administration paradigms and environmental context (Post et al., 1981). Therefore, care must be taken when surveying the data as differences in these parameters are likely to account for many of the inconsistencies found in the literature. Nonetheless, collectively the data suggest that many neural underpinnings play a role in this complex phenomenon.

Despite its complexity, cocaine-induced sensitization is undoubtedly most clearly visualized as a physical phenomenon, whereby rodents enhance their locomotor response upon repeated cocaine administration. The D1 dopamine receptor has been implicated as a potential neural mediator in the expression of cocaine-induced increases in locomotor activity and the development of sensitization. For example, pre-treatment with the D1 receptor antagonist SCH 23390 antagonized the locomotor activating effects of repeated cocaine administration and prevented the development of sensitization. Pre-treatment with a D2 receptor antagonist had no such effect (McCreary and Marsden, 1993). These data suggest that sensitization involves activation of D1 receptors. Although the D2 receptor probably also plays a supportive role in this drug-induced phenomenon, the exact nature of the role of this receptor is less clear. Therefore, it can be surmised that the two receptor subtypes are probably differentially involved in sensitization.

The role of the D1 receptor in cocaine-induced locomotor activation was subsequently supported by a study in D1 receptor null mice. In this study, D1 receptor knockout mice and wild-type control mice were given twice daily injections of either saline or cocaine for seven days. Locomotor activity was recorded before and after the injections. Overall, the D1 receptor knockouts displayed an attenuated locomotor response to repeated cocaine administration compared to the wild-type controls. The D1 receptor knockouts receiving cocaine did demonstrate an increase in locomotor activity across testing days, however, their response was never greater than the activity recorded for their saline exposed counterparts (Xu et al., 2000) (see Fig. 2). Together, these data depict a contributory role for the D1 receptor in the expression of the sensitizing effects of repeated cocaine exposure on locomotor activity.

Figure 2.

The D1 receptor mediates cocaine-induced increases in locomotor activity. D1 receptor knockout mice exhibit reduced locomotor responses to repeated cocaine administration as compared with wild-type controls. Data shown are total horizontal locomotor activity counts for D1 receptor null mice and wild-type controls injected twice daily with saline or cocaine (20 mg/kg, i.p.) for 7 days. Pre- and post-treatment is the first 30 min of activity recorded prior to and after saline or cocaine injection, respectively. Data are represented as mean ± SEM, where n = 8 per group. D1 receptor null mice exhibited an attenuated response to cocaine compared to their wild-type littermates, P < 0.01. (Data from Xu et al., 2000. Reprinted with permission from Brain Research.)

Neurochemical link

Although the behavioral activating effects of agents like cocaine are readily visible, the neurobiological mechanisms responsible for these alterations in behavior are much less apparent. Due to its complexity, researchers have found it very cumbersome to correlate visible changes in behavior to alterations in neurochemistry. Nevertheless, some success has been achieved in this regard. Two studies employed both behavioral and electrophysiological techniques to make such a correlation. In the first study, Henry and White (1995) demonstrate how sensitization occurs after twice daily injections of cocaine for 14 days and how this sensitization prevents nucleus accumbens neurons from responding to systemically and locally administered cocaine. They too show how this drug-induced inhibition of neuronal activity can be reversed by the D1 receptor antagonist SCH 23390. In a second study, Li et al. (2000) further linked the D1 receptor to this phenonmenon. Pre-treatment with the D1 receptor agonist SKF 81297 reversed cocaine-induced sensitization and D1 receptor sensitivity in the nucleus accumbens of rats sensitized to cocaine (Li et al., 2000). Both of these studies suggest that the D1 receptor may be one potential mediator in the expression of cocaine-induced sensitization.

Despite increasing evidence that behavioral sensitization involves alterations in the sensitivity of post-synaptic D1 receptors, the mechanism(s) by which these changes occur are not fully understood. Some researchers suggest that receptor density may offer an explanation. However, most concur that changes in receptor density alone are probably insufficient to account for such a profound phenomenon (Pierce and Kalivas, 1997), but nonetheless may play a somewhat contributory role. Several studies have investigated the effects of repeated cocaine administration on dopamine receptor expression. Increases, decreases, or no changes in dopamine receptor number have been reported (e.g., Kleven et al., 1990; Mayfield et al., 1992; Kunko et al., 1998; Sousa et al., 1999). We report that cocaine administered three times daily increases D1 receptor binding in the nucleus accumbens, olfactory tubercle, and ventral pallidum of sensitized rats (Unterwald et al., 1994). Furthermore, a significant positive correlation was found between cocaine-induced locomotion and D1 receptor number in the nucleus accumbens and olfactory tubercle. A subsequent study demonstrated that D1 receptor density is increased after twice daily cocaine injections, but not if the same daily dose is given in a single injection (Unterwald et al., 2001) (see Fig. 3). These data indicate that the frequency of cocaine administration is an important variable in establishing the effects of chronic cocaine exposure on neurochemistry. These data too suggest a role for D1 receptors in behavioral sensitization after chronic cocaine administration. However, the exact nature of this role in sensitization is still uncertain. It is very likely that sensitization results from numerous alterations beyond the receptor level, such as changes in signal transduction.

Figure 3.

D1 receptor density is increased in rats sensitized to cocaine. Data are shown as percent of control D1 receptor density in selected brain regions from rats injected intrapritoneally with saline, cocaine once daily (30 mg/kg), cocaine twice daily (15 mg/kg/injection), or cocaine three times daily (15 mg/kg/injection) in a binge pattern. Data are represented as mean ± SEM, where n = 6 per group. *denotes significance, whereby, P < 0.05 as determined by a one-way analysis of variance (ANOVA) with Bonferroni post hoc analysis. FrCx, frontal cortex; NA, nucleus accumbens; CP, caudate putamen; OT, olfactory tubercle; VP, ventral pallidum. (Data from Unterwald et al., 1996, 2001. Reprinted with permission from Brain Research.)


Indeed both behavioral and neurochemical data support a role for the D1 receptor in cocaine-induced sensitization. Nevertheless, sensitization is a very complex phenomenon, which cannot be explained by changes in a single protein alone. A number of investigators have maintained that subsequent to its effect on D1 receptor sensitivity, cocaine produces much more complex alterations in neuronal signaling (Terwilliger et al., 1991; Miserendino and Nestler, 1995; Unterwald et al., 1996). Specifically, it has been suggested that repeated cocaine exposure upregulates the cAMP system (Self and Nestler, 1995), an intracellular neuronal signaling pathway that is stimulated by D1 receptor activation. It is the concerted changes in dopamine D1 receptor sensitivity and D1 receptor-mediated signaling, which are thought to support cocaine-induced alterations in behavior and the development of sensitization (Miserendino and Nestler, 1995; Unterwald et al., 1996).

Under normal physiological conditions, D1 receptor stimulation results in Gs-dependent activation of adenylyl cyclase and cAMP production. cAMP in turn activates cAMP dependent protein kinase (PKA), which phosphorylates a number of intracellular signaling proteins including membrane ion channels, enzymes, receptors, and transcription factors, thereby, facilitating changes in gene expression. A number of reports have found that chronic repeated cocaine exposure enhances signaling thru the cAMP pathway, ultimately leading to a plethora of behavioral and neurochemical sequaele characteristic of cocaine addiction. For example, increases in the activity of adenylyl cyclase and PKA following chronic cocaine administration have been reported (Terwilliger et al., 1991). Further studies too have shown how PKA is involved with the development and/or expression of cocaine-induced sensitization; rats pre-treated concurrently with intraperitoneal cocaine and intra-accumbal 8-Bromo-cAMP, a PKA activator, demonstrated greater increases in cocaine-induced locomotor activity compared to animals pre-treated with the PKA inhibitor, RP-CPT-cAMP (Miserendino and Nestler, 1995). These data lend support to the notion that repeated cocaine exposure modulates the cAMP pathway with resultant changes in behavior, particularly increases in locomotor activity.

Consistent with this research, we found that chronic repeated cocaine administration increases D1 receptor-mediated signal transduction. In our study, cocaine administered three times daily to rats for 14 days increases the ability of dopamine and the D1 selective receptor agonist, SKF 82958, to stimulate adenylyl cyclase activity in the nucleus accumbens and caudate putamen in these sensitized animals (Unterwald et al., 1996) (see Fig. 4). Collectively, these data suggest that sensitization may occur as a result of the effect of cocaine on D1 receptor-mediated signaling through the cAMP pathway.

Figure 4.

Chronic cocaine increases D1 dopamine receptor stimulated adenylyl cyclase activity. Dopamine- and SKF 82958-stimulated adenylyl cyclase activity is enhanced in the nucleus accumbens and caudate putamen of rats administered cocaine (15 mg/kg, i.p.) three times daily in a binge pattern for 14 days. Data are expressed as percent of basal activity and are shown as mean ± SEM from 6–8 animals (each performed in triplicate). The stimulation curves for dopamine and SKF 82958 were evaluated by a two-way ANOVA, where dopamine (F (1,64) = 25.73, P < 0.001, nucleus accumbens; F (1,64) = 19.04, P < 0.0001, caudate putamen), and SKF 82958 (F (1,61) = 29.71, P < 0.001, nucleus accumbens; F (1,63) = 31.47, P < 0.0001, caudate putamen) produced significantly greater activation of adenylyl cyclase activity in the cocaine-exposed rats as compared to saline-injected controls. (Data from Unterwald et al., 1996. Reprinted with permission from Eur J Pharmacol.)


Chronic cocaine exposure perturbs the mammalian nervous system. As a means of counteracting this perturbation or insult, the nervous system responds by producing stable neurochemical alterations in the brain, particularly within dopamine responsive neurons. These long-lived changes in brain neurochemistry occur through alterations in gene expression and are maintained to ultimately underlie the behavioral abnormalities, which characterize addiction (Nestler, 1994, 2001). The earliest event in gene regulation involves the activation of transcription factors, nuclear proteins, which regulate gene transcription. Hundreds of transcription factors are believed to be intimately involved in controlling various patterns of gene expression within the mammalian brain. Only a handful of transcription factors, however, have been uniquely related to drug addiction. Two transcription factors in particular that have been implicated in the D1 receptor link to cocaine addiction are CREB (cAMP response element binding protein) and deltaFosB.

CREB is a well defined transcription factor, which mediates many of the effects of the cAMP system on gene expression (Lane-Ladd et al., 1997). Via indirect D1 receptor stimulation, psychostimulants like cocaine and amphetamine activate the cAMP pathway and increase PKA-dependent phosphorylation of CREB (Yamamoto et al., 1988; Gonzalez and Montminy, 1989; Konradi et al., 1994). Phosphorylated CREB then acts to control the transcriptional regulation of many genes, with the most often cited genes being the immediate early genes (i.e., Fos, Jun) (Kano et al., 1995). Immediate early genes and their protein products act to further transduce these transient drug-induced intracellular signals to more long-lived nuclear adaptations.

Direct support linking CREB to the rewarding effects of cocaine came about with a finding by Carlezon et al. (1998). A herpes simplex virus vector-mediated approach was used to over-express CREB in the nucleus accumbens of rats. In a place conditioning test, rats over-expressing CREB were found to spend less time in the cocaine-paired environment compared to control animals suggesting that cocaine reward is regulated by CREB, a potential intracellular messenger linked to D1 receptor activation. Further support linking chronic drug exposure to changes in gene expression within the cAMP system came from microarray analysis. In a study in non-human primates using cDNA hybrid arrays, chronic cocaine exposure was found to increase the expression of the alpha catalytic subunit of the PKA gene (Freeman et al., 2001). Together these studies implicate the cAMP system as a specific mediator of cocaine's long-lived effects. Additionally, these data too support the theory that cocaine-induced behavioral manifestations occur as a result of D1 receptor linked alterations in gene expression.

DeltaFosB has too been implicated in drug addiction (Nestler, 2001; Nestler et al., 2001). DeltaFos B is a member of the Fos family of transcription factors, which includes c-Fos, Fos B, and the Fos related antigens (FRA's) (Morgan and Curran, 1995). In contrast to its family members and CREB, which are relatively short-lived proteins, deltaFosB expression is long-lived and thus persists and accumulates following repeated drug exposure (Hope et al., 1992, 1994). Specifically, deltaFosB accumulates in striatonigral neurons containing D1 receptors and dynorphin following chronic cocaine exposure. DeltaFosB has been the target of much research because it may represent one tangible neurobiological adaptation resulting from repeated drug exposure that persists in neurons even after drug consummation ceases (Nye et al., 1995; Nestler, 2001; Nestler et al., 2001). Support for the long-term effects of deltaFosB came about with a recent finding by Kelz et al. (1999) who over-expressed deltaFosB in nucleus accumbens neurons of mice using a tetracycline-gene regulated system. deltaFosB over-expressing mice found cocaine more rewarding than did control mice. The mice that over-expressed deltaFosB also showed an augmented locomotor response to chronic cocaine. These data indicate that deltaFosB increases an animal's sensitivity to cocaine (Kelz et al., 1999). Additionally, deltaFosB may be representative of one neurobiological mechanism by which drugs of abuse like cocaine wreak havoc on the brain and perpetuate an addictive state.


Historically, researchers have revolved their theories of cocaine addiction solely around the neurotransmitter, dopamine. One can ascertain from evidence presented in this review that dopamine and the D1 receptor signaling pathway are very much involved in the pharmacology of cocaine and the pathology of cocaine addiction. The specific nature of the dopamine signaling system in the overall pharmacology of cocaine, however, is still very much uncertain. Additionally, cocaine addiction is not just a story about dopamine and aberrant changes in dopamine signaling. Recent evidence extrapolated from multiple animal models suggest that cocaine's effects are very broad and that numerous neurotransmitters mediate the actions of this molecule. Moreover, it is the intricacy with which cocaine impacts so many neurotransmitter systems that make this molecule's neuropharmacological profile so diverse and the neuropharmacological mechanisms of cocaine addiction so bewildering. Technological advances in science have enabled researchers to combine methods in basic molecular biology with methods in behavioral neuroscience to gain a better understanding of the neurobiological mechanisms involved in cocaine addiction. Nonetheless, cocaine addiction still remains a very complex phenomenon. Therefore, collectively, it can be surmised that cocaine's effect on multiple neurotransmitters summate to produce the complex neurochemical and behavioral sequlae characteristic of cocaine addiction.


NIH grant 09580 was awarded to E.M.U, and P50 DA 05130 was awarded to M.J. Kreek, (see Unterwald et al., 1996, 2001).