Introduction

Top Abstract Introduction Methods Results Discussion References

The dopaminergic system is involved in a variety of functions including motor control, sexual functions, cardiovascular homeostasis, endocrine regulation and cognition (for review see Jackson and Westlind-Danielsson, 1994). Disruption of this system is associated with neurological disorders such as schizophrenia, Parkinson's disease and Huntington's chorea. In addition, central dopaminergic pathways have been implicated in the reinforcing and subjective effects of drugs of abuse such as cocaine (Johanson and Fischman, 1989; Kuhar et al., 1991; Witkin, 1994). A major focus of research involving pharmacological treatment for all of these disorders has been the development of drugs that target specific subtypes of the various dopamine receptors (Seeman and Van Tol, 1994). Some success in this area has been attained, including the use of dopamine D₂/D₃ agonists (e.g., bromocriptine) in treating Parkinson's disease and D₂/D₃ antagonists in treating psychosis (for review see Seeman and Van Tol, 1994; Seeman, 1995).

At present, at least five dopamine receptor subtypes have been identified and have been designated as D₁-D₅ (Kebabian and Calne, 1979; Sibley and Monsma, 1992). D₃, D₄ and D₅ dopamine receptor subtypes share molecular characteristics with either D₁ or D₂ receptors and based on these similarities have been grouped into two subfamilies: "D₁-like" which include D₁ and D₅ subtypes and "D₂-like" which include D₂, D₃ and D₄ subtypes (Sibley and Monsma, 1992). Until recently, however, dopamine receptors were only divided into two subtypes: D₁ receptors that activate adenylyl cyclase and D₂ receptors that inhibit or have no effect on the enzyme (Kebabian and Calne, 1979). After the development of ligands selective for the D₁ and D₂ receptor subtypes, experimental evaluation revealed that selective stimulation of each subtype produced different types of unconditioned behavioral effects (for review see Clark and White, 1987). For example, D₁ agonists produce intense grooming and abnormal perioral movements in rats in the absence of stereotyped behaviors (Rosengarten et al., 1983; Molloy and Waddington, 1985; Murray and Waddington, 1989). In squirrel monkeys, full D₁ agonists produce increases in the frequencies of stationary postures and head movements (Rosenzweig-Lipson et al., 1994). Conversely, D₂ agonists produce stereotypies such as sniffing, rearing and head movements in rats (Molloy and Waddington, 1985; Arnt et al., 1987) and scratching in squirrel monkeys (Rosenzweig-Lipson et al., 1994; Pellón et al., 1995). Given that D₁ and D₂ agonists have been previously demonstrated to produce distinct behavioral effects, the question of whether the newly identified subtypes (D₃, D₄ or D₅) share functions with other subtypes in their respective subfamilies is of interest.

At present, the availability of ligands that selectively target each of the dopamine receptor subtypes is limited. However, drugs have been identified that bind with varying degrees of selectivity for the various receptor subtypes. For example, agonists have been shown to bind to both D_2, D₃ and D₄ receptors, including PD 128,907, (±)7-OH-DPAT and quinpirole. As for their selectivity at D₂ and D₃ receptors, these compounds display a slightly higher affinity for the high-affinity D₃ receptor than for the high-affinity D₂ receptor (Chio et al., 1994; Burris et al., 1995; Malmberg and Mohell, 1995; Pugsley et al., 1995). Under different assay conditions that did not take into account high- vs. low-affinity dopamine binding sites, the D₂-like agonist, bromocriptine, has been shown to have similar affinity for D₂ and D₃ receptors (Sokoloff et al., 1992; Freedman et al., 1994). Although quinpirole has been shown to have similar affinities for D₃ and D₄ receptors (Seeman and Van Tol, 1994), bromocriptine, 7-OH-DPAT and PD 128,907 have a greater selectivity for D₃ vs. D₄ receptor subtypes, with a rank order of potency of D₃>D₂>D₄ (Seeman and Van Tol, 1994; Pugsley et al., 1995). PD 128,907, 7-OH-DPAT bromocriptine, quinpirole are referred to throughout the manuscript as D₃/D₂ agonists. Our series of experiments was designed to examine the behavioral effects produced by these D₃/D₂ agonists. These effects were compared to those produced by dopamine agonists that have been used in the past to define dopamine-mediated behaviors, including the D₁ agonists SKF 38393 and SKF 82958, the nonselective dopamine agonist, apomorphine, and the indirect (uptake blocker) dopamine agonist, cocaine.

The ability of the dopaminergic agonists to produce climbing, immobility and gnawing was examined in mice. Climbing and gnawing previously have been suggested to require stimulation of both D₁ and D₂ receptor subtypes given that apomorphine or D₁/D₂ agonist combinations have been demonstrated to readily produce these behaviors in rodents (Arnt et al., 1987; Moore and Axton, 1988; Vasse et al., 1988; Murray and Waddington, 1989). Conversely, neither D₁ or D₂ agonists (Arnt et al., 1987; Arnt et al., 1988; Moore and Axton, 1988; Vasse et al., 1988; Murray and Waddington, 1989; Daly and Waddington, 1992; Tirelli and Witkin, 1995) nor cocaine (Tirelli and Witkin, 1994a) have been shown to readily produce climbing or gnawing when administered alone.

Reductions in locomotor activity and/or increases in immobility previously have been found in rodents after the administration of D₂ agonists (Martin and Bendesky, 1984; Martin et al., 1984; Eilam and Szechtman, 1989; Puglisi-Allegra et al., 1990; Jackson and Westlind-Danielsson, 1994) and apomorphine (Tirelli and Witkin, 1994a). Conversely, increases in locomotor activity and/or decreases in immobility have been associated with D1 agonists (Murray and Waddington, 1989; Daly and Waddington, 1992, 1993; Tirelli and Terry, 1993; Deveney and Waddington, 1995) or cocaine (Johanson and Fishman, 1989; Kelley and Lang, 1989; Pierce and Rebec, 1990; French and Witkin, 1993; Witkin, 1993) administration. Given the broad spectrum of effects that dopamine agonists have on locomotor activity and immobility, these behaviors were also measured in mice.

Scratching in squirrel monkeys was also examined after administration of these dopaminergic compounds because it has been suggested as an in vivo model of D₂ receptor stimulation (Rosenzweig-Lipson and Bergman, 1993; Rosenzweig-Lipson et al., 1994; Pellón et al., 1995). Whereas apomorphine has been shown to produce scratching in squirrel monkeys (Pellón et al., 1995), neither D1 agonists nor cocaine has been shown to produce this effect in monkeys (Rosenzweig-Lipson et al., 1994; Pellón et al., 1995).

PD 128,907, 7-OH-DPAT, quinpirole and bromocriptine generally produced behavioral effects that could be distinguished from those produced by apomorphine, cocaine and the D₁ agonists. These findings suggest that D₂ and/or D₃ receptor stimulation may be involved in producing behaviors that were once considered to be mediated exclusively by stimulation of D₂ receptors (decreased locomotor activity and scratching) or D₁/D₂ receptors (climbing and gnawing).

	Methods

Top Abstract Introduction Methods Results Discussion References

Behavioral Observations

Subjects. Male Swiss-Webster mice (Taconic Farms, Germantown, NY) were housed in groups of five in a temperature controlled vivarium (22-25°C) with a 12-hr light/dark cycle. Food (Zeigler Small Animal Feed, St. Louis, MO) and water were available ad-libitum. Behaviorally naive mice were studied only once (n = 6/dose).

Apparatus and procedure. As previously described by Tirelli and Witkin (1994a), behavioral observation cages were constructed of wire screen (1 cm mesh, 15 × 15 × 26 cm high). During testing, the bottomless cages were placed on a smooth surface. Testing sessions always involved six cages with one mouse per cage. After drug treatments (i.p.), mice were immediately placed into a cage for a 65-min session. During 5-min observation periods beginning at 10, 35 and 60 min postinjection, a subject was observed every 30 sec for the presence or absence of climbing, immobility, nose poking, switching sides and gnawing as described below. For each behavior, each subject could receive a total of 10 points during each observation period and, therefore, 30 points over the entire session. Climbing was recorded when the wire of the test chamber was grasped with all four paws for at least 15 sec (Tirelli and Witkin, 1994a). Climbing could simultaneously occur with immobility, nose poking, switching sides and gnawing. Immobility was scored when no movement occurred anywhere within the cage for at least 15 sec; immobility could simultaneously occur with climbing, but could not occur simultaneously with any other of the behaviors being scored. Nose poking was recorded when the nose was projected through the wire-mesh in the absence of gnawing. Switching sides was recorded after movement from one wall of the screen to another (including the ceiling and floor). Gnawing was recorded when the incisors were placed over the wire-mesh and was often accompanied by the typical jaw movements of biting (Tirelli and Witkin, 1994a). A single experimenter scored each experimental session.

Locomotor Activity

Subjects. Male Swiss-Webster mice (Taconic Farms, Germantown, NY) were housed and fed as above. Eight mice were tested at each dose of dopamine agonist.

Apparatus and procedure. Immediately after an i.p. injection of a dopamine agonist, subjects were individually placed into a 40-cm³ activity monitor equipped with photoelectric detectors placed 2.5 cm apart along the perimeter that were capable of sensing ambulatory activity at a height of 2.5 cm above the floor (Omnitech Electronics, Columbus, OH). Horizontal locomotor activity was recorded in 30-min intervals for a total of 180 min.

Scratching

Subjects. Four adult male squirrel monkeys (Saimiri sciureus) with mean body weights of 860 to 1,050 g were housed individually in a temperature and humidity-controlled room with a 12-hr light/dark cycle. Cages permitted two-way visual and olfactory (not tactile) access to other squirrel monkeys and to the observer. Monkeys had unrestricted access to food (Lab Diet, High Protein Monkey Diet, St. Louis, MO and Golden Squares, BioServ, Frenchtown, NJ, supplemented with fresh fruit) and water. They had been studied previously in experiments involving acute administration of drugs or cocaine self-administration, but had not participated in an experiment for 5 mo before the start of this study.

Procedure. Within a cumulative dosing procedure, monkeys (n = 4) received four successive injections (i.m. in thigh) of incremental doses of a dopamine agonist separated by an interval of 15 min. Immediately after each injection, they were observed in their home cages during three 1-min periods. Observations occurred at 5, 9 and 13 min after an injection. During each observation period, scratching was recorded for each monkey if it occurred on at least one occasion for at least 2 sec. Each monkey could receive three positive scores at each dose tested. Across subjects, there was a maximum of 12 positive scores possible at each dose. A single experimenter scored each experimental session.

Drugs

PD 128,907 (Research Biochemicals Inc., RBI, Natick, MA), 7-OH-DPAT (RBI), trans-(-)-4aR-4,4a,5,6,7,8,8a,9-Octahydro-5-propyl-1H-pyrazolo[3,4-g]quinoline HCl [quinpirole HCl (LY 171555); RBI], SKF 38393 (RBI), SKF 82958 (RBI), R(-)-10,11-Dihydroxyaporphine HCl (apomorphine; Sigma Chemical Co., St. Louis, MO) and 2B-carbomethoxy-3B-benzoyloxytropane HCl (cocaine HCl; Sigma Chemical Company, St. Louis, MO) were dissolved in distilled water. Bromocriptine mesylate (Sigma) was dissolved in distilled water and minimal HCl. If necessary, mild heat and sonication were used for dissolution. For mouse experiments, injections were given i.p. in volumes of .01 ml/g body weight. For squirrel monkey experiments, injections were given i.m. and no more than 1.5 ml was injected into each animal during each testing session. Vehicle injections were of a volume similar to that of the drugs in each experiment.

Data Analysis

For observational experiments in mice, the data taken at each of the three time intervals (10, 35 and 65 min) did not consistently vary at each interval; hence, the data were collapsed over the entire session. The total number of occurrences of each behavior made during the session were averaged across all six subjects at each dose of dopamine agonist. Percentages were obtained by dividing the average by the maximum possible score that each subject could receive during the session (30); that result was multiplied by 100. These data were analyzed using Fischer's exact probability test by comparison to vehicle control values. For locomotor activity data, individual contrasts at each 30-min interval for each dose were compared to vehicle control values using Dunnett's test (two-tailed) after significant ANOVA. Means (± S.E.M) were also calculated across two 90-min intervals (0-90 min; 90-180 min) at each dose. Individual contrasts at each dose during these 90-min periods were compared against control values using Dunnett's test (two-tailed). The percentage of observations with scratching was obtained by dividing the total number of observations that contained at least one instance of scratching by the overall maximum number of observations in all four monkeys (12). ED₅₀ values with 95% CI were obtained according to the methods described by Litchfield and Wilcoxon (1949) with specific comparisons between treatments and vehicle control values using Fischer's exact probability test. For each analysis, error probabilities of more than .05 were considered to be nonsignificant.

	Results

Top Abstract Introduction Methods Results Discussion References

Behavioral observations. Vehicle-treated control mice displayed climbing (35% ± 0.83) and few occurrences of immobility and gnawing (0 and 9% ± 2, respectively) during the 65-min observation period (figs. 1, 2, 3). The nonselective dopamine agonist apomorphine (3-30 mg/kg) and the dopamine uptake inhibitor cocaine (0.3-30 mg/kg) increased climbing at all doses studied (fig. 1). Whereas apomorphine increased immobility at the lowest dose tested (3 mg/kg), cocaine had no effect on immobility. Gnawing was increased in a dose-dependent manner by apomorphine but was not affected by cocaine.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Effects of the nonselective dopamine agonist, apomorphine, and the dopamine uptake inhibitor, cocaine, on climbing, immobility and gnawing in mice (n = 6 per dose) within a wire-mesh box for 65 min. Data presented are the percentage of observations in which these behaviors were observed. Dashed lines represent data for vehicle-treated control subjects (± S.E.M.). Statistical comparisons with control values were determined using Fisher's Exact Probability test (*P< .05).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Effects of the partial D1 agonist, SKF 38393, and the full D1 agonist, SKF 82958, on climbing, immobility and gnawing. Other details as in figure 1.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effects of the D3/D2 dopamine agonists PD 128,907, 7-OH-DPAT, quinpirole and bromocriptine on climbing, immobility and gnawing. Other details as in figure 1.

Locomotor activity. During the first 30 min of the locomotor activity session, vehicle-treated mice displayed moderate levels of activity (4000-8000 counts) that decreased over the remainder of the 180-min session (figs. 4 and 5, filled circles). Locomotor activity after apomorphine (fig. 4; 3-30 mg/kg), SKF 38393 (fig. 4; 100 mg/kg), PD 128,907 (fig 5; 0.3-30 mg/kg), 7-OH-DPAT (fig 5; 1-30 mg/kg), quinpirole (fig 5; 0.01-100 mg/kg) and bromocriptine (fig. 5; 3 and 30 mg/kg) was decreased relative to control subjects during the first 30 min. These decreases in activity were not evident when data were averaged across the first 90 min of the session (figs. 4 and 5, insets). During the remainder of the session, activity after these compounds was similar to control levels at most of the 30-min time intervals. At higher doses of these compounds (with the exception of bromocriptine) activity levels increased during at least one of the latter 30-min time intervals. Increases in activity were evident during the latter half of the session following apomorphine (fig. 4, inset; 10 and 30 mg/kg), SKF 38393 (fig. 4, inset; 30 and 100 mg/kg), 7-OH-DPAT (fig. 5, inset; 10 mg/kg) and quinpirole (fig. 5, inset; 30 and 100 mg/kg).

View larger version (40K): [in this window] [in a new window]   Fig. 4.   Effects of increasing doses of apomorphine, cocaine, SKF 38393 and SKF 82958 on horizontal locomotor activity in mice (n = 8 per data point). Injections (i.p.) were given immediately before testing and data were recorded at 30-min intervals for a total of 180 min. Activity at each 30-min interval and the mean (± S.E.M) of this activity over the first and second 90-min periods is also shown (inset). Individual contrasts at each 30-min interval were compared against saline control values (closed circles) using Dunnett's test (two-tailed; *P < .05). Individual contrasts at each dose during the 90-min periods were compared against saline control values (0 mg/kg) using Dunnett's test (two-tailed; *P < .05).

View larger version (39K): [in this window] [in a new window]   Fig. 5.   Effects of increasing doses of PD 128,907, 7-OH-DPAT, quinpirole and bromocriptine on horizontal locomotor activity in mice (n = 8 per data point). Other details as in figure 4.

Scratching. Cumulative doses of apomorphine (ED₅₀ = 0.10 µmol/kg; 95% CI = 0.03-0.23 µmol/kg), PD 128,907 (ED₅₀ = 0.10 µmol/kg; 95% CI = 0.07-0.24 µmol/kg), 7-OH-DPAT (ED₅₀ = 0.02 µmol/kg; 95% CI = 0.003-0.09 µmol/kg) and quinpirole (ED₅₀ = .12 µmol/kg; 95% CI = 0.03-0.39 µmol/kg) dose-dependently increased scratching in squirrel monkeys (fig. 6). The maximum frequency of scratching episodes after each of these compounds was approximately 80%. Neither SKF 38393, SKF 82958 nor cocaine significantly increased scratching. Within this cumulative dosing procedure, bromocriptine also did not produce scratching. However, bromocriptine did produce scratching when a single dose (0.3 mg/kg) was given at the onset of testing and scratching was recorded at 15-min intervals for a total of 90 min. At this dose, scratching episodes reached their highest levels (75%) during the 60- to 75-min interval.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   Effects of cumulative doses of dopamine agonists on scratching in squirrel monkeys (n = 4). Four successive injections were given incrementally and separated by an interval of 15 min. Data presented are the percentage of observations in which scratching was observed. The unconnected, inverted triangle above 0.3 mg/kg represents the percentage of scratching observations after this single dose of bromocriptine recorded during the 60- to 75-min interval of a 90-min session. Control data are presented to the left as unconnected points above "V." Statistical comparisons with control values were determined using Fisher's exact probability test (*P < .05).

	Discussion

Top Abstract Introduction Methods Results Discussion References

Our series of experiments examined the behavioral effects of several dopamine agonists that bind to D_3, D₂ and minimally to D4 receptors (PD 128,907, 7-OH-DPAT, quinpirole and bromocriptine). As summarized in table 1, during behavioral observations in mice, the D₃/D₂ agonists produced immobility that resulted in a decrease in locomotor activity over a wide dose range, had no effect on gnawing, and only inconsistently induced climbing. In squirrel monkeys, these D₃/D₂ agonists dose-dependently produced scratching. Although there was some overlap between the behaviors produced by the D₃/D₂ agonists and the other types of dopamine agonists tested, behavioral effects could be differentiated. For example, whereas the nonselective agonist, apomorphine, increased gnawing, the D₃/D₂ agonists had no effect on gnawing across a wide range of doses that readily produced other behavioral effects. The D₃/D₂ agonists also produced a different spectrum of effects (immobility, locomotor activity decreases and scratching) from those produced by the indirect dopamine agonist, cocaine and the D₁ agonists, SKF 38393 and SKF 82958 (scratching). Together these findings demonstrate that D₃/D₂ agonists produce behavioral effects that can be differentiated from those produced by nonselective dopaminergic agonists, indirect dopamine agonists and selective D₁ agonists.

Many of the compounds (e.g., quinpirole) used to define D₂-mediated behaviors in the past have now been shown also to bind to D₃, D₂ and D₄ receptors. That these agonists, as well as the newly identified D₃/D₂ agonists, PD 128,907 and 7-OH-DPAT, produced similar behavioral effects suggests that behaviors once attributed to D₂ receptor stimulation may also involve stimulation of D₃ receptors. For example, decreases in locomotor behavior by low doses of dopamine agonists have been attributed previously to presynaptic stimulation of D₂ autoreceptors (Puglisi-Allegra et al., 1990; Jackson and Westlind-Danielsson, 1994). That apomorphine, PD 128,907, 7-OH-DPAT, quinpirole and bromocriptine decreased locomotor activity over a wide dose range within the first 30 min of the test session suggests that D₃ receptors also may be involved. At present, it is not clear if D₃ receptor involvement in this, or any other behavior, is due to stimulation of presynaptic D₃ autoreceptors (Meller et al., 1993; Starr and Starr, 1995) or postsynaptic D₃ receptors (Waters et al., 1993; Svensson et al., 1994).

Scratching in squirrel monkeys also has been characterized previously as a D₂-mediated behavior (Rosenzweig-Lipson et al., 1994; Pellón et al., 1995). For example, Pellón et al. (1995) recently concluded that scratching in squirrel monkeys is D₂-mediated given that several dopamine agonists that had been demonstrated previously to bind to D₂ receptors (quinpirole, (-)-NPA, RU 24213, pergolide, quinelorane and piribedil) dose-dependently increased scratching in these subjects. However, quinelorane (Sokoloff et al., 1992) and pergolide (Sokoloff et al., 1990) have also been shown to bind to D₃ receptors. Given that these compounds (Pellón et al., 1995), as well as PD 128,907, 7-OH-DPAT, quinpirole, bromocriptine and apomorphine, produced scratching suggests that scratching may involve D₂ and/or D₃ receptors. Interestingly the nonselective dopamine agonist, apomorphine, produced scratching when the indirect dopamine agonist cocaine did not. Differences in the behavioral effects produced by apomorphine versus cocaine have been reported previously (Tirelli and Witkin, 1994a, b; Pellón et al., 1995) and elsewhere in the present report on measures of immobility, gnawing (fig. 1) and locomotor activity (fig. 4).

Stereotyped climbing and gnawing have been previously discussed in terms of the simultaneous stimulation of D₁ and D₂ receptors given that apomorphine or D₁/D₂ agonist combinations have been demonstrated to readily produce these behaviors (Arnt et al., 1987; Moore and Axton, 1988; Vasse et al., 1988; Murray and Waddington, 1989). In our experiment, climbing was increased after apomorphine and cocaine at each dose tested (fig. 1). Climbing was also increased at select doses of SKF 82958 (fig. 2), PD 128,907, 7-OH-DPAT and quinpirole (fig. 3); these increases, however, were not dose dependent. That climbing was observed after the D₁ and D₃/D₂ agonists, suggests that the amount of stimulation required at multiple dopamine receptor sites to produce this behavior may be less than that required to produce gnawing. Gnawing was dose-dependently produced by apomorphine and the highest dose tested of SKF 82958 (100 mg/kg; fig. 2), a full D₁ agonist (O'Boyle et al., 1989; Anderson and Jansen, 1990). Interestingly, SKF 38393, a partial D₁ agonist that binds with greater selectivity for D₁ vs. D₂ receptors than SKF 82958 (O'Boyle et al., 1989; Anderson and Jansen, 1990), did not produce gnawing at any dose tested (1-100 mg/kg). Overall, these findings support past research indicating that the production of climbing and gnawing require the simultaneous stimulation of multiple dopamine receptor sites. At this point, the extent that D₃ receptors may be contributing to the production of these behaviors is unclear.

A great deal of interest has been taken in D₃ receptors given their abundance in mesolimbic dopaminergic regions (Sokoloff et al., 1990), an area that has been associated with emotionality, the efficacy of antipsychotic drugs and psychomotor stimulant abuse. For this reason, it has been suggested that D₃ receptors may be involved in schizophrenia or cocaine abuse, and may serve as targets for drug development in these areas. Recent reports have found relationships between the D₃/D₂ agonists PD 128,907 and 7-OH-DPAT and the behavioral effects produced by cocaine (Acri et al., 1995; Spealman, 1996). For example, Acri et al. (1995) reported that PD 128,907 and 7-OH-DPAT fully substituted for the discriminative stimulus effects of cocaine in rats suggesting that D₃ agonists may serve as a substitution-type for cocaine abusers. However, as reported by Acri et al. (1995) there were also differences in the behavioral effects produced by cocaine and these D₃/D₂ agonists. The response rate reductions observed with PD 128,907 and 7-OH-DPAT (Acri et al., 1995) were unlike those after cocaine and other compounds that fully substitute for cocaine (Spealman et al., 1991; Baker et al., 1993; Witkin, 1994). Differences in overall activity levels produced by D₃/D₂ agonists vs. cocaine were also found in our study (fig. 1 vs. 3 and fig. 4 vs. 5). These side effects may mitigate against the therapeutic use of the present drugs. None-the-less, 7-OH-DPAT and quinpirole have been reported to reduce cocaine self-administration in rats at doses that alone were not reinforcing (Caine and Koob, 1993; 1995) further suggesting that D₃ receptor agonists may have potential as treatments for cocaine abuse.

The therapeutic potential of D₃/D₂ receptor ligands may extend beyond that of the treatment of cocaine abuse. As demonstrated by Rodriguez de Fonseca et al. (1995), 7-OH-DPAT can modulate the acquisition and expression of morphine-induced place-preference in rats. Together these findings suggest a potential use of D₃ agonists in the treatment of drug abuse, and possibly, other psychiatric and neurological disorders that have been associated with disruptions in dopamine neurotransmission. However, the development of additional D₃ ligands of better selectivity (Sokoloff et al., 1990; Rivet et al., 1994; Wright et al., 1995), is essential to exploring this hypothesis further.

The results of our study have shown that D₃/D₂ agonists produce behavioral effects that can be distinguished from nonselective dopamine receptor agonists, from D₁ agonists and from dopamine uptake inhibitors. Furthermore, behaviors once attributed to D₂ receptors alone (decreased locomotor activity and scratching) or attributed to the simultaneous stimulation of D₁ and D₂ receptors (climbing and gnawing) may also involve D₃ receptors. This study complements the work of others who have suggested that D₃ receptors may play a role in the control of certain behaviors, including yawning (Longoni et al., 1987; Damsma et al., 1993; Bristow et al., 1996), chewing (Daly and Waddington, 1993), sniffing (McElroy et al., 1993), hypothermia (Ahlenius and Salmi, 1994; Millan et al., 1994) and reductions in locomotor activity (Daly and Waddington, 1993; Svensson et al., 1994; Pugsley et al., 1995; Starr and Starr, 1995; Bristow et al., 1996; Depoortere et al., 1996).

Although these findings suggest that D₃ receptors play a role in the control of behavior, the identification of agonists and antagonists that more selectively target D₃ receptors and other subpopulations of dopamine receptors (e.g., D₂, D_4, D₅) is essential to obtaining a more precise understanding of the role these subtypes are playing in drug abuse and other neuropsychiatric disorders, as well as in nonpathological states.