Introduction

Top Abstract Introduction Methods Results Discussion References

Dopamine receptors are differentiated into two major types: the D1-like receptors, which include the D₁ and D₅ receptors, and the D2-like receptors, which include the D₂, D₃ and D₄ receptors (Sibley and Monsma 1992; Sokoloff and Schwartz 1995; Baldessarini and Tarazi 1996). Chronic antipsychotic drug administration is a crucial component in the current treatment of schizophrenia. The observation that antipsychotics bind to and block striatal dopamine D2-like receptors in a direct correlation with their clinically effective antipsychotic doses (Creese et al., 1976; Seeman et al., 1976), implicates a major role of dopaminergic systems in schizophrenia. A number of animal studies reported an up-regulation of striatal D2-like receptors after subchronic (3-4 wk) drug administration (Burt et al., 1977; Seeman, 1980; O'Dell et al., 1990; Tarazi et al., in press), the period during which beneficial clinical effects of antipsychotics on patients are first noticed. However, the classical "typical" antipsychotics, including haloperidol, although effectively blocking psychoses, also cause extrapyramidal movement disorders, both transient (parkinsonism, dystonia, akathisia) and chronic (tardive dyskinesia, TD) (Jeste and Wyatt, 1979; Baldessarini and Tarsey, 1980). Atypical antipsychotic drugs, the prototype of which is clozapine, have been developed that are devoid of these extrapyramidal side effects. Short treatment periods of rats with clozapine is reported not to increase striatal D2-like receptor levels (O'Dell et al., 1990; Tarazi et al., in press). This suggests that the upregulation of striatal D2-like receptors is more likely associated with drug-induced extrapyramidal motor side effects rather than mediation of their antipsychotic action.

To gain further insight into the mechanisms of antipsychotic drug action and their involvement in the development of extrapyramidal side effects, an animal study examining the effects of prolonged antipsychotic treatment (6 mo or more), which mimicks the time frame of TD developing in antipsychotic-treated schizophrenic patients, would be preferable. Dopamine receptor up-regulation has been observed in striatal areas after prolonged drug treatment with antipsychotic drugs (Clow et al., 1980; Owen et al., 1980; Murugariah et al., 1983; Rupniak et al., 1984). However, a comparison between antipsychotics with respect to regional (extrastriatal) differences and degree of receptor up-regulation has not been made. Moreover, the recent cloning of several (D_1-5) dopamine receptor subtypes (Sokoloff et al., 1990; Zhou et al., 1990; Sunahara et al., 1991; Van Tol et al., 1991), has led to the realization that the systems are more complicated than previously thought, with most of the available radioligands binding to more than one dopamine receptor subtype, therefore complicating interpretation of previous in vitro receptor binding studies. This is illustrated by the recent finding of D2-like receptor up-regulation in postmortem striatal tissue from drug-naive or antipsychotic-treated schizophrenic patients using ³H-nemonapride, but not ³H-raclopride, both D2-like receptor radioligands (Seeman et al., 1993; Murray et al., 1995; Sumiyoshi et al., 1995).

We have examined whether chronic (8 mo) treatment with the typical antipsychotic (haloperidol), the atypical antipsychotic (clozapine) and the D₂/D₃ receptor antagonist raclopride which exhibits antipsychotic activity (Farde et al., 1988; Cookson et al., 1989) differentially affect dopamine receptor subtype binding measured by in vitro receptor autoradiography, in various brain regions in the rat using a number of radioligands with different dopamine receptor subtype affinities. Changes in dopamine receptor subtype binding will be interpreted as a functional index of in vivo antipsychotic drug action. Up- or down-regulation of a receptor subtype by both typical and atypical antipsychotics in a specific brain region will be suggestive of a common locus for antipsychotic drug action. Differences in region-specific or receptor subtype regulation by typical vs. atypical antipsychotics may indicate the involvement of a specific dopamine receptor subtype in the induction of extrapyramidal side effects.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials

Radioligands were obtained from New England Nuclear-Du Pont (Wilmington, DE) and Amersham (Arlington Heights, IL). Flupenthixol, ketanserin, eticlopride and sulpiride were purchased from Research Biochemical Inc. (Natick, MA). Clozapine was a gift from Sandoz (East Hanover, NJ), SCH23390 was a gift from Schering-Plough Research (Kenilworth, NJ) and raclopride was a gift from Astra Läkemedel AB (Södertälje, Sweden). All other compounds were purchased from Sigma Chemical Co. (St. Louis, MO).

Drug Treatment

Different groups of male Sprague-Dawley rats (Charles-Rivers, VA) weighing 200 to 220 g on delivery, were maintained under controlled light and temperature conditions but given free access to food and water. These rats were housed in the animal facility of Maryland Psychiatric Research Center (Baltimore, MD) and treated for 8 mo with three different drugs in the following doses: haloperidol (1.5 mg/kg/day), clozapine (25 mg/kg/day) and raclopride (10 mg/kg/day). The drugs were given in drinking water to mimic oral administration of antipsychotic drugs in patients with schizophrenia, and the control consisted of tap water adjusted to pH = 6.0. The solutions were made based on the average weekly weight of the rats and an estimation of their daily solution consumption.

Tissue Preparation

Immediately after drug treatment, rats were killed by decapitation, their brains were quickly removed, frozen by immersion in chilled isopentane and stored in liquid nitrogen until use. Coronal sections (16 µm) were cut in a cryostat at 20°C, thaw-mounted on gelatin-coated microscopic slides and stored at 80°C until use. On the day of the experiment, slides were thawed on a slide warmer and air dried at room temperature.

Receptor Binding

A number of ligands were used to quantify the different dopamine receptor subtypes. Four different radioligands were used to quantify the D2-like receptors, ³H-nemonapride (previously known as ³H-YM-09151-2), ³H-spiperone, ³H-raclopride and ¹²⁵I-sulpiride. The first three radioligands were selected based on their high affinities for D2-like receptors. Saturation and competition experiments showed that the binding of the three ligands to striatal sections was saturable and was inhibited by D2-like receptor antagonists (Palacios et al., 1981; Köhler and Radesater, 1986; Yokoyama et al., 1994). Interestingly, these three ligands have different specificities for the different D2-like receptors. ³H-nemonapride and ³H-spiperone bind with high affinity to the three D2-like receptors (D₂, D₃ and D₄) in expression systems, although ³H-raclopride has high affinity for the D₂ and D₃ receptors and a much lower affinity for the D₄ receptor (Van Tol et al., 1991). ¹²⁵I-sulpiride was chosen to examine the response of extrastriatal (especially cortical) D2-like receptors that are expressed at low levels to chronic antipsychotic treatment.

D₃ receptors were quantified using ³H-7-OH-DPAT, the first selective radioligand that binds to the D₃ receptor with subnanomolar affinity compared to nanomolar affinities for D₂ and D₄ receptors in transfected cell lines (Levesque et al., 1992).

The D1 receptor was quantified using ³H-SCH23390 according to the method of Dawson et al. (1986). This radioligand labels both the D₁ and D₅ receptors equally, so it is considered to be a D1-like receptor ligand (Sunahara et al., 1991). Serotonin 5HT₂-like receptors were quantified using ³H-ketanserin according to O'Dell et al. (1990).

Receptor Autoradiography

D1-like receptor binding. Sections were preincubated for 1 hr in assay buffer (50 mM Tris-HCl, 120 mM NaCl, 5 mM KCl, 2 mM CaCl₂ and 1 mM MgCl₂; pH 7.4) at room temperature. Sections were then incubated for 1 hr at room temperature in assay buffer containing 1 nM ³H-SCH23390 (specific activity 73.2 Ci/mmol) and 40 nM ketanserin to block 5HT₂-like receptors. Nonspecific binding was determined in the presence of 1 µM flupenthixol. After incubation, slides were washed 2 × 5 min in ice-cold assay buffer, followed by a dip in ice-cold distilled water, then dried under a stream of cold dry air.

D2-like receptor binding. ³H-nemonapride, ³H-spiperone and ³H-raclopride binding. Sections were preincubated for 1 hr at room temperature in assay buffer. Sections were then incubated in buffer containing either 1 nM ³H-nemonapride (specific activity 81.4 Ci/mmol), 1.2 nM ³H-spiperone (specific activity 116 Ci/mmol) in the presence of 40 nM ketanserin) or 5 nM ³H-raclopride (specific activity 86.5 Ci/mmol) for 1 hr at room temperature. Nonspecific binding was determined in the presence of 10 µM sulpiride (³H-nemonapride) or 1 µM flupenthixol (³H-spiperone and ³H-raclopride). After each radioligand assay, slides were washed (2 × 5 min) in ice-cold buffer, followed by a quick dip in ice cold distilled water then dried under a stream of cold dry air.

D₄-like receptor binding. The optimal concentration of raclopride to completely block D₂/D₃ receptors only was determined by competition experiments using unlabeled raclopride vs. ³H-nemonapride on striatal sections. Computer fitted curves showed that the raclopride curve is best fitted by assuming a two site model (P < .05, Florijn et al., 1994). A concentration of 300 nM raclopride completely displaced the high affinity binding site (D₂/D₃ receptors), but did not block 15 to 22% of residual specific binding (which may represent D₄-like receptors). Nonspecific binding was determined using different concentrations of butaclamol, sulpiride, flupenthixol and eticlopride.

D₃ receptor binding. Sections were preincubated for 1 hr in 20 mM MOPS (3-[N-morpholino]propanesulfonic acid) buffer, containing 1 mM EDTA, 10 µM pargyline and 0.1% ascorbic acid; pH 7.2 at room temperature. Sections were then incubated in buffer containing 3 nM ³H-7-OH-DPAT (specific activity 116 Ci/mmol) for 1 hr at room temperature. Nonspecific binding was determined in the presence of 1 µM eticlopride. After incubation, slides were washed (2 × 3 min) in ice-cold buffer then dried under a stream of cold dry air.

Serotonin 5-HT₂-like receptor binding. Sections were preincubated for 1 hr in assay buffer. Sections were then incubated in buffer containing 3.0 nM ³H-ketanserin in the presence of 1 µM prazosin (to block ₁-adrenergic receptors) and 100 nM tetrabenazine (to block a site associated with monoaminergic nerve terminals). Nonspecific binding was determined in the presence of 1 µM methysergide. After incubation, slides were washed (2 × 30 min) in ice-cold buffer, followed by a quick dip in ice-cold distilled water then dried under a stream of cold dry air.

Autoradiography and image analysis. For the tritium radioligands, slides together with calibrated tritium standards (Amersham) were exposed to tritium sensitive films for 2 to 4 wk at 4°C. Films were then developed and fixed in D-19 Kodak (Eastman Kodak, NY). For the iodinated radioligand, standards of 10 mm³ containing known concentrations of ¹²⁵I-sulpiride were exposed along the slides to tritium sensitive films for 3 days (striatal sections), 7 days (nigral sections) or 30 days (cortical sections) at 4°C. OD of brain regions were measured using a computer-based densitometer, image analyzer (MCID-M1, Imaging Research Inc., Ontario, Canada). Brain regions of interest were outlined (fig. 1) and the OD of these regions were measured on two images representing total binding and two images representing nonspecific binding. The left and right sides of each region were measured separately and then averaged. The OD of the sampled regions were converted to nCi/mg using the calibrated standards. The values of nonspecific binding were subtracted from total binding to yield specific binding values, which were expressed in fmol/mg tissue (tritium-labeled ligands) or nCi/10 mm³ (iodine-labeled ligand).

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Diagramatic presentation of brain regions of interest used to quantify dopamine receptor subtypes.

Statistical analysis. An overall two-way analysis of variance using the four drinking-treated groups and different brain regions was first conducted. A significant two-way (P < .05) was followed by a one-way analysis of variance and post hoc Dunnett t test to identify statistically significant differences between the four different groups across brain regions.

	Results

Top Abstract Introduction Methods Results Discussion References

³H-Nemonapride and ³H-spiperone binding sites were significantly elevated in CP (+53% and +41% respectively) and NA (+75% and +47%, respectively) of haloperidol-treated rats (Tables 1 and 2). ³H-nemonapride and ³H-spiperone binding sites were also increased, but at a smaller percentage, in CP (+32% and +19%, respectively) and NA (+34% and +20%, respectively) of chronic raclopride-treated rats. Table 2 shows an increase in ³H-spiperone binding sites in the SNpc and SNpr after chronic treatment with haloperidol (+72% and +69%, respectively) and raclopride (+38% and +48%, respectively). Another effect that was specific to ³H-spiperone, and not other radioligands, was the significant decrease of ³H-spiperone binding sites in the MPC by both haloperidol (35%) and clozapine (45%), whereas haloperidol, raclopride and clozapine treatment decreased ³H-spiperone binding in the DFC (48%, 22% and 45%, respectively) (table 2). Surprisingly, an increase in ³H-raclopride binding sites was detected only in the CP (+25%) and NA (+26%) of haloperidol-treated rats (table 3).

The number of D₄-like receptors (measured as the remaining ³H-nemonapride binding sites in the presence of 300 nM raclopride) were exclusively increased in CP and NA of all antipsychotic-treated rats (table 4). The increase was most profound after haloperidol administration (+78% and +82%, respectively), but was also significant after raclopride (+41% and +44%, respectively) and clozapine treatment (+33% and +35%, respectively).

Quantification of D2-like receptors using ¹²⁵I-sulpiride revealed a significant increase in ¹²⁵I-sulpiride binding in CP and substantia nigra of haloperidol- and raclopride-treated rats. In addition, ¹²⁵I-sulpiride binding in MPC was significantly elevated by all three antipsychotics (table 5). ³H-ketanserin binding sites were significantly reduced in the cortex (dorsolateral: 55%, medial prefrontal cortex: 50%) of clozapine-treated rats only (table 6).

Chronic typical and atypical antipsychotics did not significantly change ³H-7-OH-DPAT binding to the D₃ receptor (table 7) or ³H-SCH23390 binding to the D1-like receptors (table 8) in any brain region examined.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Changes in serotonergic and catecholaminergic receptors have been detected in postmortem brain tissue from patient's with schizophrenia (review: Seeman, 1992; Joyce, 1993), but because of previous drug treatment, it is unclear whether these changes are etiological in the disease or caused by the chronic exposure to antipsychotic drugs. Different ligands have also been used in positron emission tomographic studies to examine changes in dopamine D2-like receptors in the brains of patients with schizophrenia in vivo (Wong et al., 1986; Farde et al., 1990). Increases in labeled spiperone and nemonapride, but not labeled raclopride, binding to the striata of patients diagnosed with schizophrenia have been reported in vivo and in vitro (Wong et al., 1986; Farde et al., 1990; Seeman et al., 1993; Murray et al., 1995; Sumiyoshi et al., 1995). We report an astonishing diversity in the apparent degree of dopamine receptor up- or down-regulation depending both on the dopamine receptor radioligand used and the anatomical localization of the affected binding sites.

Both chronic (table 7) and 1 mo (Tarazi et al., in press) drug treatment failed to alter ³H-7-OH-DPAT binding in any of the brain regions examined. Similarly, Levesque et al., (1995) could not detect changes in either D₃ receptor mRNA or binding after subchronic haloperidol treatment. Using another radioligand, ¹²⁵I-sulpiride, similar results were found for cerebellar dopamine receptors (Janowsky et al., 1992), which have been reported to be exclusively of the D₃ receptor type (Sokoloff et al., 1990). Differences in receptor regulatory mechanisms, such as differences in G-protein coupling (Sokoloff et al., 1990) might be responsible for the lack in D₃ receptor up-regulation in response to chronic receptor blockade. Alternatively, endogenous dopamine has a high affinity for this receptor and may permanently occupy it, preventing the binding and subsequent receptor up-regulation by haloperidol and raclopride (Schotte et al., 1992). Thus, no conclusion can be drawn from these results concerning the role of the D₃ receptor in the treatment of schizophrenia with antipsychotic drugs. However, because all other dopamine receptor subtypes are subject to up-regulation after functional blockade, the role of D₃ receptors seems not to be important.

Dopamine D2-like receptors were not, or only modestly, up-regulated in the CP of raclopride (+13%) or haloperidol (+28%) treated rats as measured by the D₂/D₃ subtype selective ligand ³H-raclopride. However, strong up-regulation of D2-like receptors was detected using ³H-spiperone (+53%) and ³H-nemonapride (+41%), and to a lesser extent ¹²⁵I-sulpiride (+35%). This difference was not due to the presence of residual antipsychotics blocking ³H-raclopride binding sites because we found no significant change in the apparent affinity of ³H-raclopride between striatal sections of control vs. drug-treated rats (Scatchard analysis, data not shown), indicating the effectiveness of our preincubation step. Moreover, with ¹²⁵I-sulpiride [whose K_D for the D₂ site is comparable to that of raclopride (Martres et al., 1985)], considerable receptor up-regulation is evident (table 5). Finally, our data are in agreement with a recent report (Schoots et al., 1995) where, in comparison with a significant increase in ³H-spiperone binding, a marginal change in the total number of binding sites was found for ³H-raclopride binding after a high dose regimen of haloperidol (5 mg i.p./day for 1 mo). Although their binding profiles to D₂ receptors expressed in cell lines, or from striatal homogenates, are in agreement with the classification of all of these radioligands as dopamine D₂ receptor antagonists (Zahniser and Dubocovick, 1983; Terai et al., 1989), a number of differences have been reported to exist between the binding characteristics of raclopride, nemonapride, sulpiride vs. spiperone, e.g., Na⁺ dependency, differences in B_max (Zahniser and Dubocovick, 1983; Köhler and Radesater, 1986; Terai et al., 1989) that are probably due to the difference in molecular interaction with the extracellular amino acids forming the receptor pocket. Mutation analysis reveals that Asp 80, Asp 114 and His 394 are crucial for the binding of substituted benzamides, whereas only Asp 80 is necessary for ³H-spiperone binding (Daniell and Strange, 1994; Dsouza and Strange, 1995). A rather speculative explanation for our data could be that in the process of adaptation to receptor blockade by antipsychotics, subtle structural changes to the binding pocket of the receptor occur (for instance indirectly via receptor phosphorylation) differentially affecting the binding of the radioligands for the receptor, depending on the location of their binding sites at the receptor.

Alternatively, nemonapride and spiperone could be binding to an additional dopamine receptor subtype, not recognized by raclopride. Expression of the cloned D₄ receptor in cell lines revealed that raclopride differs from the former D2-like receptor antagonists in its very low affinity for the D₄ receptor (Seeman and Van Tol, 1995). The number of D₄-like receptors in subcortical postmortem tissue from patients with schizophrenia, calculated by subtracting the number of binding sites defined with ³H-raclopride (a D₂/D₃ receptor antagonist) from the total binding defined with ³H-nemonapride (a nonselective D₂/D₃/D₄ antagonist), is greatly enhanced when compared to control values (Seeman et al., 1993; Murray et al., 1995; Sumiyoshi et al., 1995). Table 4 shows that in the presence of 300 nM raclopride, the optimal concentration to block D₂/D₃ receptors only (Florijn et al., 1994), the D₄-like receptor is up-regulated by the three antipsychotics, suggesting it may be a common site for antipsychotic drug action. Moreover, these data suggest that the reported up-regulation of D₄-like receptors in post mortem tissue from patients diagnosed with schizophrenia may be caused in part by antipsychotic treatment rather than being a component of the neuropathology of schizophrenia.

Because specific D₄ receptor drugs are not yet commercially available, it is not yet known if the additional up-regulated binding sites represent D₄ receptors. Caution is warranted, because these results imply a mismatch between D₄ receptor protein and mRNA levels: moderate levels of D₄-like receptors in contrast to negligible concentrations of D₄ receptor mRNA in the striatum (Schoots et al., 1995). However, higher levels of D₄ receptor mRNA have been detected in the cortex, thus the D₄-like receptors might be synthesized in the frontal cortex and subsequently be transported to presynaptic locations on corticostriatal terminals. However, haloperidol treatment did not significantly increase D₄ receptor mRNA in the cortex, whereas both striatal D₄ receptor protein and mRNA levels were increased 2-fold (Schoots et al., 1995), suggesting a regional-specific regulation of D₄-like receptors. To complicate matters, a recent study reported low levels of D₄ receptor mRNA in human striatum and cortex, and suggested that the subset of receptors defined by the receptor binding subtraction method is not D₄ receptors (Matsumoto et al., 1996), although another study localized D₄ receptors in GABAergic neurons of primate brains (Mrzljak et al., 1996). Improvements in the development of D₄ receptor protein assays, perhaps with specific radioligands or selective antibodies, will help to resolve these contrasting findings.

The mechanism by which raclopride strongly up-regulates D₄-like receptors but not, or only marginally, D₂ receptors is unclear since raclopride has low affinity for cloned D₄ receptors in expression systems (Seeman and Van Tol, 1995). Even at the dose of raclopride used in this study, which should result in a degree of D₂ receptor occupancy approximating that of haloperidol (Van Tol et al., 1991; Kakigi et al., 1995), a considerably lower percentage of receptor up-regulation was found with ³H-raclopride compared to that caused by haloperidol. This suggests that the effect of raclopride on D₄-like receptors may be via some indirect biochemical mechanism rather than by direct blockade of D₄-like receptors.

A number of studies have found that an increase in D2-like receptor binding is not always accompanied by an elevated D₂ receptor mRNA level (Van Tol et al., 1990; Matsunaga et al., 1991; Creese et al., 1992). Instead, receptor degradation was decreased, presumably due to changes in posttranslational processing induced by persistent receptor blockade (Pich et al., 1988). It is tempting to speculate that raclopride's affinity for the dopamine D₂ receptor is not sufficiently high to slow receptor degradation, resulting in functional blockade without apparent receptor up-regulation. Clozapine's inability to up-regulate striatal D₂ receptors may be similarly explained by its lower affinity for this pool of receptors. However, this does not necessarily mean that dopamine receptor function is not affected. For instance, electrophysiological studies showed that clozapine at a similar dose reduced the number of spontaneously firing dopamine neurons in the ventral tegmental area (Chiodo and Bunney, 1983; White and Wang, 1983).

Differences between haloperidol, raclopride and clozapine can provide valuable information about their mechanism of action, for instance with reference to their liability to induce extrapyramidal side effects. Kakigi et al. (1995) investigated the occurrence of behavioral changes after chronic antipsychotic treatment with a similar drug dose and regimen as was used in our study. They reported that vacuous chewing movements (VCMs), thought to be analogous to human tardive dyskinesia (Gunne et al., 1986; Casey, 1991), are significantly elevated in chronic haloperidol and raclopride-treated rats, but low after chronic clozapine treatment. At a clinically equivalent dose, raclopride (2 mg/kg) neither induces VCMs, nor up-regulates dopamine D₂ receptors or alters glucose metabolism (Ellison et al., 1987; Tarazi et al., 1993). Recently, Shirakawa and Tamminga (1994) compared the occurrence of VCM with dopamine D1-like receptor binding in the several brain areas in rats treated for 6 mo with haloperidol. Importantly, in rats with severe, but not mild VCMs, a significant decrease in ³H-SCH23390 binding was reported in the SNpr. In our study, no reduction in ³H-SCH23390 binding was observed in the SNpr or any other brain region examined in either haloperidol-, raclopride- or clozapine-treated rats (table 8). We did not, however, subdivide our drug-treated animals according to the frequency of VCM development, and this may have contributed to the discrepancy in ³H-SCH23390 binding in SNpr between the two studies.

Our data obtained from 8 mo treatment can be compared with those from 1-mo treatment (Tarazi et al., in press), due to identical drug treatment and receptor binding methodology. A remarkably comparable overall pattern of receptor regulation is found in 1- vs. 8 mo treatment. However, in the NA, receptor up-regulation seems to develop more slowly: ³H-raclopride binding is significantly elevated after haloperidol treatment and ³H-nemonapride binding is significantly increased after raclopride treatment only in the 8 mo-treated animals. It is striking that these subtle changes are found only in the NA, a brain region involved in mediating the clinical actions of antipsychotics.

Importantly, the D₂ receptor subtype in the cortex of primates (Lidow and Goldman-Rakic, 1994) and rats (our data, Janowsky et al., 1992) appears to be a significantly up-regulated after haloperidol, raclopride or clozapine treatment. Thus, this response to antipsychotic treatment seems to occur across species. In contrast, down-regulation of the D1-like receptors in cortical areas by the three antipsychotics appears to be restricted to primates, because we failed to detect any changes in D1-like receptor binding (table 8). This discrepancy might result from the substantial differences between primate and rodent mesocortical laminar distribution and the occurrence of dopaminergic afferents and from differences in dopamine-neuropeptide co-localization (Berger et al., 1991). Also, striking differences in intrinsic activity and efficacy of several dopamine D1-like drugs between primates vs. rodents have been reported (Pifl et al., 1991) presumably due to molecular differences at the (post)receptor level (Vermeulen et al., 1994). Several behavioral and neurochemical findings in animals have suggested that D1-like receptor antagonists may display antipsychotic activity (Altar et al., 1988; Coffin et al., 1989; Ellenbroek et al., 1991; Glenthoj et al., 1993). However, recent clinical trials using another D1-like receptor antagonist, SCH 39166, failed to show that D1-like receptor antagonists have significant antipsychotic actions (Debeaurepaire et al., 1995; Karlsson et al., 1995). These clinical data, together with our results and other biochemical and electrophysiological effects (Hietala et al., 1990) argue against the involvement of D1-like receptors in mediating the clinical effects of antipsychotic drugs.

³H-Spiperone binding sites were down-regulated by the three antipsychotics in dorsolateral and medial prefrontal cortical areas. The nature of these binding sites is not clear, but serotonin 5HT₂-like receptors were ruled out because ³H-ketanserin binding was decreased only after chronic clozapine treatment (table 6, O'Dell et al., 1990). ³H-Spiperone apparently labels an additional, uncharacterized site. One study reported a similar reduction in cortical spiperone binding in schizophrenia (Arora and Meltzer, 1991). Thus, this site may be present in humans and may have been down-regulated after long-term antipsychotic therapy. Most likely, the increase in ³H-spiperone and ¹²⁵I-sulpiride binding sites in the substantia nigra of haloperidol and raclopride-treated rats represents an increase in the dopamine D₂ receptor subtype. Up-regulation of the D₃ receptor is unlikely, since ³H-7-OH-DPAT binding in the substantia nigra was not affected.

In summary, careful analysis of various brain regions using different dopamine receptor radioligands revealed a perplexing heterogeneity in dopamine receptor response to chronic antipsychotic drug treatment. Our findings suggest that D1-like and D₃ receptor subtypes are less likely to be involved in mediating the effects of antipsychotic drugs. A major aim of this study was to differentiate between structures involved in the antipsychotic action of typical and atypical antipsychotics vs. those involved in the side effects of typical antipsychotics. D2-like receptor radioligand binding in the presence of raclopride revealed the existence of an additional, formerly unnoticed, subset of dopamine D2-like receptors in striatum and nucleus accumbens. These D₄-like receptors are strongly up-regulated after both typical and atypical antipsychotic treatment, suggesting them as possible common mediators of antipsychotic drug effects. We detected a common site of drug action for all three antipsychotics, D₂ receptors in the MPC, presumably another important target for their antipsychotic effects. Similarly, all three antagonists decreased ³H-spiperone binding in the dorsolateral frontal cortex, the significance of which is not clear as this binding site has not been characterized. Finally, we detected a structure affected only by typical antipsychotics (D₂ receptor subtype binding in the striatum) that possibly plays an instrumental role in the induction of tardive dyskinesia and other neurological side effects.