Molecular Psychiatry (2005) 10, 877–883 & 2005 Nature Publishing Group All rights reserved 1359-4184/05 $30.00 www.nature.com/mp ORIGINAL RESEARCH ARTICLE contribution to the action of PCP, LSD and psychotomimetics P Seeman1,2,FKo1 and T Tallerico2 1Department of Pharmacology, University of Toronto, Toronto, Ontario, ; 2Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada

Although and ketamine are used to model a hypoglutamate theory of , their selectivity for NMDA receptors has been questioned. To determine the affinities of phencyclidine, ketamine, and LSD for the functional high-affinity state High of the dopamine D2 receptor, D2 , their dissociation constants (Ki) were obtained on [3H] binding to human cloned dopamine D2 receptors. Phencyclidine had a high High affinity for D2 with a Ki of 2.7 nM, in contrast to its low affinity for the NMDA receptor, with a 3 Ki of 313 nM, as labeled by [ H]dizocilpine on rat striatal tissue. Ketamine also had a high High affinity for D2 with a Ki of 55 nM, an affinity higher than its 3100 nM Ki for the NMDA sites. High Dizocilpine had a Ki of 0.3 nM at D2 , but a Kd of 1.8 nM at the NMDA receptor. LSD had a Ki of 2nMatD2High. Because the psychotomimetics had higher potency at D2High than at the NMDA site, the action of these drugs must have a major contribution from D2 agonism. Because these drugs have a combined action on both dopamine receptors and NMDA receptors, these drugs, when given in vivo, test a combined hyperdopamine and hypoglutamate theory of . Molecular Psychiatry (2005) 10, 877–883. doi:10.1038/sj.mp.4001682; published online 26 April 2005 Keywords: ; phencyclidine; domperidone; ketamine; NMDA receptors; psychotomimetics

While the clinical anti- actions of anti- pride at striatal D2 receptors, but has a dissociation psychotic drugs are compatible with the hyperdopa- constant of 1.5 nM at the D2High receptor when the mine hypothesis of psychosis and schizophrenia,1,2 link between dopamine D1 and D2 receptors is the psychoses caused by glutamate antagonists such blocked by the D1 antagonist SCH23390.10 The link as phencyclidine or ketamine suggest a hypo-gluta- between D1 and D2 receptors arises from several mate component in psychosis.3–5 However, phency- sources, including the colocalization of dopamine D1 clidine also lowers plasma prolactin6 and elicits and D2 receptors in at least 50% of the medium spiny rotation,7 suggesting a direct or indirect dopamine- in the striatum, the cooperation and mutual mimetic action of phencyclidine. potentiation of D1 and D2 on various Although phencyclidine and ketamine are not behaviors, and the biochemical conversion of D2 selective for glutamate NMDA receptors,8 the precise receptors from their functional high-affinity state, High affinities of these drugs for dopamine D2 receptors D2 , into their low-affinity state, D2Low, in the 10–13 need to be clarified in order to determine their presence of a D1 antagonist. This D1–D2 link has dopaminergic and non-dopaminergic components of been detected in vitro in human and rat striata,10,11 action. More specifically, although phencyclidine had and in tissue culture cells transfected with both D1 a dissociation constant of 37 000 nM at the D2 and D2,12 but has not yet been demonstrated in receptor in rat striatal homogenate,8 phencyclidine humans by means of positron emission tomography had a dissociation constant of 1.3 nM for the func- and [11C]. tional high-affinity site of the cloned D2 receptor, or However, in order to compare the potencies of the D2High receptor.9 psychotomimetics at dopamine D2 receptors with This apparent discrepancy may be resolved by the their potencies at NMDA receptors, it would be better recent finding that dopamine itself has a dissociation to avoid using SCH23390. This is because SCH23390 constant of 3000 nM when competing vs [3H]raclo- can also affect the potency of drugs at the NMDA receptor, especially since it is known that D1 and NMDA receptors interact.14–16 Correspondence: Dr P Seeman, MD, PhD, Departments of Although the blockade of D1 (by SCH23390) Pharmacology and Psychiatry, Medical Science Building, Room unmasks the functional D2High states,11 the addition 4344, University of Toronto, 1 King’s College Circle, Toronto, of SCH23390 is only needed when using [3H]raclo- Ontario, Canada M5S 1A8. E-mail: [email protected] 3 Received 23 November 2004; revised 4 March 2005; accepted 21 pride to measure D2 receptors. [ H]Domperidone, March 2005; published online 26 April 2005 however, is a selective ligand for D2 that readily Dopamine receptors, phencyclidine and ketamine P Seeman et al 878 detects D2High receptors without the addition of scintillant (Ready Solve, Beckman Co., CA, USA), SCH23390.17 and were monitored 6 h later for tritium in a Beckman Therefore, in order to obtain the potencies of L5000 scintillation spectrometer at 55% efficiency. phencyclidine, ketamine, dizocilpine and LSD at the The nonspecific binding of the [3H]ligand to the functional D2High receptor, the present study used dopamine D2 receptors was defined as that which [3H]domperidone to label D2 receptors. The results occurred in the presence of 10 mM S-

indicate that the potencies of the psychotomimetics (Ravizza, Milan). The dissociation constants (Kd)of were greater at D2High than at the NMDA receptor. [3H]domperidone for the dopamine D2 receptors (0.42 nM on the cloned D2 receptor and 0.47 nM on 17 Materials and methods the rat striatal tissue ) were obtained by the above method, using a range of 12 final concentrations of 3 We used the human cloned dopamine D2Long receptor 0.1–20 nM [ H]domperidone. The dissociation con- (in Chinese hamster ovary (CHO) cells18) . The cells stant for [3H]dizocilpine was 1.8 nM (rat striatum), were harvested by gently scraping the cells off the using 100 mM phencyclidine to define nonspecific

bottom of the Petri dish, centrifuged, resuspended in binding. The Kd value was calculated by nonlinear phosphate-buffered saline (0.9% NaCl), recentrifuged regression analysis of the [3H]ligand saturation and the pellet frozen at À701. When used, the frozen curve.19 The competition data were analyzed as pellet was thawed, and the cells suspended at 200 mg previously described;20 the program provided two protein/ml. The was homogenized for statistical criteria to judge whether a two-site fit was 5 s (Brinkmann Polytron, setting 5 (maximum on better than a one-site fit, or whether a three-site fit scale was 10)), without any further washing or was better than a two-site fit. centrifugation. The drug-induced incorporation of [35S]GTP-g-S Frozen rat were purchased (Pel-Freez, (1250 Ci/mmol; final concentration 0.2–0.3 nM) Rogers, AR, USA), partly thawed and the striata was carried out using the procedure previously removed. To prepare the tissues for the drug/[3H] described.8 ligand competition-type experiments, the striatal tissues were blotted and weighed; buffer was added Results (50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 5 mM KCl,

1.5 mM CaCl2, 4 mM MgCl2, with a final concentration Phencyclidine of 120 mM NaCl added later in the final incubation As noted above, both [3H]raclopride and [3H]domper- tubes) to yield 4 mg tissue per ml. The suspension idone selectively label D2 receptors, but [3H]domper- was homogenized with a glass-Teflon homogenizer idone is more sensitive than [3H]raclopride to the (10 up–down of a piston rotating at 500 rpm). competitive action of dopamine and other drugs and The homogenate was not washed, centrifuged or readily allows the detection of D2High sites.17 There- preincubated because previous work found that these fore, using the more readily displaceable [3H]domper- procedures resulted in a loss of 30–60% of dopamine idone, phencyclidine had a dissociation constant at 19 High D2 receptors. D2 of 2.770.8 nM (n ¼ 9) (Figure 1, top) for the [3H]Dizocilpine (or [3H]( þ )MK801; 17 Ci/mmol; human cloned D2 receptor and 4.370.8 nM for the rat

3 nM final concentration in the incubation tubes) striatal tissue (Figure 1, middle). These Ki values were was purchased from PerkinElmer Life Sciences Inc. essentially unchanged when phencyclidine was (Boston, MA, USA). [3H]Domperidone was custom tested in the presence of 100 nM SCH23390 to synthesized as [phenyl-3H(N)]domperidone (42 Ci/ occlude other receptors such as dopamine D1 recep- mmol) by PerkinElmer Life Sciences Inc., and used tors and serotonin-2 receptors that may interact with at a final concentration of 1.2–1.8 nM. the D2 receptor.10

The competition-type experiment between a drug In contrast, the phencyclidine Ki for inhibiting and the [3H]ligand for binding at the receptors was the binding of [3H]dizocilpine to rat striatal homo- carried out as follows. Each incubation tube genate was 313748 nM (n ¼ 9) (Figure 1, bottom). (12 Â 75 mm, glass) received, in the following order, The addition of 2 mM glutamate did not change

0.5 ml buffer (with a final incubate concentration of the phencyclidine Ki values. either 10 mM or, more usually, 120 mM NaCl), 0.25 ml In order to determine the functional concentration [3H]ligand and 0.25 ml of tissue homogenate. The range for phencyclidine, the effect of phencyclidine tubes, containing a total volume of 1 ml, were on the incorporation of [35S]GTP-g-S was tested on incubated for 2 h at room temperature (201C), after the human cloned D2 receptors. Phencyclidine at which the incubates were filtered, using a 12-well cell concentrations 2 nM or higher stimulated the incor- 35 harvester (Titertek, Skatron, Lier, Norway) and buffer- poration of [ S]GTP-g-S, with an EC50% (relative presoaked glass fiber filter mats (No. 7034, Skatron, effective concentration for 50% incorporation) of Sterling, VA, USA). After filtering the incubate, the 90 nM, as shown in Figure 2. filter mat was rinsed with buffer for 15 s (7.5 ml buffer). The filters were pushed out and placed in Ketamine scintillation minivials (Packard Instruments, Chicago, The data for ketamine were qualitatively similar IL, USA). The minivials received 4 ml each of to those for phencyclidine. For example, using

Molecular Psychiatry Dopamine receptors, phencyclidine and ketamine P Seeman et al 879

Figure 2 Phencyclidine, ketamine, dizocilpine and dopa- mine stimulated the incorporation of 0.3 nM [35S]GTP-g-S in

the human cloned D2Long receptors in CHO cells. Vertical bars indicate SE (n ¼ 3). The final concentration of GDP was 10 mM. Maximum incorporation of 100% corresponded to 6000 cpm/filter. Nonspecific binding in the absence of drug was 2500 cpm/filter.

[3H]domperidone on the cloned D2 receptor, ketamine High had a Ki value of 55712 nM (n ¼ 3) at D2 (Figure 3, top). In the presence of 200 mM guanilylimidodipho-

sphate, the high-affinity site (Ki of 55 nM) was eliminated, consistent with the property of a G protein-linked receptor. Ketamine stimulated the incorporation of [35S]GTP-g-S into the cloned D2 receptor at concentrations above 20 nM, with 50% incorporation at 110 nM (Figure 2). At the NMDA receptor, however, ketamine was less

potent than at the D2 receptor, having a Ki value of 31007300 nM (n ¼ 6) vs the binding of [3H]dizocil- pine (Figure 3, bottom) to rat striatal tissue. The addition of 2 mM glutamate had no effect on the

ketamine Ki values.

Dizocilpine The data for dizocilpine were also qualitatively Figure 1 Top: Using 1 nM [3H]domperidone to label similar to those for phencyclidine and ketamine, but human cloned dopamine D2 receptors, phencyclidine had dizocilpine was much more potent. Dizocilpine 7 High a Ki of 2.7 0.8 nM (n ¼ 9) at D2 . This high-affinity state inhibited the binding of [3H]domperidone to the was eliminated by the presence of 200 mM guanilylimido- cloned D2High receptor with a K of 0.370.08 nM diphosphate (open circles). Total binding was 3000 dpm/ i (n ¼ 5) (Figure 4, top). This high-affinity site was filter; nonspecific binding in the presence of 10 mMS- sulpiride was 500 dpm/filter. Middle: Using 1.4–1.75 nM abolished in the presence of 200 mM guanilylimido- 3 7 diphosphate (Figure 4, top). Matching its high [ H]domperidone, phencyclidine had a Ki of 4.3 0.8 nM 3 (n ¼ 4) at D2High. Using 100 nM SCH23390 to occlude other potency to inhibit [ H]domperidone binding in the receptors (such as dopamine D1 and serotonin-2) did not 1–10 nM range, dizocilpine also stimulated the in- 35 significantly alter the Ki values. Bottom: Using 3.3 nM corporation of [ S]GTP-g-S into the cloned D2 [3H]dizocilpine to label NMDA receptors in rat striatal receptor at concentrations between 1 and 20 nM, with homogenate, phencyclidine only recognized low-affinity 50% incorporation at 8 nM (Figure 2). 7 binding sites with a Ki of 313 48 nM (n ¼ 9). In order to confirm that dizocilpine was indeed recognizing a high-affinity site for D2, as suggested by the observation that dizocilpine inhibited the binding of [3H]domperidone at dizocilpine concentrations of

Molecular Psychiatry Dopamine receptors, phencyclidine and ketamine P Seeman et al 880

Figure 4 Schematic showing that dizocilpine recognized the high-affinity state of the dopamine D2 receptor, D2High. Using 1.3 nM [3H]domperidone to label human cloned

dopamine D2 receptors, dizocilpine had a Ki of 0.370.08 nM (n ¼ 5) at D2High. This high-affinity state was eliminated in the presence of 200 mM guanilylimidodi- phosphate. Total binding was B9000 dpm/filter; nonspecific binding in the presence of 10 mM S-sulpiride was B5000 dpm/filter.

Figure 3 Top: Using [3H]domperidone to label cloned D2

receptors, ketamine recognized a high-affinity site with a Ki value of 55712 nM (n ¼ 3). Bottom: Using [3H]dizocilpine to label NMDA receptors in rat striatal homogenate, ketamine 7 had a Ki of 3100 300 nM (n ¼ 6). Nonspecific binding was defined by 100 mM phencyclidine.

3–100 nM (Figure 4), it was necessary to do the converse, namely, to compete dopaminergic drugs vs [3H]dizocilpine. This was carried out by competing a standard D2 antagonist, , vs 0.8 nM [3H]dizocilpine on CHO cells containing transfected

D2Long receptors. The data in Table 2 show that dopaminergic drugs, two antagonists and two ago- nists, inhibited the binding of [3H]dizocilpine to the

D2Long receptors with the appropriate rank order characteristic of dopamine D2 receptor agonists, namely that n-propyl-norapomorphine is more potent Figure 5 Top: Using 1.4–1.7 nM [3H]domperidone, LSD-25 7 than . had a Ki of 2 0.3 nM for human cloned dopamine D2 3 7 receptors. Bottom: LSD-25 had no effect on the binding of The Kd for [ H]dizocilpine was 1.8 0.4 nM for the 3 NMDA sites in the striatal tissue. The presence of 3.2 nM [ H]dizocilpine. Nonspecific binding was defined by 100 mM phencyclidine. 2 mM glutamate had no effect on the dizocilpine dissociation constants.

LSD-25 stimulated incorporation was 80% of that stimulated Using [3H]domperidone, LSD-25 inhibited the bind- by dopamine. High ing at D2 with a Ki value of 270.3 nM (n ¼ 3) (Figure 5, top) on human cloned D2 receptors. LSD-25 (up to 10 000 nM) had no effect on the binding of Discussion [3H]dizocilpine (Figure 5, bottom). LSD-25 stimulated the incorporation of [35S]GTP-g-S into the cloned D2 The data show that phencyclidine, ketamine, dizo- receptor at concentrations between 10 and 200 nM, cilpine and LSD all have high affinities for the high- High with an LSD-25 EC50% of 40 nM (data not shown in affinity state of dopamine D2 receptors, D2 with Figure 2, because the data points overlap those for dissociation constants of B3, 55, 0.3 and 2 nM, phencyclidine). The total absolute amount of LSD- respectively (Table 1).

Molecular Psychiatry Dopamine receptors, phencyclidine and ketamine P Seeman et al 881 Table 1 Psychotomimetic Ki values (nM)

D2High NMDA Clone (present) Striatum (present) Cortex34 Clone34

Phencyclidine 2.770.8 313748 196 124 4.370.8a Ketamine 55712 31007300 3150 2270 7 7 Dizocilpine 0.3 0.08 (Kd) 1.8 0.4 3 3 LSD 270.3 410 000 Not done aRat striatum.

3 Table 2 Dopaminergic drug Ki values using [ H]dizocilpine index of action. Phencyclidine and ketamine 35 on dopamine D2Long receptors in CHO cells also increased the incorporation of [ S]GTP-g-S into cloned D2-containing cells8 (Figure 2). 7 Dopaminergic drug Ki ( SE) (nM) [3H]Domperidone is the optimal ligand for detect- ing the high-affinity state of dopamine D2 receptors, Antagonists 3 3 7 because [ H] and [ H]raclopride in physio- Haloperidol 0.31 0.1 logical saline (B120 mM NaCl) primarily reveal the Domperidone 2879 low-affinity state of D2 receptors.17 In fact, the Agonists dissociation constant of 1.5 nM for dopamine to 3 High N-propyl-norapomorphine-(À)1174 inhibit the binding of [ H]domperidone at D2 is Apomorphine 5607100 identical to the dissociation constant of [3H]dopamine at the dopamine D2 receptor;17 the dopamine dis- SE for triplicate measurements. sociation constants to inhibit [3H]spiperone or [3H] High B Ki ¼ IC50/(1 þ C*/Kd), where IC50 is the concentration for raclopride at D2 are invariably higher ( 20–200 nM) 3 50% inhibition of [ H]dizocilpine binding to D2Long recep- and do not correspond to the dissociation constant of 3 3 High 17 tors, C* is the concentration of [ H]dizocilpine and Kd is the [ H]dopamine at the dopamine D2 receptor. 3 High dissociation constant of [ H]dizocilpine at D2 (Figure 4). Although early experiments did not reveal the D1–D2 link in any rat striatum or in every human striatum,11 recent work, using a competition method, has consistently revealed the D1–D2 link in rat and The present data are relevant to problems asso- human striata.10 This recent work showed that the ciated with testing the hypoglutamate theory of block of D1 receptors unveiled the presence of a high- psychosis, because it has been proposed that dopa- affinity state for D2 in a dopamine/[3H]raclopride mine abnormalities in psychosis are secondary to a competition-type experiment, with a dopamine dis- primary hypofunction in glutamate neurotransmis- sociation constant of 1.5 nM. Such a high-affinity sion. The hypoglutamate theory is supported by the state for D2 is not normally detected in striata, despite fact that phencyclidine and ketamine are NMDA the fact that the high-affinity state of the dopamine D2 antagonists and elicit psychotic symptoms.3,4 receptor is the physiologically functional state of the The present findings illustrate, however, that receptor in the anterior .13 In fact, phencyclidine and ketamine also act on the high- previous work found it difficult to detect high-affinity affinity state of dopamine D2 receptors with potent sites for dopamine D2 receptors in rat or human dissociation constants of 3 and 55 nM, respectively. striatum in the presence of physiological concentra- These values compare with 313 and 3100 nM for tions of sodium , because these high-affinity sites phencyclidine and ketamine at NMDA receptors, were generally less than 5% of the population in respectively. 120 mM NaCl.21 Thus, the determination of agonist Moreover, the biphasic pattern of inhibition of dissociation constants at the high-affinity site of the [3H]domperidone binding by phencyclidine and D2 receptor was generally carried out in the absence ketamine is compatible with agonist action at the of NaCl.20 dopamine D2High state of the D2 receptors, considering Considering that all three psychotomimetics, phen- that virtually all dopamine agonists have this bipha- cyclidine, ketamine and LSD, act on dopamine D2 sic pattern, in contrast to all dopamine antagonists, and other receptors (such as serotonin receptors for which reveal a monophasic pattern of ligand inhibi- LSD), these compounds are not sufficiently selective tion.13,17,20 for testing theories of psychotomimetic action, as had In addition, because receptor agonists accelerate been thought.22 LSD is a known direct-acting partial the exchange of receptor-bound GDP for GTP, the agonist at dopamine D2 receptors, lowering the stimulation of [35S]GTP-g-S incorporation is a useful release of and stimulating rotation.23,24 The

Molecular Psychiatry Dopamine receptors, phencyclidine and ketamine P Seeman et al 882 present data that LSD stimulated the incorporation of carried out for the data in Table 2, which shows that [35S]GTP-g-S less than dopamine are consistent with standard D2 agonists (NPA or N-propyl-norapor- the partial agonist action of LSD reported by Giaco- phine-(À), and apomorphine) and antagonists (halo- melli et al.23 Phencyclidine also stimulates dopamine peridol and domperidone) inhibited the binding of 3 receptors by a direct or indirect action, lowering the [ H]dizocilpine to D2Long in the cloned CHO cells. 25 6,7 release of prolactin and causing rotation. While it is known that drug Ki values generally 3 17 Although prolactin release from hemi-pituitary depend on the [ H]ligand used, nevertheless, the Ki glands in vitro was not directly inhibited by 1 mM value for the standard D2 antagonist, haloperidol, for phencyclidine,25 lower concentrations of phencycli- example, inhibited the binding of [3H]dizocilpine at a

dine need to be tested to determine whether phency- haloperidol Ki of 0.31 nM. This value of 0.31 nM 3 clidine can directly alter the release of prolactin closely matches the [ H]haloperidol Kd of 0.4 nM and 26 release in vitro. This is because Pampillo et al found the haloperidol Ki value of 0.7 nM when using 3 35 that L-glutamate stimulates prolactin release via [ H]raclopride on cloned D2Long receptors. More ionotropic glutamate receptors, but inhibits prolactin important than the absolute Ki values for NPA and release via metabotropic glutamate receptors. It will apomorphine in Table 2 is the observation that the be essential, therefore, to test the direct in vitro action rank order of the agonists NPA and apomorphine of phencyclidine on the release of prolactin from cells was characteristic for typical dopamine D2 re- in culture, using various methods to separate phen- ceptors, namely that NPA was more potent than cyclidine’s actions on both the glutamate receptors apomorphine. and the dopamine D2 receptors. The analogue of As discussed by Javitt and Zukin,32 phencyclidine phencyclidine, ketamine, directly reduces the basal elicits a variety of behaviors in animals, including release of prolactin.27 unique discrimination of phencyclidine, hyperactiv- Because the present data are entirely in vitro, they ity, stereotypy and tranquilization, the latter effect should not be confused or admixed with separate being especially pronounced in monkeys. While it is clinical data. Nevertheless, the present in vitro data not possible to attribute each of these behaviors to are at least consistent with the clinical data on this single systems, the pharmacology of topic. That is, the clinical agonist actions of phency- phencyclidine-induced hyperlocomotion helps to clidine and LSD in eliciting psychotic symptoms are clarify the role of dopamine. In particular, the work selectively treated in the hospital emergency room by of O¨ gren and Goldstein36 has shown that three haloperidol and other dopamine D2-selective antago- dopamine antagonists (haloperidol, raclopride and nists,28,29 supporting the idea that dopamine D2 ) blocked locomotion elicited by low and receptors are the primary target for eliciting psychotic high doses of phencyclidine. Because these three symptoms. Although it has been claimed that psy- dopamine antagonists have very low affinity for the chotic symptoms elicited by ketamine are not blocked NMDA receptor,37 with raclopride and remoxipride by haloperidol,3 Giannini et al30 have found that 5 mg being highly selective for dopamine D2 receptors,38,39 of intramuscular haloperidol caused significant re- dopamine must be an important component in the ductions in psychotic symptoms.31 phencyclidine hypoglutamate model of schizophre- It is important to note that the serum concentration nia. of phencyclidine that elicits psychotomimetic action In conclusion, with phencyclidine and ketamine is between 10 and 50 nM.32 This is precisely the active having considerable potency at D2High receptors, the range of phencyclidine that directly acts on D2High use of these compounds as provoking agents in (Figure 1) and stimulates D2 (Figure 2). In contrast, animals or humans actually tests a combined dopa-

the NMDA receptor has phencyclidine Ki values of mine/glutamate theory of psychosis. 97 nM34 and 124 nM35 on rat membranes, and 196 nM35 on the human cloned NMDA receptor (Table 1). Acknowledgements A similar situation exists for ketamine, where the psychotomimetic concentration in the serum is of the We thank Dr H-C Guan for excellent technical order of 100–500 nM,8 a value that matches the assistance. Franc¸oise Ko was supported by the ketamine action (100–500 nM) at D2High (Figure 3, University of Toronto Adel Sedra Scholarship and top), but not that at the NMDA receptor, which has a an Ontario Graduate Studentship. This research was

ketamine Ki of 3100 nM for striatal tissue (Figure 3, supported by the Ontario Mental Health Foundation bottom), 760–2270 nM for brain membranes33,34 and (Regular and Special Initiatives grants), NARSAD (the 3150 nM34 for cloned NMDA receptors (Table 1). National Alliance for Research on Schizophrenia and The data in Figure 4 show that dizocilpine , to PS and FL), the CIHR (Canadian recognized a high-affinity site for the D2 receptor at Institutes of Health Research, to FL and PS), the dizocilpine concentrations of 3–100 nM when com- CPRF (Canadian Psychiatric Research Foundation, to peting vs the D2 ligand, [3H]domperidone. As men- FL and PS), NIDA (the National Institute on Drug tioned in Results, it was necessary to confirm the Abuse), the SMRI (Stanley Medical Research Insti- existence of this high-affinity site by using dopami- tute, to PS) and by donations from Dr Karolina Jus and nergic drugs to compete vs [3H]dizocilpine. This was the Medland and O’Rorke families.

Molecular Psychiatry Dopamine receptors, phencyclidine and ketamine P Seeman et al 883 References 20 Seeman P, Watanabe M, Grigoriadis D, Tedesco JL, George SR, Svensson U et al. Dopamine D2 receptor binding sites for agonists. 1 Van Rossum JM. The significance of dopamine-receptor blockade A tetrahedral model. Mol Pharmacol 1985; 28: 391–399. for the action of neuroleptic drugs. In: Brill H, Cole J, Deniker P, 21 Grigoriadis D, Seeman P. Complete conversion of brain D2dopa- Hippius H, Bradley PB (eds). Neuropsychopharmacology, Proceed- mine receptors from the high- to the low-affinity state for ings of the 5th Collegium Internationale Neuropsychopharmacolo- dopamine agonists, using sodium ions and guanine nucleotide. gicum. Excerpta Medica Foundation: Amsterdam, 1967, pp 321–329. J Neurochem 1985; 44: 1925–1935. 2 Seeman P. Dopamine receptors and the dopamine hypothesis of 22 Svenningsen P, Nomikos GG, Greengard P. Response to comment schizophrenia. Synapse 1987; 1: 133–152. on ‘Diverse psychotomimetics act through a common signaling 3 Tamminga CA, Holcom HH, Gao X-M, Lahti AC. Glutamate pathway’. Science 2004; 305:180. pharmacology and the treatment of schizophrenia: current status 23 Giacomelli S, Pamery M, Romanelli R, Cheng CY, Silvestrini B.

and future directions. Int Clin Psychopharmacol 1995; 10(Suppl Lysergic acid diethylamide (LSD) is a partial agonist of D2 3): 29–37. dopaminergic receptors and it potentiates dopamine-mediated 4 Jentsch JD, Roth RH. The neuropsychopharmacology of phency- prolactin secretion in lactotrophs in vitro. Life Sci 1998; 63: clidine: from NMDA receptor hypofunction to the dopamine 215–222. hypothesis of schizophrenia. Neuropsychopharmacology 1999; 20: 24 Meltzer HY, Fessler RG, Simonovic M, Doherty J, Fang VS. 201–225. Lysergic acid diethylamide: evidence for stimulation of pituitary 5 Goff DC, Coyle JT. The emerging role of glutamate in the dopamine receptors. Psychopharmacology (Berl) 1977; 54: 39–44. pathophysiology and treatment of schizophrenia. Am J Psychiat 25 Meltzer HY, Simonovic M, Gudelsky GA. Phencyclidine-induced 2001; 158: 1367–1377. inhibition of rat prolactin secretion: increased portal blood 6 Lozovsky D, Saller CF, Bayorh MA, Chiueh CC, Rice KC, Burke Jr dopamine. Eur J Pharmacol 1985; 110: 143–146. TR et al. Effects of phencyclidine on rat prolactin, dopamine 26 Pampillo M, Theas S, Duvilanski B, Seilicovich A, Lasaga M. receptor and locomotor activity. Life Sci 1983; 32: 2725–2731. Effect of ionotropic and metabotropic glutamate agonists and D- 7 Mele A, Wozniak KM, Hall FS, Pert A. The role of striatal aspartate on prolactin release from cells. Exp dopaminergic mechanisms in rotational behavior induced by Clin Endocrinol 2002; 110: 138–144. phencyclidine and phencyclidine-like drugs. Psychopharmacol- 27 Login IS. Direct stimulation of pituitary release by glutamate. Life ogy (Berl) 1998; 135: 107–118. Sci 1990; 47: 2269–2275. 8 Kapur S, Seeman P. NMDA receptor antagonists ketamine and PCP 28 Giannini AJ, Eighan MS, Loiselle RH, Giannini MC. Comparison of have direct effects on the dopamine D2 and serotonin 5-HT2 haloperidol and in the treatment of phencyclidine receptors—implications for models of schizophrenia. Mol Psychiat psychosis. J Clin Pharmacol 1984; 24: 202–204. 2002; 7: 837–844. 29 Giannini AJ, Nageotte C, Loiselle RH, Malone DA, Price WA. 9 Seeman P. Comment on ‘Diverse psychotomimetics act through a Comparison of chlorpromazine, haloperidol and in the common signaling pathway’. Science 2004; 305: 180. treatment of phencyclidine psychosis: DA-2 receptor specificity. 10 Seeman P, Tallerico T. Link between dopamine D1 and D2 receptors J Toxicol Clin Toxicol 1984–85; 22: 573–579. in rat and human striatal tissues. Synapse 2003; 47: 250–254. 30 Giannini AJ, Underwood NA, Condon M. Acute ketamine 11 Seeman P, Niznik HB, Guan HC, Booth G, Ulpian C. Link between intoxication treated by haloperidol: a preliminary study. Am J D1 and D2 dopamine receptors is reduced in schizophrenia and Ther 2000; 7: 389–391. Huntington diseased brain. Proc Natl Acad Sci USA 1989; 86: 31 Krystal JH, D’Souza DC, Karper LP, Bennett A, Abi-Dargham A, 10156–10160. Abi-Saab D et al. Interactive effects of subanesthetic ketamine and 12 Seeman P, Sunahara RK, Niznik HB. Receptor–receptor link in haloperidol in healthy humans. Psychopharmacology (Berl) 1999; membranes revealed by ligand competition: example for dopa- 145: 193–204. mine D1 and D2 receptors. Synapse 1994; 17: 62–64. 32 Javitt DC, Zukin SR. Recent advances in the phencyclidine model 13 George SR, Watanabe M, Di Paolo T, Falardeau P, Labrie F, Seeman P. of schizophrenia. Am J Psychiat 1991; 148: 1301–1308. The functional state of the dopamine receptor in the anterior pituitary 33 Murray F, Kennedy J, Hutson PH, Elliot J, Huscroft I, Mohnen K et is in the high-affinity form. Endocrinology 1985; 117: 690–697. al. Modulation of [3H]MK-801 binding to NMDA receptors in vivo 14 Lee FJ, Xue S, Pei L, Vukusic B, Chery N, Wang Y et al. Dual re- and in vitro. Eur J Pharmacol 2000; 137: 263–270. gulation of NMDA receptor functions by direct protein–protein in- 34 Grimwood S, Le Bourdelles BL, Atack JR, Barton C, Cockett W, teractions with the dopamine D1 receptor. Cell 2002; 111: 219–230. Cook SM et al. Generation and characterization of stable cell lines 15 Pei L, Lee FJ, Moszczynska A, Vukusic B, Liu F. Regulation of expressing recombinant human N-methyl-D-aspartate receptor dopamine D1 receptor function by physical interaction with the subtypes. J Neurochem 1996; 66: 2239–2247. NMDA receptors. J Neurosci 2004; 24: 1149–1158. 35 Seeman P. Atypical : . Can J 16 Salter MW. D1 and NMDA receptors hook up: expanding on an Psychiat 2002; 47: 27–38. emerging theme. Trends Neurosci 2003; 26: 235–237. 36 O¨ gren SO, Goldstein M. Phencyclidine- and dizocilpine-induced 17 Seeman P, Tallerico T, Ko F. Dopamine displaces [3H]domperidone hyperlocomotion are differentially mediated. Neuropsychophar- from high-affinity sites of the dopamine D2 receptor, but not macology 1994; 11: 167–177. 3 3 [ H]raclopride or [ H]spiperone in isotonic medium. Implications 37 Reynolds IJ, Miller RJ. block N-methyl-D- for human positron emission tomography. Synapse 2003; 49: aspartate receptors: similarities to the action of . Br J 209–215. Pharmacol 1988; 95: 95–102. 18 Liu IS, George SR, Seeman P. The human dopamine D2(Longer) 38 Ko¨hler C, Hall H, gren S-O, Gawell L. Specific in vitro and in vivo receptor has a high-affinity state and inhibits adenylyl cyclase. binding of 3H-raclopride. Biochem Pharmacol 1985; 34: 2251– Mol Brain Res 2000; 77: 281–284. 2259. 19 Seeman P, Ulpian C, Wreggett, KA, Wells J. Dopamine receptor 39 Hall H, Sa¨llemark M, Jerning E. Effects of remoxipride and some parameters detected by 3H-spiperone depend on tissue concentra- related new substituted on rat brain receptors. Acta tion: analysis and examples. J Neurochem 1984; 43: 221–235. pharmacol toxicol 1986; 58: 61–70.

Molecular Psychiatry