Proc. Nati. Acad. Sci. USAVol. 91, pp. 4353-4356, May 1994Neurobiology

A common action of clozapine, haloperidol, and remoxipride on D1-and D2-dopaminergic receptors in the primate cerebral cortex

(neuroleptic drugs/schizophrenia/asociatlon cortex/receptor autoradiography/receptor regulation)

MICHAEL S. LIDOW AND PATRICIA S. GOLDMAN-RAKICSection of Neurobiology, Yale University School of Medicine, New Haven, CT 06510

Contributed by Patricia S. Goldman-Rakic, January 24, 1994

ABSTRACT The potencies of the major neuroleptics usedin the treatment of schizophrenia, including haloperidol andremoxipride, correlate with their ability to bind D2-dopa-minergic receptors in subcortical structures. On the otherhand, the neuroleptic clozapine has a low affinity for these sites,and the pharmacological basis of its beneficial action is lessclear. We have found that chronic treatment with clozapine,haloperidol, and remoxipride up-regulates D2 receptors inspecific cortical areas of the rhesus monkey frontal, parietal,temporal, and occipital lobes. Of particular interest, all threeneuroleptics down-regulated D1 receptors in prefrontal andtemporal association regions-the two areas most often asso-ciated with schizophrenia. This latter finding raises the possi-bility that down-regulation of D1 receptors in prefrontal andtemporal cortex may be an important component of the ther-apeutic response to neuroleptic drugs. Further, the commoneffects of three neuroleptics with different pharmacologicalprofiles in the cerebral cortex is consistent with the idea thatthis structure is a major therapeutic target in the treatment ofschizophrenia.

The therapeutic effect of neuroleptics has generally beenassociated with their binding to the D2 class of dopaminergicreceptors in the brain (1). However, the beneficial actions ofclozapine have not been satisfactorily explained, given thatthis drug does not interact with D2 sites in most regionsexamined (2, 3). Numerous previous studies of the effects ofprolonged neuroleptic treatment on dopaminergic receptorshave focused on subcortical structures (3-8), while theregulation of cerebral cortical dopaminergic sites by similartreatment has received little attention. Such data are partic-ularly important in view of the mounting evidence for theinvolvement of the cortical dopaminergic system in schizo-phrenia (for reviews see refs. 9 and 10). To gain further insightinto the mechanism of action of clozapine, we compared itseffect on dopamine receptors in the cerebral cortex of rhesusmonkeys with those of two representative neuroleptics,haloperidol and remoxipride. Both of these drugs display ahigh affinity for the D2-dopaminergic receptor sites in theneostriatum and nucleus accumbens, though remoxipride ismore selective (11).

MATERIALS AND METHODSSixteen rhesus monkeys (Macaca mulatta), 5-7 years of age,were divided into four groups ofequal size. Group H receivedhaloperidol (0.2 mg/kg per day), group C1 received clozapine(5.2 mg/kg per day), and group R received remoxipride (3.7mg/kg per day). These doses fall within the recommendedrange for therapeutic effects in schizophrenic patients (12,13). The drugs were given orally (in fruit treats) twice a day

for 6 months to approximate the maintenance of medicationin clinical practice (14). A control group (group CO) receivedonly fruit treats. Five days after final administration ofdrugs,the animals were sacrificed, and cortical dopamine receptorswere assayed by quantitative receptor autoradiography. Thecortical regions examined are shown in Fig. 1A. They includeprefrontal association [cytoarchitectonic areas 46 and 9 ofWalker (15)], temporal association [area 21 of Brodmann(16)], primary motor [area 4 of Brodmann (16)], somatosen-sory [areas 1, 2, and 3 of Brodmann (16)], and primary visual[area 17 of Brodmann (16)] cortex.Dj-dopaminergic sites were labeled with the antagonist

[3H]SCH23390 (17, 18). An example of [3H]SCH23390 bind-ing in the prefrontal cortex is shown in Fig. 1B. Tissuesections were first preincubated for 20 min at room temper-ature in 50 mM Tris'HCl buffer, pH 7.4. They were thenincubated with 0.1-10 nM [3H]SCH23390 for 90 min at roomtemperature in 50mM Tris HCl buffer, pH 7.4, containing 120mM NaCl, 5 mM KC1, 2 mM CaCl2, 1 mM MgCl2, and 1 pMmianserin. The latter was added to block binding to 5-HT2,5-HTic, and a2 sites. After incubation, the tissue was rinsedtwice (10 min each) in ice-cold 50 mM Tris HCl buffer, pH7.4. Nonspecific binding was determined in the presence of1 jmM cis-flupentixol.D2-dopaminergic receptors, illustrated in the prefrontal

cortex (Fig. 1C), were labeled with the antagonist [1251]epi-depride (18). Sections were preincubated as described for[3H]SCH23390 binding and then incubated for 45 min at roomtemperature with 0.1-3.0 nM [1251]-epidepride in 50 mMTris HCl buffer, pH 7.4, containing 120 mM NaCl, 5 mMKCl, 2 mM CaCl2, 1 mM MgCl2, 0.1% ascorbic acid, and 0.2,uM idazoxan to prevent labeling of a2 sites. After incuba-tion, the sections were rinsed as described for [3H]SCH23390binding. Nonspecific labeling was determined in the presenceof 10 AuM (+)-butaclamol.At the end of the labeling assays, all tissue sections were

dipped in distilled water and apposed to 3H-sensitive Ultro-film (Amersham) for 6 weeks ([3H]SCH23390) or 1 week([125I]epidepride). After development for autoradiography,the tissue sections were stained with cresyl violet to allowanalysis of the cytoarchitecture. For all animals, bindingassays were repeated twice. At least five tissue sections wereprocessed for every concentration of ligand, three for totalbinding and two for nonspecific binding.The autoradiograms were analyzed with a BDS computer

system (Biological Detection Systems, Pittsburgh; ref. 17) andthe statistical analysis of saturation binding was performedwith the nonlinear curve-fitting computer programs KINETIC/EDBA/LIGAND/LOWRY from Elsevier-Biosoft (Cambridge,U.K.). The analysis is based on specific radiolabeling in tissueexposed to seven different concentrations of radioactive li-gand in incubating solutions. This number of data points issufficient to allow a relatively accurate estimation of B..(maximal binding) and Kd (steady-state dissociation constant)values for a one-site receptor model (17). Examples of satu-ration and Scatchard plots of [3H]SCH23390 and [125I]_

4353

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 20

, 202

0

4354 Neurobiology: Lidow and Goldman-Rakic

A

FIG. 1. (A) Diagram of the lateral surface of the rhesus monkeycerebral cortex. Different patterns indicate the cortical areas exam-ined in the present study. The plane of section for the slices shownin B and C is indicated by the line. (B) Color-coded autoradiogramshowing the laminar distribution of [3H]SCH23390 binding in thepresence of mianserin (Dj-dopaminergic receptors) in a coronalsection from the prefrontal cortex of an untreated rhesus monkey.(C) Color-coded autoradiogram of [125I]epidepride labeling in thepresence of idazoxan (D2-dopaminergic sites) in cortical sectionsimilar to that shown in B. Note that the [3H]SCH23390 binding isdensest in layers I, II, MIla, V, and VI, while [125I]epidepride bindingis prominent in layer V. Red coloring indicates the highest density ofbinding; black indicates the lowest density of binding. PS, principalsulcus.

epidepride are presented in Fig. 2. Bm.. values for control andeach of the treated groups were compared with two-tailedStudent t tests. The effects of neuroleptic treatments on Kdvalues were evaluated with a one-way ANOVA.

RESULTS AND DISCUSSIONSaturation analysis of [1251]epidepride binding showed that allthree neuroleptics produced statistically significant increases(47-52%) in the density of D2-dopaminergic receptor sites inall layers of the temporal association, primary motor, sOma-tosensory, and primary visual cortical regions (Fig. 3A). D2receptor sites were also increased in the prefrontal cortex butby a smaller percentage (11-18%) which did not reach sta-tistical significance (Fig. 3A). None ofthe changes in receptor

A 0 total

X0 48 ° nonspecifc

° 36

N E24_

12 18

Bound (fmol/mg prot.)

0 3 6 9I3HISCH23390 free(nM)

B 3.2 °

§.z 2.4-

000

w0 0.9 1.8 2.7

Ur%. s~~~~~~ound (final/mg prot.)

o~~~.2 0.4

[1251epidepride free (nM)

FIG. 2. Examples of saturation binding of [3H]SCH23390 (A) and[125I]epidepride (B) in the area 46 of neuroleptic naive monkey. Ineach assay, consecutive brain sections were labeled with sevenconcentrations of the radioligand as described in the text. Sectionswere then apposed to 3H-sensitive Ultrofilm and autoradiogramswere generated. Nonspecific labeling for [3H]SCH23390 was definedin the presence of 1 ,uM cis-flupentixol and for [125I]epidepride in thepresence of 10 pM (+)-butaclamol. For each concentration of theradioligand, the total binding was obtained from three consecutivebrain sections, with the fourth and fifth sections providing nonspe-cific binding. Insets represent the corresponding Scatchard plots.These plots yielded Kd = 0.14 nM andB. = 38.6 fmol/mg ofproteinfor [3H]SCH23390 and Kd = 0.029 nM and Bm..c = 2.94 fmol/mg ofprotein for [125I]epidepride.

density were accompanied by changes in the affinity for[1251]epidepride (Table 1). Our findings with haloperidol andremoxipride are in line with the widely reported D2 up-

regulation by these drugs in subcortical structures (3-7).However, the up-regulation of cortical D2 receptors byclozapine in widespread areas of the cerebral cortex differsfrom the apparent lack of such effect of clozapine on these

Table 1. Changes in apparent affinity values for binding of radioligands in the cortex of monkeys chronically treatedwith neuroleptics

Affinity, -log Kd P by one-way

Radioligand Cortical area Control Haloperidol Clozepine Remoxipride ANOVA[1251]Epidepride Prefrontal 10.17 + 0.62 10.12 + 0.43 10.05 + 0.67 10.15 ± 0.83 <0.05

Temporal 10.21 ± 0.43 10.08 + 0.99 9.91 ± 0.97 10.32 ± 0.76 <0.05Motor 10.33 ± 1.01 9.97 ± 0.96 10.02 ± 0.60 10.37 ± 0.83 <0.05Somatosensory 10.25 ± 0.31 9.86 + 0.81 9.77 ± 0.99 10.17 ± 0.69 <0.05Visual 10.18 + 0.54 10.06 ± 1.11 9.86 ± 0.88 9.99 ± 0.71 <0.05

[3H]SCH23390 Prefrontal 9.31 ± 0.75 9.97 ± 0.82 9.78 ± 0.93 8.03 ± 1.20 <0.05Temporal 9.86 + 0.88 8.13 ± 1.13 9.88 ± 1.00 9.74 ± 0.78 <0.05Motor 9.93 ± 0.89 9.85 ± 0.81 9.99 ± 0.70 8.05 ± 0.97 <0.05Somatosensory 8.02 ± 1.10 8.05 ± 0.96 9.85 ± 0.62 9.91 ± 1.11 <0.05Visual 9.65 ± 0.77 8.12 ± 1.10 9.51 ± 0.83 9.33 ± 1.26 <0.05

Since apparent affinities of the ligands have log-normal distributions, the data are presented and analyzed as -log Kd.

Proc. Natl. Acad Sci. USA 91 (1994)

Dow

nloa

ded

by g

uest

on

Nov

embe

r 20

, 202

0

Proc. Natl. Acad. Sci. USA 91 (1994) 4355

A [ 12251]EpIdepride (D2 receptors)

0._0)0

E-

m--

B

Prefrontal Temporal Motor Somato- Visualcortex cortex cortex sensory cortex

(area46) (area2l) (area4) cortex (area 17)(area 1)

[ 3H]SCH23390 (Dl receptors)

00)

E-6-

0Ex

m

Prefrontal Temporal Motor Somato- Visualcortex cortex cortex sensory cortex

(area 46) (area 21) (area 4) cortex (area 1 7)(area 1)

FIG. 3. Bar graphs representing changes in the density of D2-specific [1251]epidepride (A) and Di-specific [3H]SCH23390 (B) bind-ing in response to chronic haloperidol, clozapine, and remoxipridetreatment in different cortical regions ofrhesus monkey. The data forprefrontal areas 9 and 12 were similar to those for area 46, and thedata for somatosensory areas 2 and 3 were similar to those for area1. The B.,, value for each column is an average of four animals.Error bars are ± SEM. Statistically significant differences betweencontrol and treated groups (two-tailed Student t test; P < 0.05) aremarked by asterisks. Note that all three neuroleptics up-regulate the[125I]epidepride binding in most cortical areas. PH]SCH23390 bind-ing is down-regulated only in the prefrontal and temporal regions.

sites in striatum and the nucleus accumbens (3-5). Anincrease in the density of D2 receptors in response to cloz-apine treatment has also been observed in the medial pre-frontal cortex ofthe rat (19). It is ofinterest that the prefrontalcortex was the only area in our study where the D2 receptorwas least affected by neuroleptics, including clozapine, per-haps indicating a difference between primates and rodents.The similarity in the effects of clozapine, haloperidol, and

remoxipride on cortical D2-dopaminergic receptors may berelated to the unique pharmacological and electrophysiolog-ical properties of the dopamine neurons that project to thecerebral cortex as compared with those of the nigrostriataland mesolimbic dopamine systems (20). Also, the cortex hasa substantial proportion of the D4 subtype of the family of D2receptors (21), whereas the striatum and nucleus accumbenscontain mainly D2A sites (22). Clozapine has a high affinity forthe D4 receptor subtype (21) and could easily up-regulate it,while the regulation of D2A receptors by clozapine would be

negligible due to its low affinity for these sites (21). Indeed,our preliminary binding studies in cell cultures expressingD4-dopaminergic receptors indicate that under the conditionsof the present experiment, [125I]epidepride may label D4 sites(M.S.L., unpublished results). However, it remains to bedetermined whether the general increase in the density of theD2 sites in cortical areas described here is related in any wayto the unique composition of D2 dopaminergic receptorsubtypes in the cortex.A major finding that could not have been predicted is that

all three neuroleptics produced decreases in the density ofD1-dopaminergic receptor sites. These reductions were sub-stantial (30-34%) and were found exclusively in the prefron-tal and temporal association cortices (Fig. 3B) and affected alllayers of these cortical areas (Fig. 4). No significant effectswere observed in the somatosensory, visual, or motor areas(Fig. 3B), nor were alterations in the affinity of D1 receptorsto [3H]SCH23390 observed in any of the cortical areasstudied (Table 1). It is unlikely that the down-regulation ofthese receptors results from competitive binding by residualneuroleptics, as the animals were not sacrificed until 5 daysafter the last administration of drug. Indeed, gas chromato-graphic detection analysis of haloperidol and clozapine incortical tissue, conducted for us by National Medical Serv-ices (Willow Grove, PA), failed to detect concentrations ofthese neuroleptics in the prefrontal and temporal cortices oftreated animals above 0.01 nmol/g of tissue, values notsufficient to interfere with [3H]SCH23390 binding (11). More-over, any significant degree of competition with residualneuroleptics would have affected the Kd values recorded for[3H]SCH23390 binding in the tissue of the treated animals.This, however, was not the case (Table 1).The mechanisms for decrease in the density of cortical

D1-dopaminergic receptor sites by chronic neuroleptic treat-ment are now open for investigation. For example, down-regulation of D1 sites may be a compensatory reaction to achronic increase in cortical dopamine release resulting fromthe blockade of D2 pre- and postsynaptic sites by neurolepticdrugs (23, 24). It should be mentioned, however, that the

03 Control 0 Clozapine

50c

400.O 40

30

E 20

210m

0I II IIla mb IV V VI

Layers

FIG. 4. Bar graph representing changes in the density of D1-specific [3H]SCH23390 binding in response to chronic clozapinetreatment in individual layers of area 46 in the prefrontal cortex. Inall cases, B, values are averaged for four animals. Error bars are+ SEM. Receptor densities significantly different from controlvalues (two-tailed Student t test; P < 0.05) are marked by asterisks.Note that clozapine significantly down-regulates [3H]SCH23390binding in all layers of the prefrontal cortex. Haloperidol andremoxipride have similar effects on [3H]SCH23390 receptors in alllayers of the same cortical area.

Neurobiology: Lidow and Goldman-Rakic

Dow

nloa

ded

by g

uest

on

Nov

embe

r 20

, 202

0

4356 Neurobiology: Lidow and Goldman-Rakic

effect of chronic neuroleptic treatment on cortical dopami-nergic release after treatments as prolonged as that used inthe present study is not presently known. It is possible,therefore, that very prolonged exposure to neuroleptics mayproduce depolarization block of the cortical dopaminergicinnervation (25). Depolarization block can also down-regulate D1 receptors (26), although this down-regulationwould be related to the elimination of phasic dopaminerelease (27). Alternatively, the decrease in D, receptor den-sity may be secondary to a neuroleptic-induced increase inthe level of responsiveness ofD2-mediated adenylate cyclaseto dopamine (6). As a strong interaction between D1 and D2second messenger systems is well established (28, 29), it isconceivable that an increase in the sensitivity of D2-associated adenylate cyclase would be accompanied by anincrease in the sensitivity of the D1-coupled enzyme which,in turn, would downregulate D1 receptors (8, 30).Our findings that haloperidol, remoxipride, and clozapine

have virtually identical effects in all cortical areas examinedmay provide the missing link in understanding the therapeuticactions of these very different neuroleptic drugs. It seemsparticularly relevant that the D1 down-regulation was foundselectively in the prefrontal and temporal cortical regionswhich, in rhesus monkeys, have especially high concentra-tions of dopamine receptors among cortical areas studied(31). In humans, these areas are also critically involved incognitive and affective functions (for review see refs. 32 and33) and are the areas most often compromised in schizophre-nia (9, 34, 35). All of these considerations suggest that the D,receptors in these regions may be one of the major sites towhich neuroleptic treatment is and/or should be directed.

Note Added in Proof. Since this paper was submitted for publication,an article by Malmberg et al. (36) came to our attention. It shows thatclozapine binds to the short isoform of the D2A-dopaminergic recep-tor with an affinity comparable to that for D4 receptors. It has alsobeen shown (37) that while the majority of the striatal D2A receptorsbelong to the long type, the short isoform is predominant among D2Areceptors in the cerebral cortex (particularly in those areas where wedetected neuroleptic-induced increase in D2A receptors). Thus, theclozapine-induced up-regulation of D2 sites, detected in the presentstudy, may reflect an interaction of this drug with the D4 receptor,the short isoform of the D2A receptor, or both.

This work was supported by National Institute of Mental HealthCenter Grant P50-MH44866-O5. Haloperidol was provided by Mc-Neil Pharmaceutical (Spring House, PA), clozapine was provided bySandoz Pharmaceutical (East Hanover, NJ), and remoxipride wasreceived from Astra Pharmaceutical (Sodertalje, Sweden).

1. Creese, I., Burt, D. R. & Snyder, S. H. (1976) Science 192,481-483.

2. Seeman, P., Chau-Wong, M., Tedesco, J. & Wong, K. (1975)Proc. Natl. Acad. Sci. USA 72, 4376-4380.

3. Rupniak, N. M. J., Hall, M. D., Mann, S., Fleminger, S.,Kilpatrick, G., Jenner, P. & Marsden, C. D. (1985) Biochem.Pharmacol. 34, 2755-2763.

4. O'Dell, S. J., Hoste, G. J., Widmark, C. B., Shapiro, R. M.,Potkin, S. G. & Marshall, J. F. (1990) Synapse 6, 146-153.

5. See, R. E., Toga, A. W. & Ellison, G. (1990) J. Neural Transm.82, 93-109.

6. Memo, M., Pizzi, M., Missale, C., Carruba, M. 0. & Spano,P. F. (1987) Neuropharmacology 26, 477-480.

7. Hall, H. & Sallemark, M. (1987) Pharmacol. Toxicol. 60,359-363.

8. Laruelle, M., Jaskiw, G. E., Lipska, B. K., Kolachana, J. E.& Weinberger, D. R. (1992) Brain Res. 575, 47-56.

9. Weinberger, D. R. (1987) Arch. Gen. Psychiatr. 44, 660-669.10. Davis, K. L., Kahn, R. S., Grank, K. 0. & Davidson, M.

(1991) Am. J. Psychiatr. 148, 1474-1486.11. Seeman, P. (1990) Acta Psychiatr. Scand. 82 (Suppl. 358),

14-20.12. Pflug, B., Bartels, M., Baner, H., Bunse, J., Gallhofer, B.,

Haas, S., Konzow, W. T. H., Lkieser, E., Kufferle, B., Stein,D., Steinberger, K. A. & Weiselmann, G. (1990) Acta Psychi-atr. Scand. 82 (Suppl. 358), 142-146.

13. Dowd, A. L., ed. (1993) Physicians Desk Reference (MedicalEconomics, Montvall, NJ), 47th Ed.

14. Hyman, S. E. & Arana, G. W. (1987) Handbook ofPsychiatricDrug Therapy (Little-Brown, Boston).

15. Walker, A. E. (1940) J. Comp. Neurol. 73, 59-86.16. Brodmann, K. (1905) J. Psychol. Neurol. (Leipzig) 9, 177-226.17. Lidow, M. S., Goldman-Rakic, P. S., Gallager, D. W. &

Rakic, P. (1991) Neuroscience 40, 657-671.18. Lidow, M. S. (1993) in Receptor Autoradiography: Principles

and Practice, eds. Wharton, J. & Polak, J. M. (Oxford Univ.Press, Oxford, U.K.), pp. 217-236.

19. Janowsky, A., Neve, K. A., Kinzie, J. M., Taylor, B., De-Paulis, T. & Belknar, J. K. (1992) J. Pharmacol. Exp. Ther.261, 1282-1290.

20. Bannon, M. J. & Roth, R. H. (1983) Pharmacol. Rev. 35,53-67.

21. VanTol, H. H. M., Bunzow, J. R., Guan, H. C., Sunahara,R. K., Seeman, P., Niznik, H. B. & Civelli, 0. (1991) Nature(London) 350, 610-614.

22. Meador-Woodruff, J. H., Mansour, A., Healy, D. J., Kuehn,R., Zhou, Q. Y., Bunzow, J. R., Akil, H., Civelli, 0. &Watson, S. J. (1991) Neuropsychopharmacology 5, 231-242.

23. Scatton, B., Glowinski, J. & Julou, L. (1987) Brain Res. 109,184-189.

24. Bacopoulos, N. C., Spokes, E. G., Bird, E. & Roth, R. H.(1979) Science 205, 1405-1407.

25. Essig, E. C. & Kilpatrick, I. C. (1991) Psychopharmacology104, 194-200.

26. Gerfen, C. R., Engler, T. M., Mahan, L. C., Susel, Z., Chase,T. N., Monsma, F. J. & Sibley, D. R. (1990) Science 250,1429-1431.

27. Grace, A. A. (1991) Neuroscience 41, 1-24.28. Seeman, P., Niznik, H. B., Guan, H. C., Booth, G. & Ulpian,

C. (1989) Proc. Natl. Acad. Sci. USA 86, 10156-10160.29. Strange, P. G. (1991) Trends Pharmacol. Sci. 12, 48-49.30. Creese, I. & Hess, E. S. (1986) Clin. Neuropharmacol. 9

(Suppl. 4), 14-16.31. Lidow, M. S., Goldman-Rakic, P. S., Rakic, P. & Innis, R. B.

(1989) Proc. Natl. Acad. Sci. USA 86, 6412-6416.32. Fuster, J. M. (1980) The Prefrontal Cortex (Raven, New York).33. Goldman-Rakic, P. S. (1987) in Handbook ofPhysiology, ed.

Plum, F. (Am. Physiol. Soc., Bethesda, MD), Section 1, Vol.5, pp. 373-417.

34. Weinberger, D. R. (1988) Trends Neurosci. 11, 367-370.35. Shenton, M. E., Kikins, R., Jolesz, F. A., Pollak, S. D.,

LeMay, M., Wible, C. G., Hokama, H., Martin, J., Metcalf,D., Coleman, M. & McCarley, R. W. (1992) N. Engl. J. Med.327, 604-612.

36. Malmberg, A., Jackson, D. M., Eriksson, A. & Mohell, N.(1993) Mol. Pharmacol. 43, 749-754.

37. O'Malley, K. L., Mack, K. J., Gangelman, K. Y. & Tood,R. D. (1990) Biochemistry 29, 1367-1371.

Proc. Natl. Acad. Sci. USA 91 (1994)

Dow

nloa

ded

by g

uest

on

Nov

embe

r 20

, 202

0