changes in serotonin2a and gabaa receptors in schizophrenia : studies on the human dorsolateral...

Changes in Serotonin2A and GABAA Receptors inSchizophrenia: Studies on the Human

Dorsolateral Prefrontal Cortex

Brian Dean, Tabasum Hussain, Wendy Hayes, Elizabeth Scarr, Susie Kitsoulis, Christine Hill,Kenneth Opeskin, and David L. Copolov

Rebecca Cooper Research Laboratories, Division of Molecular Schizophrenia,Mental Health Research Institute, Parkville, Victoria, Australia

Abstract: Having shown a decrease in serotonin2A re-ceptors in the dorsolateral prefrontal cortex (DLPFC) fromschizophrenic subjects, we have now determined if thischange was reflective of widespread changes in neuro-chemical markers in DLPFC in schizophrenia. In Brod-mann’s area (BA) 9 from 19 schizophrenic and 19 controlsubjects, we confirmed a decrease in the density of[3H]ketanserin binding to serotonin2A receptors in tissuefrom the schizophrenic subjects [39 6 3.3 vs. 60 6 3.6fmol/mg estimated tissue equivalents (ETE); p , 0.005].In addition, the density of [3H]muscimol binding toGABAA receptors was increased in the schizophrenicsubjects (526 6 19 vs. 444 6 28 fmol/mg ETE; p , 0.02).[3H]YM-09151-2, N-[1-(2-thienyl)cyclohexyl]-3,4-[3H]pip-eridine, [3H]SCH 23390, [3H]mazindol, and NG-nitro-L-[3H]arginine binding to BA 9 did not differ betweengroups, and there was no specific binding of [3H]raclo-pride or 7-hydroxy-2-(di-n-[3H]propylamino)tetralin to BA9 from either cohort of subjects. This suggests the den-sity of dopamine D1-like and NMDA receptors, the dopa-mine transporter, and nitric oxide synthase activity arenot altered in BA 9 from schizophrenic subjects. Theselective nature of the changes in serotonin2A andGABAA receptors in DLPFC could indicate that thesechanges are involved in the pathology of schizophrenia.Key Words: Schizophrenia—Dopamine receptors—Serotonin2A receptor—Dorsolateral prefrontal cortex—GABAA receptors—Nitric oxide synthase.J. Neurochem. 72, 1593–1599 (1999).

The location and nature of the biological changes thatunderlie schizophrenia and contribute to the pathology ofthe illness have yet to be elucidated. However, variouslines of evidence suggest a change in the functioning ofthe dorsolateral prefrontal cortex (DLPFC) is likely tocause some symptoms of the illness (Egan and Wein-berger, 1997). At the molecular level, the interactionsbetween the inhibitory dopaminergic, serotonergic, andGABAergic neurons and the excitatory glutamatergicneurons have been suggested to be central to manyfunctions of the cortex (Fuster, 1989). It is significant

that all these neurotransmitter systems have been pro-posed to be involved in the pathology of schizophrenia(Meltzer, 1987). It would therefore seem that changes inthese neurotransmitter systems in the DLPFC in schizo-phrenic subjects could be particularly important in thepathology of the illness.

Current theories on possible molecular changes in thebrain of subjects with schizophrenia frequently refer toresults from studies using brain tissue obtained at au-topsy. One such study showed a decrease in the densityof [3H]lysergic acid diethylamine ([3H]LSD) binding toDLPFC from schizophrenic subjects (Bennett et al.,1979). Although not replicated (Whitaker et al., 1981),this study has proven to be the first direct indication thatchanges in serotonin (5-HT) receptors were present inDLPFC from schizophrenic subjects. More recent studieshave shown a decrease in the density of [3H]ketanserinbinding in the DLPFC from schizophrenic subjects (Mitaet al., 1986; Arora and Meltzer, 1991; Hashimoto et al.,1993; Dean and Hayes, 1996), a finding that again hasbeen disputed (Joyce et al., 1993). [3H]Ketanserin bindswith high affinity and specificity to the 5-HT2A receptor(Leysen et al., 1982); therefore, most studies using thisradioligand would suggest that there is a decrease in5-HT2A receptors in the DLPFC from schizophrenic sub-jects.

There have also been reports of changes in otherneurotransmitter systems in the DLPFC from subjects

Received August 26, 1998; revised manuscript received November25, 1998; accepted November 25, 1998.

Address correspondence and reprint requests to Dr. B. Dean, Divi-sion of Molecular Schizophrenia, Mental Health Research Institute,Locked Bag 11, Parkville, Victoria 3052, Australia.

Abbreviations used:BA, Brodmann’s area; DA, dopamine; DLPFC,dorsolateral prefrontal cortex; DOI, duration of illness; ETE, estimatedwet weight tissue equivalents; FT, freezer time; 5-HT, serotonin; LSD,lysergic acid diethylamine; NOS, nitric oxide synthase; NSB, nonspe-cific binding; [3H]7-OH-DPAT, 7-hydroxy-2-(di-n-[3H]propylamino)-tetralin; PMI, postmortem interval; [3H]TCP, N-[1-(2-thienyl)cyclo-hexyl]-3,4-[3H]piperidine.

1593

Journal of NeurochemistryLippincott Williams & Wilkins, Inc., Philadelphia© 1999 International Society for Neurochemistry

with schizophrenia. An increase in the density ofGABAA receptors in the DLPFC from schizophrenicsubjects has been reported (Hanada et al., 1986; Benes etal., 1996). By contrast, studies either have failed to showa change (Pandey et al., 1997) or have shown a reduction(Squires et al., 1993) in the number of benzodiazepinebinding sites on the GABAA receptor in schizophrenia.In addition, the numbers of both dopamine (DA) D1-likereceptors (Knable et al., 1996) and NMDA receptors(Ishimaru et al., 1992; Simpson et al., 1992) have beenshown to be increased in DLPFC from schizophrenicsubjects. Thus, current data cannot exclude changes inGABAergic, glutamatergic, or dopaminergic neurons inthe frontal cortex from subjects with schizophrenia.

Nitric oxide has been suggested to be important in thecontrol of dopaminergic, GABAergic, and serotonergicfunction in the brain (Lorrain and Hull, 1993; Lin et al.,1995; Ohkuma et al., 1995). Changes in activity ofNADPH diaphorase or nitric oxide synthase (NOS) havebeen reported in the temporal lobe and hypothalamusfrom subjects with schizophrenia (Akbarian et al., 1994;Bernstein et al., 1998). This raises the possibility thatchanges in this enzyme may also be present in theDLPFC from subjects with schizophrenia. It is importantthat it is unlikely that direct evidence could be obtainedfor a change in levels of nitric oxide itself in DLPFCfrom schizophrenic subjects. However, identifying con-comitant changes in dopaminergic, serotonergic, andGABAergic markers in association with changes inmarkers of nitric oxide activity, such as NOS, couldstrongly point to an involvement of nitric oxide in thepathology of schizophrenia.

In summary, there is a growing body of evidencesuggesting that components of the 5-HT, DA, GABA,and glutamate systems are altered in the DLPFC fromschizophrenic subjects. In addition, it seems that changesin the nitric oxide system could be a modulatory factorinvolved in such changes. Although there have beenseveral studies examining important neurochemicalmarkers in the DLPFC from subjects with schizophreniaindividually, there appears to have been no systematicstudy looking at multiple, potentially interactive, neuro-chemical markers within the DLPFC. We have thereforebegun studies to determine if there are interrelatedchanges in important neurotransmitter markers in theDLPFC in schizophrenia.

MATERIALS AND METHODS

Materials7-Hydroxy-2-(di-n-[3H]propylamino)tetralin ([3H]7-OH-

DPAT), [3H]SCH 23390,NG-nitro-L-[3H]arginine HCl, [3H]-Micro-scales, and Hyperfilm-3H were obtained from Amer-sham Australia Pty. Ltd. (Sydney, Australia). [3H]Ketanserin,[3H]mazindol, [3H]raclopride, [3H]YM-09151-2, [3H]musci-mol, and N-[1-(2-thienyl)cyclohexyl]-3,4-[3H]piperidine([3H]TCP) were obtained from New England Nuclear viaAMRAD Biotech (Melbourne, Australia). All other chemicalswere obtained from Sigma Aldrich Pty Ltd. (Castle Hill, NewSouth Wales, Australia).

Tissue collectionWith the ethical approval of the Human Ethics Committee of

the Victorian Institute of Forensic Medicine, Brodmann’s area(BA) 9 was collected at autopsy from the left brain hemisphereof 19 schizophrenic subjects who were initially diagnosed ashaving schizophrenia in a police report (Table 1). Tissue wasalso collected from the same brain regions from 19 nonschizo-phrenic individuals (controls) who were matched for sex andwho were of a similar age to the schizophrenic subjects (Table1). Because of this collection process there was no difference inthe sex distribution or mean age of the schizophrenic andcontrol subjects (Table 1). In cases where death was witnessed,the time between death and autopsy was taken at the postmor-tem interval (PMI; Table 1). Where death was not witnessed,tissue was only taken from individuals who had been seen aliveup to 5 h before being found dead. In those cases, the PMI wasthe interval that was halfway between the donor being founddead and being last seen alive. There was no difference in themean PMI for the tissue from schizophrenic and control sub-jects (Table 1).

In all cases, the cadavers were refrigerated within 5 h ofbeing found, and tissue was rapidly frozen to270°C within 30min of autopsy and stored until required. The tissue storagetime at 270°C to the time when the tissue was sectioned forthis study is shown as the freezer time (FT; Table 1) and did notdiffer for tissue from the schizophrenic and control subjects. Inaddition, to attempt to address any effects of agonal state on thetissue collected, the pH of the brain tissue was measured asdescribed previously (Kingsbury et al., 1995) and did not differbetween the two cohorts (Table 1).

Diagnostic evaluationFollowing the collection of tissue, an extensive review of

case histories for the subjects with a provisional diagnosis ofschizophrenia was carried out using a structured instrument(Hill et al., 1996a,b) by a senior psychologist and psychiatrist.Following the case history review the diagnosis of schizophre-nia was confirmed according to theDiagnostic and StatisticalManual of Mental Disorders(3rd edit., revised) criteria (Amer-ican Psychiatric Association, 1987). After the case historyreview the duration of illness (DOI) for each schizophrenicsubject was calculated as the time from first hospital admissionto death, and the final recorded dose of antipsychotic drug wasconverted to chlorpromazine equivalents (Foster, 1998).

Measurement of radioactive drug binding to tissuesections by autoradiography

Ethical approval.Before commencement, approval for thesestudies was obtained from the Western Health Care NetworkBehavioural and Psychiatric Human Research Committee andthe Western Health Care Network Behavioural and PsychiatricHuman Ethics Committees.

Autoradiography.Radioligand binding to sections from BA9 was measured at a single concentration at three times thebinding KD, or higher, for each radioligand. Thus, this studyuses single-point saturation analysis where the specific bindingof the radioligand provides a good estimate of the density ofbinding sites.

All studies involved the use of frozen sections and autoradiog-raphy. Hence, 5-3 20-mm frozen tissue sections (220°C) werecut from BA 9 from each subject for each of the nine radioligandbinding assays and mounted on gelatin-coated slides. After expo-sure to radioligand in the absence (total binding; three sections) orpresence [nonspecific binding (NSB); two sections] of nonradio-active drugs, all sections were washed twice in ice-cold assay

J. Neurochem., Vol. 72, No. 4, 1999

1594 B. DEAN ET AL.

TA

BLE

1.D

em

og

rap

hic

da

taa

nd

ne

uro

tra

nsm

itte

rre

cep

tor

an

dtr

an

spo

rte

rd

en

sitie

sfo

rth

esc

hiz

op

hre

nic

an

dco

ntr

olsu

bje

cts

fro

mw

ho

mB

A9

wa

so

bta

ine

da

ta

uto

psy

Sex

Age

(yea

rs)

DO

I(y

ears

)C

ause

ofde

ath

Ant

ipsy

chot

icdr

ugs

Fin

alre

cord

eddr

ugdo

se(C

hlor

.E

qu.)a

Tis

sue

pHP

MI

(h)

FT

(yea

rs)

fmol

/mg

ofE

TE

5-H

T 2A

GA

BA

AN

MD

AD

1Y

MD

AT

NO

S

Sch

izop

hren

ia1

M53

30In

test

inal

isch

emia

Flu

phen

azin

e,ch

lorp

rom

azin

e1,

700

5.98

373.

717

518

46.0

7.5

15.0

9.9

2.0

2M

4721

Sui

cide

:m

ultip

lein

jurie

sC

hlor

prom

azin

e,ha

lope

ridol

1,40

06.

5242

2.8

1860

378

.013

.017

.048

.07.

0

3M

6736

Pne

umon

iaF

luph

enaz

ine

756.

4621

3.0

3642

026

.020

.036

.018

.03.

04

M22

3P

eric

ardi

tisT

rifluo

pera

zine

,flu

pent

hixo

l45

06.

0737

2.5

4448

169

.024

.023

.07.

02.

0

5F

4024

Sui

cide

:po

ison

Flu

phen

azin

e55

05.

7161

3.2

6754

944

.032

.014

.06.

710

.06

F81

40A

spira

tion/

food

Trifl

uope

razi

ne10

06.

3125

2.5

3459

043

.013

.011

.016

.09.

07

M46

12Is

chem

iche

art

dise

ase

Hal

oper

idol

160

5.53

422.

342

564

43.0

2.5

77.0

48.0

9.0

8M

6947

Isch

emic

hear

tdi

seas

eT

rifluo

pera

zine

100

6.38

452.

330

574

65.0

19.0

37.0

27.0

2.0

9M

7153

Asp

iratio

n/fo

odT

hior

idaz

ine

150

6.45

482.

322

471

35.0

13.0

63.0

32.0

10.0

10F

342

Sui

cide

:dr

owni

ngR

ispe

ridon

e1,

200

6.23

422.

127

527

10.0

16.0

51.0

28.0

4.0

11M

6138

Isch

emic

hear

tdi

seas

eN

odr

ugs

for

2ye

ars

N/A

6.46

382.

155

647

58.0

19.0

29.0

43.0

4.0

12M

223

Sui

cide

:dr

ugov

erdo

seC

hlor

prom

azin

e20

06.

1737

2.1

4649

278

.031

.02.

027

.03.

013

F68

40Is

chem

iche

art

dise

ase

Trifl

uope

razi

ne40

05.

7342

2.2

4965

035

.019

.036

.018

.010

.014

M42

15S

uici

de:

drow

ning

Flu

pent

hixo

l,ch

lorp

rom

azin

e61

06.

2635

2.2

3752

943

.023

.044

.023

.03.

0

15M

384

Men

ingo

ence

phal

opat

hyC

loza

pine

506.

0250

2.0

5633

348

.015

.03.

05.

93.

016

M53

7A

spira

tion/

food

Chl

orpr

omaz

ine

200

6.23

432.

359

488

57.0

14.0

12.0

20.0

9.0

17M

6520

Per

fora

tega

stric

ulce

rF

luph

enaz

ine

400

6.56

413.

445

428

46.0

11.0

30.0

21.0

5.0

18M

7344

Chr

onic

card

iac

failu

reN

odr

ugs

for

5ye

ars

N/A

6.07

733.

730

620

4.0

18.0

12.0

8.8

2.0

19M

3215

Sui

cide

:ca

rbon

mon

oxid

epo

ison

ing

Hal

oper

idol

670

6.05

171.

929

510

26.0

13.0

17.0

24.0

2.0

Mea

n6

SE

M52

64.

124

63.

849

5612

06.

176

0.06

416

2.9

2.66

0.1

396

3.3

5266

1945

64.

617

61.

728

64.

623

63.

05.

260.

7C

ontr

ol1

M53

—Is

chem

iche

art

dise

ase

——

6.56

452.

638

458

15.0

23.0

28.0

20.0

7.0

2M

50—

Isch

emic

hear

tdi

seas

e—

—6.

4369

2.7

5549

747

.020

.041

.00.

95.

03

M65

—R

uptu

red

thor

acic

aort

a—

—6.

4721

2.7

6445

737

.015

.021

.08.

49.

04

M22

—E

xsan

guin

atio

n—

—6.

5851

2.4

3931

750

.025

.034

.035

.02.

05

F38

—Is

chem

iche

art

dise

ase

——

6.14

512.

883

217

43.0

47.0

8.0

3.5

1.0

6F

80—

Isch

emic

hear

tdi

seas

e—

—6.

2855

2.4

6845

841

.021

.014

.031

.04.

07

M46

—A

cute

myo

card

ial

infa

rctio

n—

—6.

0847

2.2

6540

446

.013

.024

.029

.06.

0

8M

69—

Isch

emic

hear

tdi

seas

e—

—6.

2338

1.9

3633

956

.012

.07.

039

.03.

09

M72

—P

yelo

neph

ritis

——

5.60

451.

751

583

55.0

9.0

34.0

10.0

5.0

10F

33—

Mul

tiple

inju

ries

——

6.41

421.

369

494

44.0

24.0

2.0

69.0

8.0

11M

63—

Isch

emic

hear

tdi

seas

e—

—6.

3738

1.9

6746

887

.011

.053

.016

.03.

012

M22

—Ia

trog

enic

hem

orrh

age

——

6.39

622.

063

594

33.0

24.0

29.0

13.0

3.0

13F

66—

Acu

tem

yoca

rdia

lin

farc

tion

——

6.37

431.

588

549

99.0

23.0

48.0

13.0

5.0

14M

43—

Dro

wni

ng—

—6.

2545

1.3

5367

078

.060

.05.

016

.04.

015

M42

—C

ardi

omeg

aly

——

6.34

631.

667

209

66.0

15.0

1.0

6.2

8.0

16M

48—

Isch

emic

hear

tdi

seas

e—

—5.

9952

1.4

7455

458

.016

.010

.055

.07.

017

M35

—A

cute

myo

card

ial

infa

rctio

n—

—6.

4027

3.2

7739

522

.015

.053

.037

.04.

0

18M

65—

Isch

emic

hear

tdi

seas

e—

—6.

5641

3.4

5436

66.

410

.06.

021

.05.

019

M42

—C

oron

ary

arte

rial

thro

mbo

sis

——

6.34

632.

134

413

50.0

10.0

21.0

39.0

3.0

Mea

n6

SE

M50

63.

86.

306

0.05

476

2.8

2.26

0.1

606

3.6

4446

2849

65.

321

63.

023

64.

024

64.

24.

860.

9

N/A

,no

tav

aila

ble;

YM

,[3H

]YM

-091

51-2

;D

AT

,D

Atr

ansp

orte

r.a

Chl

or.

Equ

.,ch

lorp

rom

azin

eeq

uiva

lent

s.


1595NEUROCHEMICAL MARKERS IN SCHIZOPHRENIA

buffer, dipped into ice-cold distilled water, and then thoroughlydried. Sections of tissue from the matched schizophrenic andcontrol subjects, as well as a set of high and low autoradiographic[3H]Micro-scales, were then apposed to a sheet of Hyperfilm-3Huntil an image of appropriate optical density was obtained. Expo-sure time related to both the density of binding sites and thespecific activity of the radioligand used. Resulting images werethen analyzed with reference to the images from the [3H]Micro-scales using an MCID image analysis system. Results could thenbe expressed as dpm per milligram of estimated wet weight tissueequivalents (ETE) and converted to femtomoles per milligram ofETE.

Assay of ligand-gated ion channels.The density of GABAAreceptors was measured as the difference in binding of[3H]muscimol (90 nM) in the absence or presence of 1026 MSR-95531 after a 30-min incubation at 4°C in 50 mM Tris-citrate (pH 7.1). This method was adapted from that used withparticulate membrane (Dodd et al., 1992). The GABAA recep-tor antagonist SR-95531 (Gruen et al., 1995) was used insteadof GABA as, in preliminary experiments, it proved slightlymore effective than GABA at the same concentration in dis-placing [3H]muscimol from human BA 9. Before the estimationof [3H]muscimol binding all sections were first washed threetimes in ice-cold buffer (50 mM Tris-citrate, pH 7.1) at 4°C for5 min and then air-dried using a stream of cool air at roomtemperature.

The difference in the binding of [3H]TCP (20 nM) alone orin the presence of MK-801 (1026 M) was taken as a measure ofthe NMDA receptor (Vignon et al., 1983) after incubating in 5mM Tris (pH 8.2) containing glycine (1 mM), EGTA (10 mM),and EDTA (10 mM) after a 30-min incubation at 4°C.

Assay of seven-transmembrane receptors.The density of5-HT2A receptors, measured as described previously (Dean andHayes, 1996), was taken as the difference in the binding of[3H]ketanserin (10 nM) in the absence or presence of spiperone(1025 M).

To attempt to determine the makeup of DA D2-like receptorsin BA 9, three radioligands were used to measure the density ofDA D2-like receptors. For all the radioligands the incubationswere carried out at room temperature for 60 min in 50 mM Trisbuffer containing 120 mM NaCl, 5 mM KCl, 2 mM CaCl2, and1 mM MgCl2 at pH 7.4. The difference between the binding of[3H]YM-09151-2 (4 nM) (Murray et al., 1995) or [3H]raclo-pride (8 nM) (Dean et al., 1997) in the absence or presence of(1)-butaclamol (1025 M) was compared as a measure of levelsof DA D2-like receptors. The difference in binding between[3H]7-OH-DPAT (2 nM) alone and in the presence of 1026 Mhaloperidol was taken as a measure of levels of DA D3 recep-tors (Herroelen et al., 1994). For both [3H]raclopride and[3H]7-OH-DPAT binding, tissue sections were washed twicefor 5 min at room temperature in buffer and dried thoroughlybefore being used to measure radioligand binding. The differ-ence in binding between [3H]SCH 23390 (3 nM) alone and inthe presence of 1026 M cis-flupenthixol was taken as a measureof levels of DA D1 receptors (Knable et al., 1996).

Assay of DA transporter.The density of the DA transporterwas measured essentially as described previously using auto-radiography and [3H]mazindol (Javitch et al., 1983). Thus, thetotal binding of [3H]mazindol (15 nM) was measured in thepresence of 0.3mM desmethylimipramine in 50 mM Tris buffercontaining 300 mM NaCl and 5 mM KCl after an incubation of60 min at 4°C. NSB was measured by incubating tissue sec-tions under identical conditions but in the presence of 1026 Mmazindol.

Assay of NOS.The density of NOS was measured as de-scribed previously (Doyle and Slater, 1995). Thus, tissue wasfirst washed at room temperature for 90 min in assay buffer (50mM Tris-HCl containing 3 mM CaCl2 and 0.025% TritonX-100). The tissue sections were then thoroughly dried, and thedensity of NOS was measured as the difference in binding ofNG-nitro-L-[3H]arginine HCl (20 nM) in the absence or pres-ence ofNv-nitro-L-arginine (1026 M) after incubation with BA9 for 120 min at room temperature.

Statistical analysis.All statistical analysis was carried outusing the Minitab Statistical Software Release II. Statisticaldifferences were identified by comparing age, PMI, DOI, FT,and radioligand binding between the schizophrenic and controlgroups using the Mann–WhitneyU test. The relationshipsbetween age, DOI, PMI, pH, and final recorded drug dose withthe binding of each radioligand was assessed using the Pearsonproduct–moment correlation coefficients calculated using anassumed straight-line fit. An analysis of covariance was alsocarried out to determine if age, PMI, pH, or FT was a con-founding variable influencing the apparent relationship be-tween radioligand binding data from the schizophrenic andcontrol subjects.

RESULTS

There was a significant decrease in the binding of[3H]ketanserin to the 5-HT2A receptor in BA 9 fromschizophrenic compared with control subjects (mean6 SEM, 39 6 3.3 vs. 606 3.6 fmol/mg of ETE;p, 0.001; Table 1 and Fig. 1). By contrast, there was anincrease in the mean6 SEM density of [3H]muscimolbinding to BA 9 from the schizophrenic compared withcontrol subjects (5266 19 vs. 4446 28 fmol/mg ofETE; p 5 0.02). There were no significant differencesbetween the binding of [3H]TCP, [3H]YM-09151-2,[3H]SCH 23390, [3H]mazindol, orNG-nitro-L-[3H]argi-nine HCl to BA 9 from schizophrenic and control sub-jects (Fig. 1). There was no detectable specific binding ofeither [3H]raclopride or [3H]7-OH-DPAT in BA 9 fromeither the schizophrenic or control subjects.

FIG. 1. Density of NOS, DA D1-like receptors, [3H]YM-09151-2binding (YM), the DA transporter (DAT), the NMDA receptor, the5-HT2A receptor, and the GABAA receptor in BA 9 from schizo-phrenic and control subjects. Data are mean 6 SEM (bars)values.


1596 B. DEAN ET AL.

Relationships between radioligand binding andconfounding factors

There was a significant negative correlation betweenthe density of [3H]TCP binding to BA 9 from the non-schizophrenic subjects and FT (r 5 20.74), which wasnot as apparent in the tissue from the schizophrenicsubjects (r 5 20.2; Fig. 2). There were no other corre-lations between radioligand binding and donor age, PMI,pH, FT, DOI, or final drug dose. Analysis of covarianceshowed that there was no significant effect of donor age,PMI, pH, FT, DOI, or final recorded antipsychotic drugdose on the comparison of radioligand binding in theschizophrenic and control subjects.

DISCUSSION

Using tissue from new cohorts of subjects, this studyhas confirmed our earlier finding (Dean and Hayes,1996) that there is a decrease in the binding of [3H]ketan-serin to the 5-HT2A receptor in BA 9 from subjects withschizophrenia. In addition, we have shown a significantincrease in the binding of [3H]muscimol to the GABAAreceptor in the same brain region.

The finding of decreased 5-HT2A receptor density inthe DLPFC from subjects with schizophrenia is in linewith the majority of other studies on this receptor (Mitaet al., 1986; Arora and Meltzer, 1991; Hashimoto et al.,1993; Dean and Hayes, 1996). Indeed, this would now beone of the most consistent findings relating to changes in

neurotransmitter receptors in schizophrenia and wouldsuggest a disturbance of serotonergic function in theDLPFC in subjects with schizophrenia.

In an early study, [3H]LSD binding in BA 4, 10, and11 was reported not to be changed in tissue from medi-cated schizophrenic subjects but increased in tissue from“unmedicated” subjects (Whitaker et al., 1981). In a laterstudy 125I-LSD binding similarly showed no change inBA 9 from schizophrenic subjects receiving antipsy-chotic drugs until death (Joyce et al., 1993). That studyreported an increase in125I-LSD binding in the middlelaminae of the posterior cingulate from the schizophrenicsubjects. In a more recent study using125I-LSD, thedensity of binding was reported as decreased in BA 26and 6 tissue from schizophrenic subjects whether or notthey were receiving antipsychotic drugs at death (Gurev-ich and Joyce, 1997). By contrast, in the other 11 corticalareas studied (including BA 9), a decrease in125I-LSDbinding was only detectable in tissue from the subjectstreated with antipsychotic drugs up until death. Thus,although results using radioactive LSD are not consis-tent, they do provide some evidence that changes in5-HT2A receptors in tissue from subjects with schizo-phrenia may be due to either antipsychotic drug treat-ment during life or the presence of residual antipsychoticdrugs in the tissue studied. It is significant that a studyusing [3H]ketanserin has reported that binding is de-creased in BA 9 in both medicated and “unmedicated”subjects (Mita et al., 1986). Thus, the limited availabledata on [3H]ketanserin binding in DLPFC from schizo-phrenic subjects suggest that if binding changes are dueto the effect of antipsychotic drugs, they are presentseveral months after the cessation of antipsychotic drugtreatment.

This study has also shown an increase in the density ofGABAA receptors in BA 9 from subjects with schizo-phrenia, a finding that is again consistent with that ofothers (Hanada et al., 1986; Benes et al., 1996). It issignificant that other studies have either failed to show achange (Pandey et al., 1997) or have shown a reduction(Squires et al., 1993) in the density of benzodiazepinebinding to the GABAA receptors in the DLPFC fromschizophrenic subjects. It is important that the two stud-ies that reported an increase in GABAA receptors used[3H]muscimol, which binds to thea subunit of theGABAA receptor. By contrast, the radioactive benzodi-azepines would bind to theb subunit of the GABAAreceptor (McKernan and Whiting, 1996). Thus, currentdata from subjects with schizophrenia could indicate thatabnormalities in subunit assembly of the GABAA recep-tor in the DLPFC and other brain regions could beimportant in the pathology of the illness.

One of the objectives of this study was to determine ifconcomitant changes in neurotransmitter receptors ortransporters occur in the brains of subjects with schizo-phrenia. No such changes were observed. This study didshow selective changes in the 5-HT2A and GABAA re-ceptors in BA 9 from subjects with schizophrenia. It hasbeen suggested that the increase in GABAA receptor

FIG. 2. Relationship between the density of [3H]TCP binding tothe NMDA receptor in BA 9 from schizophrenic (top panel) andcontrol (bottom panel) subjects with the time that the tissue hadbeen at 270°C.



density in the DLPFC from subjects with schizophreniareflects receptor up-regulation as a result of decreasedGABAergic activity (Benes et al., 1996). It is significantthat the activity of GABAergic neurons has been shownto be facilitated by activation of the 5-HT2A receptor(Cozzi and Nichols, 1996). Thus, it seemed possiblefrom our data that a decrease in 5-HT2A receptor contentin the DLPFC could be associated with a decrease inGABAergic neuronal activity and hence an increase inGABAA receptors. However, there is no relationshipbetween the density of 5-HT2A receptors and GABAAreceptors in BA 9 from schizophrenic (r 2 5 0.02) orcontrol (r 2 5 0.01) subjects. This suggests that if there isa relationship between levels of 5-HT2A and GABAAreceptors in the DLPFC, it is too complex to be identifiedin this study.

This study failed to identify a change in the density ofNMDA receptors in BA 9 from subjects with schizophre-nia, a finding that is consistent with other reports on theNMDA receptors in the DLPFC from subjects withschizophrenia (Kornhuber et al., 1989; Ishimaru et al.,1992). It is significant that one of these reports (Ishimaruet al., 1992) and another study (Simpson et al., 1992)have shown increases in NMDA receptors in other cor-tical regions from schizophrenic subjects. Thus, it wouldappear that changes in NMDA receptors may be re-stricted to specific regions of the cortex from subjectswith schizophrenia.

No specific binding of [3H]raclopride or [3H]7-OH-DPAT was detected in the DLPFC from eitherschizophrenic or control subjects used in this study.This finding is consistent with previous reports of lowlevels of DA D2-like receptors in the human DLPFC(Hall et al., 1994; Lahti et al., 1995). By contrast, therewere detectable levels of [3H]YM-09151-2 binding inthe DLPFC. In studies of the caudate-putamen it hasbeen suggested that differences between [3H]raclo-pride and [3H]YM-09151-2 binding may be a measureof the DA D4 receptor (Seeman et al., 1993). Unfor-tunately, in the DLPFC such an interpretation is notpossible as [3H]YM-09151-2 also binds to thes re-ceptor (Helmeste et al., 1996). Thus, further experi-ments will be required to determine the nature of[3H]YM-09151-2 receptor binding in the humanDLPFC.

There was no difference in the binding of [3H]SCH23390 to DA D1-like receptors in the DLPFC fromschizophrenic compared with control subjects. Thus, thisstudy failed to replicate the recently reported finding ofan increase in DA D1-like receptors in DLPFC fromsubjects with schizophrenia who had received antipsy-chotic drugs (Knable et al., 1996). This study also failedto show a difference in the density of the DA transporterin the DLPFC from subjects with schizophrenia.

Finally, this study failed to demonstrate a change inNG-nitro-L-[3H]arginine binding to NOS in the DLPFCfrom subjects with schizophrenia, a finding consistentwith a study of the cerebellum using this radioligand(Doyle and Slater, 1995). Thus, at present, changes in

NOS have only been found in the temporal lobe (Akbar-ian et al., 1994) and hypothalamus (Bernstein et al.,1998) from subjects with schizophrenia.

There are several proposals suggesting how changes inthe neurochemistry of the DLPFC may be involved in thepathology of schizophrenia. This study does not presentevidence that would support the hypothesized role forchanges in the dopaminergic system in the DLPFC inschizophrenia (Knable and Weinberger, 1997). However,this study does support a role for changes in GABAergicsystems in the DLPFC in schizophrenia (Benes et al.,1996). It has been proposed that changes in the interac-tions between GABAergic, dopaminergic, and glutama-tergic systems are important in the pathology of schizo-phrenia (Goldman-Rakic and Selemon, 1997). Our studydoes not readily provide evidence to support this hypoth-esis. This study does provide preliminary evidence forthe hypothesis that an altered interaction betweenGABAergic and serotonergic neurons in the DLPFC maybe involved in the pathology of schizophrenia.

Acknowledgment: Dr. Kenneth Opeskin is a neuropatholo-gist at the Victorian Institute of Forensic Medicine, Southbank,and the Department of Forensic Medicine, Monash University,Clayton, Victoria, Australia. This work has been supported inpart by grants-in-aid from the State Government of Victoria,the Rebecca L. Cooper Medical Research Foundation, and theWoods Family Trust. The authors would like to thank Mr.Geoffrey Pavey for his excellent technical assistance and Pro-fessor Nicholas Keks for his assistance with postmortem diag-nosis.

REFERENCES

Akbarian S., Vinuela A., Kim J. J., Potkin S. G., Bunney W. E., andJones E. G. (1994) Distorted disturbance of nicotinamide adeninedinucleotide phosphate-diaphorase neurons in temporal lobe ofschizophrenics implies anomalies in cortical development.Arch.Gen. Psychiatry50, 178–187.

American Psychiatric Association (1987)Diagnostic and StatisticalManual of Mental Disorders,3rd edit., revised. American Psychi-atric Association, Washington, D.C.

Arora R. C. and Meltzer H. Y. (1991) Serotonin2 (5-HT2) receptorbinding in the frontal cortex of schizophrenic patients.J. NeuralTransm. Gen. Sect.85, 19–29.

Benes F. M., Vincent S. L., Marie A., and Khan Y. (1996) Up-regulation of GABAA receptor binding on neurons of the prefron-tal cortex in schizophrenic subjects.Neuroscience75,1021–1031.

Bennett J. P., Enna S. J., Bylund D. B., Gillin J. C., Wyatt R. J., andSnyder S. H. (1979) Neurotransmitter receptors in frontal cortexof schizophrenics.Arch. Gen. Psychiatry36, 927–934.

Bernstein H. G., Stanarius A., Baumann B., Henning H., Krell D.,Danos P., Falkai P., and Bogerts B. (1998) Nitric oxide synthase-containing neurons in the human hypothalamus: reduced numberof immunoreactive cells in the paraventricular nucleus of depres-sive patients and schizophrenics.Neuroscience83, 867–875.

Cozzi N. V. and Nichols D. E. (1996) 5-HT2A receptor antagonistsinhibit potassium-stimulated gamma-aminobutyric acid release inrat frontal cortex.Eur. J. Pharmacol.309,25–31.

Dean B. and Hayes W. (1996) Decreased frontal cortical serotonin2Areceptors in subjects with schizophrenia.Schizophr. Res.21,133–139.

Dean B., Pavey G., and Opeskin K. (1997) [3H]Raclopride binding tobrain tissue from subjects with schizophrenia: methodologicalaspects.Neuropharmacology36, 779–786.


1598 B. DEAN ET AL.

Dodd P. R., Thomas G. J., Harper C. G., and Kril J. J. (1992) Aminoacid neurotransmitter receptor changes in cerebral cortex in alco-holism: effect of cirrhosis of the liver.J. Neurochem.59, 1506–1515.

Doyle C. A. and Slater P. (1995) Application of [3H]NG-nitro-L-arginine labelling to measure cerebellar nitric oxide synthase inpatients with schizophrenia.Neurosci. Lett.202,49–52.

Egan M. F. and Weinberger D. R. (1997) Neurobiology of schizophre-nia. Curr. Opin. Neurobiol.7, 701–707.

Foster P. (1998) Neuroleptic equivalence.Pharmaceut. J.243, 431–432.

Fuster J. M. (1989)The Prefrontal Cortex: Anatomy, Physiology andNeuropsychology of the Frontal Lobe.Raven Press, New York.

Goldman-Rakic P. S. and Selemon L. D. (1997) Functional and ana-tomical aspects of prefrontal pathology in schizophrenia.Schizo-phr. Bull. 23, 437–458.

Gruen R. J., Wenberg K., Elahi R., and Friedhoff A. J. (1995) Alter-ations in GABAA receptor binding in the prefrontal cortex fol-lowing exposure to chronic stress.Brain Res.684,112–114.

Gurevich E. V. and Joyce J. N. (1997) Alterations in the corticalserotonergic system in schizophrenia: a postmortem study.Biol.Psychiatry42, 529–545.

Hall H., Sedvall G., Magnusson O., Kopp J., Halldin C., and Farde L.(1994) Distribution of D1- and D2-dopamine receptors, and do-pamine and its metabolites in the human brain.Neuropsychophar-macology11, 245–256.

Hanada S., Mita T., Nishino N., and Tanaka C. (1986) [3H]Muscimolbinding in autopsied brains of chronic schizophrenics.Life Sci.40,259–266.

Hashimoto T., Kitamura N., Kajimoto Y., Shirai Y., Shirakawa O.,Mita T., Nishino N., and Tanaka C. (1993) Differential changes inserotonin 5-HT1A and 5-HT2 receptor binding in patients withchronic schizophrenia.Psychopharmacology (Berl.)112 (Suppl.),S35–S39.

Helmeste D. M., Tang S. W., Bunney W. E. Jr., Potkin S. G., and JonesE. G. (1996) Decrease ins but no increase in striatal dopamine D4sites in schizophrenic brains.Eur. J. Pharmacol.314,R3–R5.

Herroelen L., De Backer J.-P., Wilczak N., Flamez A., Vauquelin G.,and De Keyser J. (1994) Autoradiographic distribution of D3-typedopamine receptors in human brain using [3H]7-hydroxy-N,N-di-n-propyl-2-aminotetralin.Brain Res.648,222–228.

Hill C., Keks N. A., Roberts S., Opeskin K., Dean B., and CopolovD. L. (1996a) Postmortem brain studies in schizophrenia: theproblems of diagnosis.Am. J. Psychiatry153,533–537.

Hill C., Roberts S., Keks N. A., Dean B., Mackinnon A., and CopolovD. L. (1996b) Diagnostic Instrument for Brain Studies.MentalHealth Research Institute, Melbourne.

Ishimaru M., Kurumaji A., and Toru M. (1992) NMDA-associatedglycine binding site increases in schizophrenic brain.Biol. Psy-chiatry 32, 379–380.

Javitch J. A., Blaustein R. O., and Snyder S. H. (1983) [3H]Mazindolbinding associated with neuronal dopamine uptake sites in corpusstriatum membranes.Eur. J. Pharmacol.90, 461–462.

Joyce J. N., Shane A., Lexow N., Winokur A., Casanova M. F., andKleinman J. E. (1993) Serotonin uptake sites and serotonin recep-tors are altered in the limbic system of schizophrenics.Neuropsy-chopharmacology8, 315–336.

Kingsbury A. E., Foster O. J., Nisbet A. P., Cairns N., Bray L., EveD. J., Lees A. J., and Marsden C. D. (1995) Tissue pH as anindicator of mRNA preservation in human post-mortem brain.Mol. Brain Res.28, 318.

Knable M. B. and Weinberger D. R. (1997) Dopamine, the prefrontalcortex and schizophrenia.J. Psychopharmacol.11, 123–131.

Knable M. B., Hyde T. M., Murray A. M., Herman M. M., andKleinman J. E. (1996) A postmortem study of frontal corticaldopamine D1 receptors in schizophrenics, psychiatric controls,and normal controls.Biol. Psychiatry40, 1191–1199.

Kornhuber J., Mack-Burkhardt F., Riederer P., Hebenstreit G. F.,Reynolds G. P., Andrews H. B., and Beckmann H. (1989)[3H]MK-801 binding sites in postmortem brain regions of schizo-phrenic patients.J. Neural Transm. Gen. Sect.77, 231–236.

Lahti R. A., Roberts R. C., and Tamminga C. A. (1995) D2-familyreceptor distribution in human postmortem tissue: an autoradio-graphic study.Neuroreport6, 2505–2512.

Leysen J. E., Niemegeers C. J. E., van Nueten J. M., and Laduron P. M.(1982) [3H]Ketanserin (R41468), a selective3H-ligand for sero-tonin S2 receptor binding sites.Mol. Pharmacol.21, 301–314.

Lin A. M.-Y., Kao L.-S., and Chai C. Y. (1995) Involvement of nitricoxide in dopaminergic transmission in rat striatum: an in vivoelectrochemical study.J. Neurochem.65, 2043–2049.

Lorrain D. S. and Hull E. M. (1993) Nitric oxide increases dopamineand serotonin release in medial preoptic area.Neuroreport 5,87–89.

McKernan R. M. and Whiting P. J. (1996) Which GABAA-receptorsubtypes really occur in the brain.Trends Neurosci.19,139–143.

Meltzer H. Y. (1987) Biological studies in schizophrenia.Schizophr.Bull. 13, 77–107.

Mita T., Hanada S., Nishino N., Kuno T., Nakai H., Yamadori Y., andTanaka C. (1986) Decreased serotonin S2 and increased dopamineD2 receptors in chronic schizophrenics.Biol. Psychiatry 21,1407–1414.

Murray A. M., Hyde T. M., Knable M. B., Herman M. M., BigelowL. B., Carter J. M., Weinberger D. R., and Kleinman J. E. (1995)Distribution of putative D4 dopamine receptors in postmortemstriatum from patients with schizophrenia.J. Neurosci.15,2186–2191.

Ohkuma S., Narihara H., Katsura M., Hasegawa T., and Kuriyama K.(1995) Nitric oxide–induced [3H]GABA release from cerebralcortical neurons is mediated by peroxynitrite.J. Neurochem.65,1109–1114.

Pandey G. N., Conley R. R., Pandey S. C., Goel S., Roberts R. C.,Tamminga C. A., Chute D., and Smialek J. (1997) Benzodiazepinereceptors in the post-mortem brain of suicide victims and schizo-phrenic subjects.Psychiatry Res.71, 137–149.

Seeman P., Guan H.-C., and Van Tol H. H. M. (1993) Dopamine D4receptors elevated in schizophrenia.Nature365,441–445.

Simpson M. D. C., Slater P., Royston M. C., and Deakin J. F. W.(1992) Alterations in phencyclidine and sigma binding sites inschizophrenic brain.Schizophr. Res.6, 41–48.

Squires R. F., Lajtha A., Saederup E., and Palkovits M. (1993) Reduced[3H]flunitrazepam binding in cingulate cortex and hippocampus ofpostmortem schizophrenic brains: is selective loss of glutamater-gic neurons associated with major psychoses?Neurochem. Res.18, 219–223.

Vignon J., Chicheportiche R., Chicheportiche M., Kamenka J.-M.,Geneste P., and Lazdunski M. (1983) [3H]TCP: a new tool withhigh affinity for the PCP receptor in rat brain.Brain Res.280,194–197.

Whitaker P. M., Crow T. J., and Ferrier I. N. (1981) Tritiated LSDbinding in frontal cortex in schizophrenia.Arch. Gen. Psychiatry38, 278–280.



changes in serotonin2a and gabaa receptors in schizophrenia : studies on the human dorsolateral...

Documents