Decreased muscarinic1 receptors in the dorsolateral ... - Nature

Molecular Psychiatry (2002) 7, 1083–1091  2002 Nature Publishing Group All rights reserved 1359-4184/02 $25.00 www.nature.com/mp

ORIGINAL RESEARCH ARTICLE

Decreased muscarinic1 receptors in the dorsolateral prefrontal cortex of subjects with schizophrenia B Dean, M McLeod, D Keriakous, J McKenzie and E Scarr The Rebecca L Cooper Research Laboratories, The Mental Health Research Institute of Victoria, Parkville, Victoria, Australia To test the hypothesis that muscarinic receptors are involved in the pathology of schizophrenia, we measured muscarinic1 (M1R) and muscarinic4 (M4R) protein and mRNA as well as [3H]pirenzepine binding in Brodmann’s areas (BA) 9 and 40 obtained postmortem from 20 schizophrenic and 20 age/sex-matched control subjects. There was a significant decrease in [3H]pirenzepine binding to BA 9 (mean ± SEM: 151 ± 15 vs 195 ± 10 fmol mg−1 ETE; P ⬍ 0.02), but not BA 40 (143 ± 13 vs 166 ± 11 fmol mg−1 ETE), from subjects with schizophrenia. The level of M1R protein (0.11 ± 0.007 vs 0.15 ± 0.008 OD; P ⬍ 0.01), but not M4R protein, was decreased in BA9 from schizophrenic subjects with neither receptor protein being altered in BA 40. The level of M1R mRNA was decreased in BA 9 (30 ± 7.0 vs 79 ± 14 dpm × 103 mg−1 ETE, P ⬍ 0.01) and BA 40 (28 ± 5.9 vs 99 ± 14, P ⬍ 0.01) with schizophrenia but M4R mRNA was only decreased in BA 40 (48 ± 6.6 vs 89 ± 9.9, P ⬍ 0.005). These data suggest that the M1R, at least in the dorsolateral prefrontal cortex, may have a role in the pathology of schizophrenia. Molecular Psychiatry (2002) 7, 1083–1091. doi:10.1038/sj.mp.4001199 Keywords: postmortem; cortex; human; acetylcholine; cholinergic

Introduction It has been suggested that the cholinergic systems of the central nervous system (CNS) are involved in the pathology of schizophrenia.1 This hypothesis is supported by a growing understanding that the deficits of CNS function associated with schizophrenia, such as cognitive function, memory, motor activity and sleep,2 are significantly influenced by cholinergic systems. Interest in cholinergic systems in schizophrenia was reawakened with the demonstration that drugs such as clozapine and olanzapine, both of which bind to muscarinic receptors,3 are of therapeutic benefit in treating symptoms of the disorder.4,5 Thus, from therapeutic pharmacology, postmortem tissue research and other lines of evidence it has now been suggested that muscarinic receptors are important in the pathology of schizophrenia.2 Five muscarinic receptors have been identified in the human CNS2 but investigations on the involvement of changes in muscarinic receptors in schizophrenia have been hampered by the lack of receptor-specific radioligands and drugs.6 This laboratory has reported a decrease in [3H]pirenzepine binding in the frontal cor-

Correspondence: Associate Professor B Dean, The Rebecca L Cooper Research Laboratories, The Mental Health Research Institute of Victoria, Parkville, Victoria 3042, Australia. E-mail: [email protected] Received 3 December 2001; revised 27 February 2002; accepted 1 March 2002

tex,7,8 hippocampus9 and caudate-putamen10 from subjects with schizophrenia. [3H]pirenzepine binds selectively to muscarinic1 (M1R) and muscarinic4 (M4R) receptors11 and we therefore suggested that deficits in M1R and/or M4R are present in various CNS regions from subjects with schizophrenia. Given the different roles for these receptors in the human CNS2 it was important to determine the changes in M1R and M4R that were contributing to changes in [3H]pirenzepine binding in the CNS of subjects with schizophrenia. To begin to address the issue of which muscarinic receptors are altered in the CNS of subjects with schizophrenia, we focused on human cortex obtained postmortem. The cholinergic system innervates the entire cortex, but the input to the frontal cortex is more influenced by direct projections from the basal forebrain whilst cholinergic inputs into the parietal and occipital cortex are more modulated by the thalamus.12 To determine if these differences in input could affect the status of M1R and M4R in cortical regions in schizophrenia, we measured [3H]pirenzepine binding in Brodmann’s area (BA) 9 (dorsolateral prefrontal cortex) and 40 (supramarginal area of the parietal lobe) obtained postmortem. In addition, to determine which of the two muscarinic receptors that bind [3H]pirenzepine are altered in schizophrenia, we measured the levels of M1R and M4R by Western blots with receptor specific antibodies and levels of mRNA using in situ hybridisation.

Muscarinic receptors in schizophrenia B Dean et al

1084

Material and Methods Materials [N-methyl-3H]pirenzepine, [35S]-deoxyadenosine 5⬘-(␣thio)triphosphate ([35S]ATP) and Oligonucleotide 3⬘ end labeling system were obtained from NEN Life Sciences Product Inc, Boston, MA, USA. Rabbit antiM1R and anti-M4R antibodies were obtained from Research and Diagnostic Antibodies, Berkley, CA, USA; goat anti-rabbit IgG: HRP conjugated were obtained from DAKO (Australia) Pty Ltd, Botany, NSW, Australia (M4R) and Vector Laboratories, Inc, Burlingame, CA, USA (M1R). Quinuclidinyl xanthene-9carboxylate hemioxalate (QNX) was obtained from Research Biochemicals International, Natick, MA, USA. Complete protease inhibitor cocktail (CPIC) was obtained from Boehringer Mannheim Pty Ltd, Castle Hill, NSW, Australia. Oligonucleotide probes directed against mRNA for the M1R and M4R were synthesised by GIBCO BRL, Grand Island, NY, USA. Electrophoretic grade acrylamide, glycine, Tween 20, 2-mercaptoethanol, ammonium persulphate, N,N,N⬘,N⬘-tetramethyl-ethylenediamine (TEMED), sodium dodecyl sulfate (SDS), TRIS and N,N⬘-methylene-bis-acrylamide (BIS) as well as DC Protein Assay Kits, ImmunoBlot PVDF membrane, Trans-Blot Transfer Medium, Bio-spin columns and broad range SDS-Page molecular weight standards were obtained from Bio-Rad, Hercules, CA, USA. All other chemicals were obtained from Sigma, Castle Hill, NSW, Australia. ECL Western Blot Detection System, Hyperfilm-␤max, Hyperfilm ECL, [3H]micro-scales and [14C]micro-scales were obtained from Amersham Pharmacia Biotech Australia Pty Ltd, Castle Hill, NSW, Australia. BAS-TR2025 Fuji Imaging Plates were obtained from Berthold Australia Pty Ltd, Bundoora, Victoria, Australia. Tissue collection Consent for this study was obtained from the Ethics Committee of the Victorian Institute of Forensic Medicine and the North Western Mental Health Program Behavioural and Psychiatric Research and Ethics Committee. After gaining consent, blocks of tissue, containing BA 9 and 40 were excised from the left hemisphere of subjects who were listed as having schizophrenia in reports of death. Tissue was also collected from the same regions from subjects with no history of psychiatric illness (controls) who were matched for sex and closely matched for age to the subjects with schizophrenia (Tables 1 and 2). In cases where death was witnessed, the time between death and autopsy was taken as the postmortem interval (PMI). Where death was not witnessed, tissue was only taken from individuals who had been seen alive up to 5 h before being found dead. In these cases PMI was taken as the interval half way between the donor being found dead and last being seen alive. In all cases, cadavers were refrigerated within 5 h of being found and tissue was rapidly frozen to −70°C within 30 min of autopsy. The pH of the CNS tissue was measured as described previously.13

Molecular Psychiatry

Diagnostic evaluation For a subject to be included in this study, sufficient information needed to be available from clinical case records to enable a psychologist and psychiatrist to reach a diagnostic consensus using the Diagnostic Instrument for Brain Studies (DIBS). The DIBS is a semi-structured protocol for postmortem assessment14 allowing the diagnosis of schizophrenia according to DSM-IV criteria.15 Subjects with schizoaffective disorder were excluded. When available, postmortem toxicology was reviewed to exclude recent substance misuse and levels of antipsychotic and anticholinergic drugs in blood were recorded. Duration of illness (DOI) was calculated as the time from first admission to death and a comprehensive medication history was obtained for each subject. Most recently prescribed antipsychotic and anticholinergic drugs and their final record prescribed doses were recorded and then converted to standardised drug doses16 (Tables 1 and 2). [3H]Pirenzepine binding In situ radioligand binding and autoradiography was performed using five 20-␮m frozen tissue sections from each block of BA 9 and BA 40. The binding of [3H]pirenzepine (15 nM) was measured in the presence (Nonspecific binding (NSB): two sections) or absence (Total binding: three sections) of 10−6 M QNX10 after incubating in 10 mM sodium-potassium phosphate buffer (10 mM KH2PO4, 10 mM Na2HPO4; pH 7.4) at 25°C for 30 min. Sections were then washed twice for 2 min in icecold 10 mM sodium-potassium phosphate buffer, dipped in ice-cold water and thoroughly dried prior to being fixed overnight in paraformaldehyde fumes in a desiccator. The sections, and a set of [3H]micro-scales, were apposed against a BAS-TR2025 imaging plate until an image of appropriate intensity was obtained for scanning in the BAS 5000 phosphoimager. Exposure time related to both the density of binding sites and the specific activity of the radioligand utilised. The intensity of the phosphoimages was then measured by comparison to the intensity of the blocks of radioactivity on the [3H]microscales using AIS image analysis software, with results being expressed as dpm mg−1 estimated wet weight tissue equivalents (ETE) and then converted to fmol mg−1 ETE. In this way, [3H]pirenzepine binding was measured using a single point saturation analysis which gives a good approximation of the density of radioligand binding sites in tissue sections. Western blot analyses Tissue for the analysis of M1R and M4R by Western blot was prepared from each block of BA 9 and BA 40 used to prepare frozen sections for in situ radioligand binding and in situ hybridisation. Each sample of tissue was homogenized with five strokes of a Teflonglass homogenizer into 20 mM TRIS-HCl pH 7.4 containing 0.2 mM EGTA and 0.1 mM EDTA. Importantly, 3 × CPIC, phenylmethylsulfonylfluoride (1 mM), leupeptin (10 ␮g ml−), benzamidine (1 mM), bacitracin (1 mg ml−1), pepstatin (10 ␮g ml−1) and chymostatin (10


Table 1 Demographic, clinical and pharmacological data on schizophrenic subjects from whom tissue was used for the study of muscarinic receptors Schizophrenia ID Sex Age (yr)

pH

PMI (h)

DOI (yr)

Cause of death

Antipsychotic

FAD

BAD

Risperidone Remoxipride Haloperidol Chlorpromazine Thioridazine Fluphenazine Trifluperazine Chlorpromazine Haloperidol Free for 5 yrs

6 750 128 100 1250 473 300

ND 75.0 ND ND 7.9 ND ND

160

ND ND

500

ND

160 600 200 160 700 460

1 2 3 4 5 6 7

F F M M M F M

35 26 42 19 48 38 53

6.36 6.11 6.44 6.22 6.62 6.43 6.29

39.0 41.0 47.0 43.0 30.0 20.0 9.0

9.0 11.0 8.0 3.0 24.0 17.0 9.0

Incised wrist injury Combined drug toxicity Hanging Unascertained Bronchopneumonia Burning Carotid artery atheroma

8 9

M M

38 69

5.52 40.0 6.44 48.0

15.0 6.0

10

M

41

6.20 31.0

11.0

11 12 13 14 15

F F M M F

35 30 22 46 48

6.26 6.37 6.17 5.53 6.21

15.0 48.0 37.0 42.0 52.5

7.0 10.0 4.0 12.0 22.0

16

M

65

6.29 42.0

36.0

17 18 19 20

M M M M

38 65 41 42

6.02 6.57 6.64 6.26

4.0 35.0 15.0 15.0

Mediastinitis Carbon monoxide poisoning Combined drug toxicity Fluphenazine Trifluperazine Carotid artery thrombosisHaloperidol Hanging Flupenthixol Combined drug toxicity Pimozide Ischaemic heart disease Haloperidol Pulmonary Flupenthixol thromboembolism Chlorpromazine Bronchopneumonia Trifluperazine Haloperidol Meningo encephalitis Free for 2 mths Ischaemic heart disease Fluphenazine Hanging Free for 8 yrs Drowning Flupenthixol Chlorpromazine

50.0 41.0 48.0 34.5

150 610

Antichol

FCD BCD

Benztropine Benztropine

2.0 1.0

ND ND

Pricyclidine

6.0

ND

ND ND 3.0 NA 0.8

Benztropine Benztropine

2.0 2.0

ND ND

ND

Benzhexol

1.6

ND

1085

ND ND ND ND

FAD = final recorded antipsychotic drug dose (mg chlorpromazine equivalents); BAD = blood antipsychotic drug levels (mg chlorpromazine equivalents l−1); Antichol = Anticholinergics; FCD = final recorded anticholinergic drug dose (mg benztropine equivalents); BCD = blood anticholinergic drug levels (mg benztropine equivalents l−1); ND = not detected.

␮g ml−1) were added to the homogenisation buffer on the day of use. Homogenised tissue was stored on ice until the concentration of protein in each sample was measured using the BioRad protein assay and then frozen at −70°C until required. Duplicate measurements of M1R and M4R were made for each sample and the samples from schizophrenics and their age- and sex-matched controls were run on the same gel. Thus, on the day of assay each tissue homogenate was thawed and the protein concentration adjusted to 1.0 mg ml−1. A sample of each homogenate was than denatured by heating at 95°C for 4 min in an equal volume of electrophoresis buffer (0.5 M TRIS-HCl pH 6.8, containing 20% glycerol, 10% SDS, 10% 2-␤-mercaptoethanol and 0.05% bromophenol blue). The denatured tissue was then loaded onto 4% SDS stacking gel (20 ␮g protein per well) and separated on 10% (M1R) or 7.5% (M4R) SDS mini-gels (200 V constant). Proteins were then transferred from the mini-gel to either Trans-blot Transfer Medium (nitrocellulose membrane: M1R) or Immuno-Blot PVDF (PVDF) membrane (M4R) in Towbin transfer buffer (25 mM Tris pH 8.3 containing 192 mM glycine and 20% methanol) overnight in a Bio-Rad Mini Transblot electrophoretic transfer cell (40 mA constant).

The proteins on each nitrocellulose and PVDF membrane were then stained with 0.2 % Ponceau S Red in 3% trichloroacetic acid to verify transfer had occurred. For the measurement of M1R, the nitrocellulose membranes containing the separated proteins were placed in blocking buffer A (Tween Tris buffered saline (TTBS: 100 mM Tris pH 7.5, 0.9% sodium chloride and 0.1% Tween 20) containing 5% blotting grade non-fat milk) for 45 min at room temperature (RT) with agitation. The proteins on the membrane were then exposed to anti-M1R antibody (diluted 1:1600) in blocking buffer A for 45 min at RT with agitation and then washed 6 × 5 min in TTBS. The washed membranes were exposed to 1: 2000 dilution Vector goat anti-rabbit IgG conjugated to horseradish peroxidase in blocking buffer A for 45 min at RT and then washed 6 ×5 min at RT with TTBS. Each membrane was then incubated with Amersham ECL at RT for 60 s followed by exposure to Amersham Hyperfilm—ECL so that the density of the internal control was within the previously defined optical density (OD) range (see below). For the measurement of M4R, the PVDF membranes containing the separated proteins were placed in blocking buffer B (blocking buffer A containing 2% heatMolecular Psychiatry


1086

Table 2 Demographic, clinical and pharmacological data on control subjects from whom tissue was used for the study of muscarinic receptors Controls ID Sex

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

F F M M M F M M M M F F M M F M M M M M

Age (yr)

pH

PMI Cause of death (h)

36 21 43 27 48 56 52 38 67 42 62 47 26 57 39 68 60 68 43 30

6.40 6.03 6.25 6.46 6.37 5.88 5.98 6.42 6.14 6.32 6.45 5.89 6.37 6.43 6.23 6.59 6.11 6.06 6.43 5.86

60.0 58.0 45.0 45.0 24.0 24.0 22.0 46.0 32.0 26.0 40.0 24.0 46.4 27.0 65.0 27.0 27.5 41.0 51.0 27.0

Dilated cardiomyopathy Myocarditis Drowning Drowning Coronary artery atheroma Pericardial tamponade Pulmonary thromboembolism Trauma/asphyxia Acute myocardial infarction Coronary artery atheroma Ischaemic heart disease Pulmonary embolism Electrocution Ischaemic heart disease Mitral valve prolapse Coronary artery atheroma Pulmonary embolism Aortic stenosis Coronary artery atheroma Coronary artery atheroma

inactivated goat serum) overnight at 4°C with agitation. The proteins on the membrane were then exposed to anti-M4R (diluted 1:500) in blocking buffer B for 4 h at RT with agitation and then washed 3 × 5 min in blocking buffer A. Each membrane was then exposed to 1:6000 dilution DAKO goat anti-rabbit IgG antibody conjugated to horseradish peroxidase in blocking buffer B for 2 h at RT and then washed 4 × 5 min at RT with blocking buffer A and twice with TTBS. PVDF membranes were then processed as described for nitrocellulose membranes to obtain an image on Amersham Hyperfilm—ECL. All ECL film was developed in Kodak D-19 X-ray developer and fixed with Ilford Hypam X-ray fixer. When the ECL film was washed and dried, it was imaged with a digital CCD camera and the OD of the bands measured using an MCID image analysis system. To control for blot-to-blot variation, a single membrane preparation (internal control) was initially prepared from BA 9 from a single donor. For both M1R and M4R measurements, this sample was run in all 15 wells on two gels (30 measurements) to establish the intra- and inter-gel variation in the measurement of M1R and M4R protein for the internal control. The internal control was subsequently run on every gel from which measurements of M1R or M4R protein were taken, with each gel being exposed to the ECL film for a time that allowed the OD of the internal control to fall within the range established, by calculating the mean ± two standard deviations of the initial 30 repeated measurements. Molecular Psychiatry

In situ hybridisation All solutions for in situ hybridisation used 0.1% diethyl pyrocarbonate-treated water. For in situ hybridisation, 6 × 10 ␮m frozen tissue sections were cut from BA 9 and BA 40 from each subject at the same time as those cut for in situ radioligand binding. These sections were stored frozen at −70°C until required. On the day of assay, the frozen sections were removed from cold storage and dried for at least 1 h at RT. The sections were then placed in 4% paraformaldehyde in phosphate buffered saline (PBS: 136 mM NaCl, 2 mM KH2PO4, 1 mM Na2HPO4 and 2.5 mM KCl) for 5 min at RT. The fixed sections were rinsed twice in PBS then placed in fresh 0.25% acetic anhydride in 0.1 M triethanolamine HCl, pH 8.0 for 10 min at RT, dehydrated by transfer through ethanol, delipidated in chloroform and equilibrated into 95% ethanol before being dried at RT. As in an earlier study,17 to increase hybridisation specificity each oligonucleotide probe contained a mix of three antisense sequences that were designed to hybridise to three regions of the receptor specific mRNA. Thus, for the M1R the oligonucleotide probes were complimentary to bases 4–51, 721–768, 811–853 of the human m1 mRNA (accession No: X52068) whilst the M4R were complimentary to bases 4–51, 721–768, 789–839 of human m4 mRNA (accession No: NM000741). The probes were labelled with [35S]ATP using the NEN Oligonucleotide 3⬘ end labeling system with each labelled probe being purified using a Micro Bio-Spin Bio Gel P30 column. Prior to hybridisation with the tissue sections, 2 ␮l of radiolabeled probe was placed into 4 ml scintillation fluid and the radioactivity of the probe measured. A sufficient quantity of probe was then prepared by dispersing the radiolabeled probe at 1 × 106 dpm of radioactivity per 50 ␮l of hybridisation buffer (formamide (50%), NaCl (600 mM), Tris-HCl (80 mM), EDTA (4 mM), sodium pyrophosphate (0.1%), SDS (0.2%), sodium heparin (0.2 mg ml−), dextran sulphate (10%) and DTT (100 mM) ), to give enough solution to cover all tissue sections. The radiolabeled probes were then incubated with tissue sections under optimized conditions to give maximum hybridisation (20 h at 37°C for M1R and M4R). To prevent dehydration each section was covered with parafilm and incubations were completed in a humidified chamber containing filter paper soaked in formamide (50%) in 4 × saline sodium citrate buffer (4 × SSC is diluted from a stock of 20 × SSC: 3 M NaCl plus 0.3 M Na citrate at pH 7.2). Two sections were incubated with the relevant radioactive oligonucleotide probe (total binding), whilst two sections were incubated with the radioactive oligonucleotide probe in the presence of an excess of non-radioactive probe (NSB). In addition, the specificity of oligonucleotide probe hybridisation to mRNA was confirmed by incubating two sections which had been exposed to RNase prior to hybridisation with the radioactive probes (negative control). On completion of the hybridisation procedure, all tissue sections were washed as described previously.18


The washed and dried sections were then apposed to Amersham Hyperfilm-␤max along with 14  [ C]microscales for a period sufficient to generate an image of appropriate density (苲6 weeks). Each image was then analysed as described for sections undergoing in situ radioligand binding. The specific binding of the oligonucleotide probes were calculated as total binding minus NSB expressed as dpm × 103 mg−1 ETE. Statistical analysis All sets of data were first analysed using the Kolmogorov–Smirnov test to determine whether they had a Gaussian distribution. Age, PMI and CNS pH for the two diagnostic cohorts were compared using Student’s t-test. Two-way ANOVA was used to identify variations between diagnoses or CNS region in all experimental measures. Post-hoc analyses using Student’s ttests were then used to identify the source of variation between diagnostic groups, CNS regions or between drug treatments within diagnostic groups. Pearson product-moment correlations, assuming a straight-line best fit, were used to determine if the experimental data were influenced by donor age, PMI, CNS pH, DOI or final recorded antipsychotic drug dose.

Results Demographic data BA 9 and BA 40 were obtained from 14 male and six female schizophrenic and 14 male and six female control subjects (Table 2). There was no statistical difference in the mean age of the schizophrenic (mean ± SEM: 42 ± 3.0 yrs) and control (47 ± 3.2 yrs) subjects (P = 0.3). Similarly, there were no significant differences in CNS pH (6.24 ± 0.07 vs 6.23 ± 0.05; P = 0.87) or PMI (37.9 ± 2.6 vs 37.9 ± 3.0 h; P = 0.99). Due to restrictions on the availability of tissue, analysis in BA 40 could not be completed on tissue from one schizophrenic (ID 5) and one control (ID 29) subject. [13H]Pirenzepine binding There was no layering in [3H]pirenzepine binding in either cortical region (Figure 1a). Two-way ANOVA

showed that there was a significant variation in [3H]pirenzepine binding between diagnoses (F = 7.45, df = 1,75, P = 0.01) which was not apparent between CNS regions (F = 2.15, df = 1,75, P = 0.15). Analysis with Student’s t-test within CNS regions showed that this was due to a significant decrease in the density of [3H]pirenzepine binding in BA 9 (151 ± 15 vs 195 ± 10 fmol mg−1 ETE; P ⬍ 0.02), but not BA 40 (143 ± 13 vs 166 ± 11 fmol mg−1 ETE; P = 0.18), from the subjects with schizophrenia (Figure 1b).

1087

Western blot analyses The Western blots for the M1 and M4 receptor in human CNS were essentially as obtained with other human tissue using the same quality controlled commercially available antibodies (Figure 2a, c).19,20 Thus, the protein bands representing the M1 and M4 receptors were identified by careful measurement of Rf values prior to quantification of the optical density values. Two-way ANOVA showed there was significant variation in the levels of M1R protein with diagnoses (F = 3.88, df = 1.75, P = 0.05) and CNS region (F = 22.92, df = 1.75, P ⬍ 0.01) (Figure 2b). Further analysis within diagnosis showed this was due to a decrease in M1R in BA9 (0.11 ± 0.007 vs 0.15 ± 0.008 OD; P ⬍ 0.01) but not BA 40 (0.10 ± 0.002 vs 0.09 ± 0.004 OD; P = 0.084) from subjects with schizophrenia. Analysis within diagnosis but between regions showed that the variance between regions was due to differences in levels of M1R in BA 9 and BA 40 (0.15 ± 0.008 vs 0.09 ± 0.004) from the control subjects (P ⬍ 0.01). Two-way ANOVA showed no significant differences between M4R proteins with diagnosis (F = 0.9, df = 1.75, P = 0.35) or CNS region (F = 0.61, df = 1.75, P = 0.44) (Figure 2d). In situ hybridisation After incubating for 苲6 weeks images were obtained showing the distribution of mRNA for the M1R (Figure 3a) and M4R (Figure 3c) in BA9 and BA 40 to be diffuse across laminae II–VI of the cortex. In the majority of cases, the image in the presence of either excess unlab-

Figure 1 A typical autoradiograph showing the binding of [3H]pirenzepine binding to human cortex (a) obtained postmortem which was measured (mean ± SEM) in BA 9 and BA 40 from 20 schizophrenic and 20 control subjects (b). **P ⬍ 0.02. Molecular Psychiatry


1088

Figure 2 Typical ECL image following Western blot analysis of M1 (a) and M4 (c) receptors in cortex obtained postmortem. M1 (b) and M4 (d) receptors were measured (mean ± SEM) in BA 9 and BA 40 from 20 schizophrenic and 20 control subjects. *** P ⬍ 0.01.

elled probe or from tissue exposed to RNase was not of sufficient intensity to analyse using the [14C]microscales. In those instances, total binding was taken as specific binding. In the remaining cases, binding of the radiolabelled probe to sections exposed to RNase or to sections along with cold labelled (NSB) was less than 10% of that of total binding. In these cases, the specific binding of the radiolabelled probe was calculated as total binding minus NSB. These data demonstrate that, in all cases, the radiolabelled probes were binding to mRNA with a sequence complimentary to that probe. Two-way ANOVA showed a significant difference in the levels of mRNA for the M1R with diagnoses (F = 29.2, df = 1.75, P ⬍ 0.001) but not between CNS regions (F = 0.73, df = 1.75, P = 0.39) (Figure 3b). The difference in M1R mRNA with diagnosis was due to a significant Molecular Psychiatry

decrease in mRNA in BA 9 (30 ± 7.0 vs 79 ± 14 dpm × 10−3 mg−1 ETE, P ⬍ 0.01) and BA 40 (28 ± 5.9 vs 99 ± 14, P ⬍ 0.01) in the tissue from the subjects with schizophrenia. Two-way ANOVA also showed a significant variation in levels of mRNA for the M4R with diagnosis (F = 8.95, df = 1.75, P ⬍ 0.005) but not between CNS regions (F = 0.95, df = 1.75, P = 0.33) (Figure 3d). The variance with diagnoses was due to a significant decrease in levels of mRNA encoding M4R in BA 40 from subjects with schizophrenia (48 ± 6.6 vs 89 ± 9.9, P ⬍ 0.005). Confounding factors For both cohorts of subjects, there were no significant correlations between experimental variables in either CNS region with donor age (r2 0.006–0.20), CNS pH (r2 ⬍0.001–0.30) or PMI (r2 ⬍0.001–0.34). For subjects


1089

Figure 3 Typical autoradiographs showing the distribution of mRNA for the M1 (a) and M4 (c) in cortex obtained postmortem. Levels of mRNA for the M1 (b) and M4 (d) receptor were measured (mean ± SEM) in BA 9 and BA 40 from 20 schizophrenic and 20 control subjects. ***P ⬍ 0.01.

with schizophrenia, there were no significant correlations between experimental variables and DOI illness (r2 0.003–0.24). All the schizophrenic subjects from whom tissue had been obtained for this study had been treated with antipsychotic drugs, however, blood levels of drugs at autopsy were not detected in 16 of the 20 subjects (Table 1). There were no significant correlations between experimental variables and final recorded antipsychotic drug dose (r2 0.01–0.18) in either CNS region. Measurable levels of antipsychotic drug were detected in the blood of only three individuals and

hence a meaningful regression of this factor with the experimental variables could not be completed. Another confounding factor was that six of the 20 schizophrenic subjects had a record in their case histories of receiving anticholinergic drugs close to death (Table 1). However, the levels of anticholinergic drugs in blood from these individuals that was obtained at autopsy were too low to be measured. In neither CNS region was there a correlation between the final recorded dose of anticholinergic drug and any of the experimental variables (r2 0.02–0.49).



1090

Discussion As in earlier studies,8,21 this study shows [3H]pirenzepine binding to be diffuse across several laminae of the human cortex. Moreover, we have confirmed a previous finding of decreased [3H]pirenzepine binding in BA 9 from subjects with schizophrenia,7,8 but in tissue from new cohorts of subjects. In addition, this study shows decreased [3H]pirenzepine binding is not ubiquitous throughout the cortex in schizophrenia, since it was not detected in BA 40. Our findings with [3H]pirenzepine are consistent but an increase in [3H]quinuclidinyl benzilate binding ([3H]-QNB) to muscarinic receptors has been reported in the frontal cortex of subjects with schizophrenia being treated with antipsychotic drugs until death.22 There are several explanations for this apparent discrepancy, the most obvious being that our study uses a M1R/M4R selective ligand11 with autoradiography, whilst the other uses a nonselective muscarinic receptor radioligand with particulate membrane.22 Thus, differences between studies could be purely due to methodological variation. Alternatively, it is possible that increases in M2, M3 and/or M5 in BA 9 from schizophrenic subjects, that would not be detected with [3H]pirenzepine, are of a sufficient order of magnitude to give an overall increase in [3H]-QNB.22 To understand the cause of the decrease in [3H]pirenzepine binding in the cortex of subjects with schizophrenia, we measured M1 and M4R proteins using receptor-specific antibodies. These studies show there is a decrease in M1R in BA 9, but not BA 40, from subjects with schizophrenia whilst the levels of M4R are not changed in either region. These data suggest that decreases in [3H]pirenzepine binding in BA 9 from subjects with schizophrenia reflect a reduction in M1R. The experimental design of our study does not allow a comparison of relative levels of M1R and M4R protein within BA 9 or BA 40. By contrast, an earlier study using immunoprecipitation of radiolabelled muscarinic receptors with receptor-specific antibodies23 showed that M1R is the predominant muscarinic receptor in the human cortex. Unfortunately in the immunoprecipitation study, muscarinic receptors were only measured in one cortical region, so it is not possible to compare our results on variations of muscarinic receptors between cortical regions. Our study has shown a decrease in M1R mRNA and [3H]pirenzepine binding in BA 9 from subjects with schizophrenia. These findings contrast with our studies in the caudate-putamen where we showed there was a decrease in [3H]pirenzepine binding10 that was not associated with a decrease in levels of M1R mRNA.17 Hence, further studies will be needed to identify the causes of changed [3H]pirenzepine binding in other CNS regions in schizophrenia. Significantly, the decreases we have shown in M1R and M4R mRNA in BA 40 from subjects with schizophrenia are not accompanied by a decrease in receptor protein or radioligand binding. Dissociation between receptor mRNA levels and other measures of receptor protein have


been reported previously10,17,24 in studies using tissue obtained postmortem. It remains to be determined if changes in mRNA alone have physiological significant outcomes that would result in altered CNS function. One confounding factor deserves consideration in relation to this study, which is that all individuals with schizophrenia from whom tissue was obtained for this study had a record of receiving antipsychotic drugs. Significantly, the absence of a relationship between any experimental measure and final recorded antipsychotic drug dose suggests that this confounding factor was not a significant contributor to the reduction of M1R in BA 9. A similar analysis suggests that antipsychotic drugs were not the cause of the reduction in M4R mRNA in both CNS regions studied. Moreover, [3H]pirenzepine binding has been shown to be only marginally reduced (苲10%) in the cortex of rats treated for 4 weeks with haloperidol but not clozapine.8 In that study, rats treated for longer periods (up to 7 months) had increased cortical [3H]pirenzepine binding, a finding that is in line with the study of muscarinic receptors in schizophrenics receiving antipsychotic drugs at death.22 Thus, available data would not support the hypothesis that differences relating to M1R and M4R in schizophrenia in this study are simply due antipsychotic drug treatment. Unlike in our previous study,8 this study did not find a significant difference in [3H]pirenzepine binding in tissue from subjects who had a record of receiving anticholinergic drugs in their clinical history. Notably, none of the subjects in this study had detectable levels of anticholinergic drugs in their blood at autopsy, suggesting anticholinergic drugs were not administered close to death. Thus, it is possible that any effect of these drugs on [3H]pirenzepine binding was no longer apparent. Levels of M4R mRNA were decreased in both BA 9 and BA 40 from subjects with schizophrenia and hence a role for that receptor in the pathology of the illness cannot be excluded. However, the M4R protein was not altered in either cortical region and radioligand binding was normal in BA 40. Thus, despite the change in levels of mRNA for the M4R, the presence of normal levels of receptors in subjects with schizophrenia should nullify changes in M4R expression at the level of mRNA. The decrease in M1R protein, mRNA and radioligand binding makes it likely there are changes in M1R function in BA 9 from subjects with schizophrenia. Importantly, the effects of a deficiency in M1R can now be studied in an M1R knockout mouse.25 Studies in these mice have shown that deletion of the M1R leads to disturbances in learning, memory25 and circadian rhythms.26 Since these are functions that are all affected in subjects with schizophrenia27,28 it would seem that data from M1R knockout mice support the hypothesis that abnormalities in M1R could play a role in the pathology of the illness. In this study, we examined two cortical regions to investigate whether M1R and M4R in the frontal cortex, which is more influenced by cholinergic projections


from the basal forebrain,12 may be more affected in schizophrenia. The localisation of changes in M1R to BA 9 is preliminary data to support this hypothesis. These findings support the hypothesis that drugs which target specific muscarinic receptors will be beneficial in treating the symptoms of schizophrenia.29,30 Further investigations of changes in specific muscarinic receptors in a larger number of cortical regions will be required to fully determine the regional specificity of changes in muscarinic receptors in schizophrenia. Acknowledgements The authors would like to acknowledge Dr Ken Opeskin and the Staff at the Victorian Institute of Forensic Medicine for their assistance with the collection of tissue, Professor Nicholas Keks and Ms Christine Hill for providing diagnostic confirmation and Mr Geoffrey Pavey for his efforts in curating the tissue collection at the Mental Health Research Institute. This work was financed in part by Grants-in-Aid from the Rebecca L Cooper Research Laboratories, the Stanley Foundation and the Wood’s Family Trust and NH&MRC Project (grant 209046).

References 1 Tandon R. Cholinergic aspects of schizophrenia. Br J Psychiatry Suppl 1999; 37: 7–11. 2 Hyde TM, Crook JM. Cholinergic systems and schizophrenia: primary pathology or epiphenomena. J Chem Neuroanatomy 2001; 22: 53–63. 3 Bymaster FP, Calligaro DO, Falcone JF, Marsh RD, Moore NA, Tye NC et al. Radioreceptor binding profile of the atypical antipsychotic olanzapine. Neuropsychopharmacology 1996; 14: 87–96. 4 Kane JM, McGlashan TH. Treatment of schizophrenia. Lancet 1995; 346: 820–825. 5 Reus VI. Olanzapine: a novel atypical neuroleptic agent. Lancet 1997; 349: 1264–1265. 6 Eglen RM, Choppin A, Watson N. Therapeutic opportunities from muscarinic receptor research. Trends Pharmacol Sci 2001; 22: 409–414. 7 Dean B. A predicted cortical serotonergic/cholinergic/GABAergic interface as a site of pathology in schizophrenia. Clin Exp Pharmacol Physiol 2001; 28: 74–78. 8 Crook JM, Tomaskovic-Crook E, Copolov DL, Dean B. Low muscarinic receptor binding in prefrontal cortex from subjects with schizophrenia: a study of Brodmann’s areas 8, 9, 10, and 46 and the effects of neuroleptic drug treatment. Am J Psychiatry 2001; 158: 918–925. 9 Crook JM, Tomaskovic-Crook E, Copolov DL, Dean B. Decreased muscarinic receptor binding in subjects with schizophrenia: a study of the human hippocampal formation. Biol Psychiatr 2000; 48: 381–388. 10 Dean B, Crook JM, Opeskin K, Hill C, Keks N, Copolov DL. The density of muscarinic M1 receptors is decreased in the caudateputamen of subjects with schizophrenia. Mol Psychiatry 1996; 1: 54–58.

11 Hulme EC, Birdsall NJM, Buckley NJ. Muscarinic receptor subtypes. Annu Rev Pharmacol Toxicol 2000; 30: 633–673. 12 Perry E, Walker M, Grace J, Perry R. Acetylcholine in mind: a neurotransmitter correlate of consciousness? Trends Neurosci 1999; 22: 273–280. 13 Kingsbury AE, Foster OJ, Nisbet AP, Cairns N, Bray L, Eve DJ et al. Tissue pH as an indicator of mRNA preservation in human postmortem brain. Mol Brain Res 1995; 28: 311–318. 14 Keks N, Hill C, Opeskin K, Copolov DL, Dean B. Psychiatric diagnosis after death: the problems of accurate diagnosis. In: Dean B, Hyde TM, Kleinman JE (eds). Using CNS Tissue in Psychiatric Research: A Practical Guide. J Gordon & Breach Science Publishers: Sydney, 1999, pp 19–37. 15 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (Fourth edition). American Psychiatric Association: Washington, DC, 1994. 16 Foster P. Neuroleptic equivalence. Pharmaceut J 1998; 243: 431– 432. 17 Dean B, Crook JM, Pavey G, Opeskin K, Copolov DL. Muscarinic1 and 2 receptor mRNA in the human caudate-putamen: no change in m1 mRNA in schizophrenia. Mol Psychiatry 2000; 5: 203–207. 18 Loiacono RE, Gundlach AL. In situ hybridisation histochemistry: application to human brain tissue. In: Dean B, Hyde TM, Kleinman JE (eds). Using CNS Tissue in Psychiatric Research: A Practical Guide. Harwood Academic Press: Sydney, 1999, pp 85–106. 19 Ndoye A, Buchli R, Greenberg B, Nguyen VT, Zia S, Rodriguez JG et al. Identification and mapping of keratinocyte muscarinic acetylcholine receptor subtypes in human epidermis. J Invest Dermatol 1998; 111: 410–416. 20 Buchli R, Ndoye A, Rodriguez JG, Zia S, Webber RJ, Grando SA. Human skin fibroblasts express m2, m4, and m5 subtypes of muscarinic acetylcholine receptors. J Cell Biochem 1999; 74: 264–277. 21 Vogt BA, Plager MD, Crino PB, Bird ED. Laminar distributions of muscarinic acetylcholine, serotonin, GABA and opioid receptors in human posterior cingulate cortex. Neuroscience 1990; 36: 165–174. 22 Watanabe S, Nishikawa T, Takashima M, Toru M. Increased muscarinic cholinergic receptors in prefrontal cortices of medicated schizophrenics. Life Sci 1983; 33: 2187–2196. 23 Flynn DD, Ferrari-DiLeo G, Mash DC, Levey AI. Differential regulation of molecular subtypes of muscarinic receptors in Alzheimer’s disease. J Neurochem 1995; 64: 1888–1891. 24 Burnet PW, Eastwood SL, Harrison PJ. 5-HT1A and 5-HT2A receptor mRNAs and binding site densities are differentially altered in schizophrenia. Neuropsychopharmacology 1996; 15: 442–455. 25 Hamilton SE, Loose MD, Qi M, Levey AI, Hille B, McKnight GS et al. Disruption of the m1 receptor gene ablates muscarinic receptordependent M current regulation and seizure activity in mice. Proc Natl Acad Sci U S A 1997; 94: 13311–13316. 26 Gillette MU, Buchanan GF, Artinian L, Hamilton SE, Nathanson NM, Liu C. Role of the M1 receptor in regulating circadian rhythms. Life Sci 2001; 68: 2467–2472. 27 Heinrichs RW, Zakzanis KK. Neurocognitive deficit in schizophrenia: a quantitative review of the evidence. Neuropsychology 1998; 12: 426–445. 28 Keshavan MS, Reynolds CF III, Miewald MJ, Montrose DM, Sweeney JA, Vasko RC Jr et al. Delta sleep deficits in schizophrenia: evidence from automated analyses of sleep data. Arch Gen Psychiatry 1998; 55: 443–448. 29 Felder CC, Porter AC, Skillman TL, Zhang L, Bymaster FP, Nathanson NM et al. Elucidating the role of muscarinic receptors in psychosis. Life Sci 2001; 68: 2605–2613. 30 Shannon HE, Rasmussen K, Bymaster FP, Hart JC, Peters SC, Swedberg MD et al. Xanomeline, an M(1)/M(4) preferring muscarinic cholinergic receptor agonist, produces antipsychotic-like activity in rats and mice. Schizophr Res 2000; 42: 249–259.

1091
