Comparative Analysis of Group II Metabotropic Glutamate Receptor

Home , Brodmann area 46

Molecular Psychiatry (2002) 7, 157–164  2002 Nature Publishing Group All rights reserved 1359-4184/02 $25.00 www.nature.com/mp ORIGINAL RESEARCH ARTICLE Comparative analysis of group II metabotropic glutamate receptor immunoreactivity in Brodmann’s area 46 of the dorsolateral prefrontal cortex from patients with schizophrenia and normal subjects JM Crook, M Akil, BCW Law, TM Hyde and JE Kleinman

Section on Neuropathology, Clinical Brain Disorders Branch, National Institute of Mental Health, Bethesda, MD 20892, USA

Glutamate is the primary excitatory neurotransmitter in the mammalian central nervous system, and a key neurotransmitter in prefrontal cortical function. Converging lines of evidence implicate prefrontal cortical dysfunction in the neurobiology of schizophrenia. Thus, aberrant glutamate neurotransmission may underlie schizophrenia and other complex disorders of behavior. Group II metabotropic receptors (mGluRs) are important modulators of glutamatergic and non-glutamatergic neurotransmission. Moreover, in an animal model, an agonist for group II mGluRs has been shown to reverse the behavioral, locomotor, and cognitive effects of the psychotomimetic drug phencyclidine. Accordingly, group II mGluRs constitute attractive targets for the pharmacotherapeutics and study of schizophrenia. Using immunocytochemistry and Western immunoblotting, we compared the localization and levels of group II mGluRs in Brodmann’s area 46 of the dorsolateral prefrontal cortex from patients with schizophrenia and normal subjects. Consistent with previous reports, we found that immunolabeling of group II mGluRs is prominent in Brodmann’s area 46. The majority of labeling was present on axon terminals distributed in a lamina-speciﬁc fashion. No apparent difference in the cellular localization or laminar distribution of immunoreactive group II mGluRs was noted between the two diagnostic groups. Similarly, the levels of receptor immunoreactivity determined by quantitative Western immunoblotting were comparable between schizophrenic patients and normal subjects. We conclude that while the function of group II mGluRs in Brodmann’s area 46 of dorsolateral prefrontal cortex may be altered in patients with schizophrenia, this is not evident at the level of protein expression using an antibody against mGluR2 and mGluR3. Molecular Psychiatry (2002) 7, 157–164. DOI: 10.1038/sj/mp/4000966 Keywords: mGluR2/3; Western immunoblotting; immunocytochemistry; glutamate neurotransmission; Brodmann’s area 46

Introduction brain.1 Taken together with the putative involvement of BA 46 in the pathophysiology of schizophrenia,5–11 Glutamate is a primary excitatory neurotransmitter in the general role of glutamate in CNS function and the the mammalian central nervous system (CNS), and a anatomy of glutamatergic neurons in BA 46 support a putative transmitter of many clinically important path- key role for glutamate dysfunction in DLPFC of schizo- ways.1 Quantitatively, glutamate is the most abundant phrenia. amino acid transmitter in the CNS,1–4 and glutamate Hypothetical models of glutamate dysfunction in receptors are located on virtually all neurons of the schizophrenia include hypofunction12,13 and hyper- CNS.4 In Brodmann’s area (BA) 46 of the dorsolateral function14,15 of glutamate neurotransmission. The prefrontal cortex (DLPFC), pyramidal neurons appear model of glutamate hypofunction is based, in part, on to be glutamatergic, forming efferent corticofugal pro- the psychotomimetic properties of phencyclidine jections to other cortical and subcortical regions of the (PCP) and ketamine. These N-methyl-D-aspartate (NMDA) ionotropic glutamate receptor antagonists can induce positive and negative symptoms in healthy Correspondence: Dr TM Hyde, Section on Neuropathology, Clini- individuals,16–18 and the exacerbation of psychosis in cal Brain Disorders Branch, National Institute of Mental Health, patients with schizophrenia.19,20 In addition, PCP and Bldg 10, Room 4N 306 MSC 1385, Bethesda, MD 20892, USA. E-mail: hydetȰintra.nimh.nih.gov ketamine impair performance on prefrontal cortex- 21–23 Received 26 February 2001; revised 16 May 2001; accepted 27 dependent cognitive tasks, by interacting with June 2001 dopamine neurotransmission,14,22,24 mirroring cogni- Metabotropic glutamate receptors in schizophrenia JM Crook et al 158 tive impairment in schizophrenia. Assuming that PCP Materials and methods and ketamine act postsynaptically, and the resulting Tissue collection behavioral dysfunction is a valid model of schizo- After gaining the consent of a donor’s legal next of kin, phrenia, the behavioral abnormalities produced by human DLPFC was obtained at autopsy through the these drugs suggest reduced glutamate tone in the Medical Examiners’ Office of the District of Columbia DLPFC of patients with schizophrenia. Paradoxically, (Washington DC, USA). All tissue was obtained and recent studies have shown that psychotomimetic processed in accordance with a research protocol NMDA receptor antagonists increase glutamate efflux, reviewed and approved by the Institutional Review perhaps via presynaptic autoreceptors,25–27 thereby Board of the National Institute of Mental Health, in potentiating PFC-glutamate neurotransmission at post- conformance with the standards and guidelines of the synaptic non-NMDA (alpha-amino-3-hydroxy-5- National Institutes of Health. Tissue specimens were methyl-4-isoxazolepropionic acid (AMPA), and collected from 20 patients with a diagnosis of schizo- kainate) receptors.14 Importantly, antagonism of iono- phrenia and 20 subjects with no clinical history of psy- tropic AMPA/kainate receptors in PFC attenuates mne- chiatric illness or neuroleptic exposure (normal monic and other behavioral effects of NMDA receptor subjects). Patient diagnoses were confirmed after an antagonism.14 independent review of medical records by two board While most studies of glutamate receptor mediated certified psychiatrists according to DSM-IV criteria.42 neurotransmission in schizophrenia have focused on A board certified neuropathologist conducted an exam- ionotropic glutamate receptors, recently, G-protein ination of each case in order to exclude any subject coupled mGluRs have received increased attention.28,29 with significant neuropathological abnormalities. As modulators of synaptic neurotransmission,15,30,31 Microscopic examination of brain tissue sections taken mGluRs are functionally related to NMDA, AMPA, and from multiple cerebral areas and stained with Biel- kainate receptors. Also, mGluRs are involved in neur- schowsky’s silver stain was performed to exclude the oplasticity,30,32,33 are characterized by a diversity of presence of pathology associated with Alzheimer’s dis- receptor subtypes and heterogeneous localization,30,34 ease. Toxicology screens conducted on blood or brain and constitute attractive targets for the pharmacothera- samples showed no evidence of alcohol or drug abuse peutics of psychiatric disorders associated with by any of the subjects studied. increased or decreased glutamate neurotransmis- Wherever possible, normal subjects were matched sion.34,35 for gender, age at death, postmortem interval (PMI; the time between death and freezing of brain tissue), and The mGluRs currently recognized are classified into pH of tissue to subjects who had schizophrenia (Table three groups according to sequence homology, signal 1). Where death was not witnessed (normal subjects: transduction mechanism, and agonist selectivity.34,36 ID 7, 8; schizophrenic patients: ID 4, 6; Table 1), PMI Group II mGluRs consist of mGluR2 and mGluR3, was taken as the interval half way between the last which modulate synaptic transmission through inhi- 37,38 sighting of a subject while still alive and being found bition of cyclic AMP or calcium ion channels. dead (less than 12 h), to the time of tissue freezing. For Group II mGluRs are typically presynaptic autorecep- 30,34 34 schizophrenic patients, duration of illness (DOI; the tors, but also serve as presynaptic heteroceptors. time from first hospital admission to death), and final Moreover, they provide neuroprotection against gluta- and average daily antipsychotic drug doses (calculated 39,40 mate-induced excitotoxicity, and are prevalent in as chlorpromazine equivalents) were determined from 41 frontal cortex. Finally, in an animal model, group II the medical records (Table 1). mGluRs have been shown to mediate, in part, the dis- Following autopsy, the cerebral hemispheres of each ruptive behavioral, locomotor, and cognitive effects of brain were blocked coronally and flash frozen using a PCP, and the concomitant cortical glutamate mixture of dry ice and isopentane. Tissue blocks were efflux.15,30,34 stored at −80°C until required (freezer time, FT; Table Based on current understanding of the functional 1). The selection of tissue blocks containing BA 46 of and anatomical characteristics of group II mGluRs, we DLPFC was standardized between subjects. Blocks hypothesized that the levels and/or anatomical distri- extended rostral to the genu of the corpus callosum and bution of these receptors are altered in BA 46 of the contained the middle frontal gyrus. For immunocyto- DLPFC from patients with schizophrenia compared to chemistry, 14-␮m tissue sections were cut at −20°C normal subjects. To investigate this hypothesis, we using a cryo-microtome (CM 3050, Leica Microsystems, used immunocytochemistry with light microscopy to Deerfield, IL, USA), and thaw-mounted onto chrome- identify the localization of group II mGluRs in BA 46 alum/gelatin-coated glass slides and stored at −80°C of DLPFC from schizophrenic patients and normal sub- until required. jects. In addition, we performed Western immunoblot- Immunocytochemistry ting with scanning densitometry to measure mGluR2/3 Mounted tissue sections (two per subject) were receptor immunoreactivity in BA 46 of DLPFC from removed from −80°C, briefly thawed at room tempera- schizophrenic patients and normal subjects, including ture (RT), and fixed for 2 h at 4°C in 4% phosphate- those chosen for immunocytochemistry studies. buffered paraformaldehyde (pH 7.2). Sections were washed three times for 5 min in 0.02 M phosphate-

Molecular Psychiatry Metabotropic glutamate receptors in schizophrenia JM Crook et al 159 Table 1 Demographic, postmortem, and pharmacological data for schizophrenic patients and normal subjects studied, and mGluR2/3 immunoblot density in the dorsolateral prefrontal cortex

ID Gender Age at PMI pH Freezing DOI Medication Final Average daily mGluR2/3 death (hours) time (years) medication medication immunoblot (years) (months) dosea dosea densityb

Schizophrenic Patients 1c F 67 39.5 6.63 100 37 Haloperidol 80 100 1394 2c M 31 15.5 6.46 92 14 Thioridazine 200 250 1132 Diphenhydramine 3c M 23 43 6.48 92 1 Haloperidol 400 480 385 Carbamazepine Nortriptyline Benztropine Lorazepam 4c F 60 19 6.38 90 20 Haloperidol 100 100 347 5c M 30 72.5 6.32 86 14 Haloperidol 1900 500 472 Carbamazepine 6c F 81 11 6.78 80 54 Haloperidol 100 150 1765 7c M 41 32 6.63 56 21 Haloperidol 50 400 1342 8 F 71 47.5 6.41 161 56 Serentil 100 500 337 9 M 48 48.5 6.42 162 26 Haloperidol 2000 275 784 Benztropine 10 M 48 40.5 5.70 165 28 Haloperidol 900 533 275 11 M 36 13 6.56 172 15 Prolixin 400 850 1543 decanoate 12 F 44 36.5 6.51 171 15 Haloperidol 200 200 1218 13 F 71 32.5 6.63 176 49 Serentil 400 650 970 14 M 46 25 6.73 176 9 Haloperidol 300 300 148 15 M 54 24 6.31 139 31 Prolixin 1600 800 237 decanoate 16 M 48 15 6.29 126 15 Chlorpromazine 300 300 1437 Benadryl 17 F 64 20.5 6.48 101 45 Haloperidol N/A 200 447 Clonazepam 18c M 75 41.5 6.29 101 46 Haloperidol 400 400 251 Carbamazepine Nortriptyline 19c M 44 35 6.28 62 18 Haloperidol 300 350 605 Fluphenazine Benztropine 20c F 41 51.5 6.08 60 25 Haloperidol 2400 1135 670 Fluphenazine Mean ± s.e. 51 ± 3.6 33 ± 3.5 6.4 ± 0.1 118 ± 10 788 ± 115

(Continued)

buffered saline (PBS; pH 7.4), and incubated with 1% diluted in blocking solution. After washing, the sec-

H2O2 for 8 min at RT to reduce the activity of endogen- tions were incubated for 1 h at RT with biotinylated ous peroxidases. Sections were again washed, and goat anti-rabbit IgG secondary antibody (1:200 dilution; incubated in humidified chambers with 10% normal Vector Laboratories, Burlingame, CA, USA) diluted in goat serum in blocking solution (1% bovine serum blocking solution. Sections were washed again, and albumin (BSA; Sigma, St Louis, MO, USA), 0.1% Tri- incubated with avidin-biotin/peroxidase complex sol- ton X-100 (Sigma), and PBS) for 30 min at RT to block ution (VECTASTAIN ABC Elite Kit; Vector non-specific labeling. Serum was removed, and the sec- Laboratories) for 1 h at RT. Sections were then washed, tions were incubated overnight at 4°C in humidified preincubated with 3,3Ј-diaminobenzidine (DAB; chambers with affinity-purified rabbit polyclonal Sigma) for 30 min at RT, and finally reacted with DAB ␮ −1 mGluR2/3 antibody (1:200 dilution; 1.5 gml ; raised and 0.03% H2O2 in PBS for 10 min at RT. After a final against the carboxy terminus peptide of rat mGluR2 wash, sections were dehydrated, cleared in xylene, and [NGREVVDSTTSSL] conjugated to BSA with glutaral- coverslipped for light microscopic examination. dehyde; Chemicon International, Temecula, CA, USA) In addition to Western immunoblot studies of speci-

Molecular Psychiatry Metabotropic glutamate receptors in schizophrenia JM Crook et al 160 Table 1 Continued

ID Gender Age at PMI pH Freezing DOI Medication Final Average daily mGluR2/3 death (hours) time (years) medication medication immunoblot (years) (months) dosea dosea densityb

Normal Subjects 1c F 52 12 6.87 80 1217 2c M 41 14.5 6.72 79 706 3c M 24 12.5 6.59 76 653 4c M 38 32.5 6.14 75 1677 5c M 18 14.5 6.51 73 2625 6c F 57 19 6.43 56 326 7c F 59 37 6.57 52 789 8c F 67 34 6.69 52 703 9 F 53 22 5.76 155 199 10 M 34 37 6.60 165 803 11 M 68 19.5 6.57 172 242 12 M 47 24.5 6.54 178 726 13 F 58 26.5 6.54 206 821 14 F 39 41.5 6.34 206 666 15c F 52 26.5 6.38 63 763 16 M 45 17 6.61 124 563 17 M 47 59.5 6.03 137 454 18 F 60 10 6.40 125 696 19c M 56 34 6.09 68 798 20 M 77 20.5 6.09 126 672 Mean ± s.e. 50 ± 3.3 26 ± 2.8 6.4 ± 0.1 113 ± 12 805 ± 120

PMI, postmortem interval; DOI, duration of illness; M, male; F, female; s.e., standard error; N/A, not available. aChlorpromazine equivalents (mg day−1). bArbitrary units. cSubjects selected for qualitative immunocytochemistry studies.

ficity of antibody labeling, immunocytochemistry was samples (50 ␮g of protein) were stored in Laemmli performed on sections with the exclusion of the pri- reducing sample buffer45 at −80°C until required. mary or secondary antibody. Isolated samples (two per subject) were subjected to Tissue preparations were examined by light SDS-PAGE (7.5% polyacrylamide gels) and transferred microscopy (Nikon, Eclipse E400). BA 46 was histo- by electroblotting to nitrocellulose membrane over- logically characterized by applying cytoarchitectonic night. Blots were blocked for 2 h at RT in blocking criteria to Nissl stained sections from each subject buffer (5% non-fat dry milk, 0.1% Tween 20, 50 mM studied.43 A few sections were counterstained with the Tris-buffered saline (TBS; pH 7.5)), and incubated for modified Giemsa method previously described44 in 1 h at RT with the above described mGluR2/3 antibody order to further facilitate characterization of laminar (1:80 dilution, 1.25 ␮gml−1) in blocking solution. Blots distribution. were washed six times for 5 min in TBS with 0.1% Tween 20, incubated for 1 h at RT with horseradish Western immunoblotting with scanning densitometry peroxidase-conjugated secondary antibody (1:5000 Tissue samples were removed from −80°C and immedi- dilution, 0.08 ␮gml−1, goat anti-rabbit IgG) prepared in ately homogenized at 4°C in buffer (20 mM Tris-HCl blocking buffer, and washed again. The immunoreac- (pH 7.4) containing 10% sucrose, 1 mM EDTA, 1 mM tive proteins were detected using enhanced chemi- EGTA, 1 mM PMSF, 20 mg ml−1 leupeptin, 10 mg ml−1 luminescence (Amersham Pharmacia Biotech, Piscata- pepstatin A, 20 mg ml−1 chymostatin, 2 mg ml−1 aproti- way, NJ, USA) with exposure to Kodak BioMax film nin, and 10 mM benzamidine). Homogenates were cen- (Eastman Kodak Company, Rochester, NY, USA). trifuged at 1000 g for 10 min at 4°C. The supernatant Specificity of primary and secondary antibodies was was re-centrifuged in buffer without sucrose, at verified by performing Western immunoblot studies on 114 000 g for 20 min at 4°C. The resulting pellet was samples under the following conditions: (1) an optimal washed in the same buffer three times by resuspension, (1:80 dilution) concentration of primary antibody pre- and centrifuged at 114 000 g for 20 min at 4°C. The absorbed for 2 h at 4°C with an excess of synthetic pellet was resuspended in buffer and placed on ice. mGluR2 (6.25 ␮gml−1) peptide (Chemicon After determining the protein concentrations of iso- International); and (2) exclusion of the primary anti- lated samples using Bio-Rad DC Protein Assay (Bio- body. Rad Laboratories, Hercules, CA, USA), aliquots of Film-based images of Western immunoblots were

Molecular Psychiatry Metabotropic glutamate receptors in schizophrenia JM Crook et al 161 scanned using a Hewlett-Packard Scanjet at 600 d.p.i. of mGluR2 and mGluR3 (Figure 2a–c). Preabsorbtion resolution. Quantitative analysis of mGluR2/3 immun- of an optimal concentration (1:80 dilution) of primary opositive bands was performed on a Macintosh com- antibody with an excess of synthetic peptide (Figure puter using the public domain NIH Image program 2d–e) or omission of the primary antibody (Figure 2f) (developed at the US National Institutes of Health and resulted in an absence of immunoreactivity. available on the Internet at http://rsb.info.nih.gov/nih- There was no difference between the mean levels of image/). Scanning densitometry was performed blind mGluR2/3 immunoreactivity in tissue from schizo- to diagnosis. Briefly, densitometric analysis was perfor- phrenic and normal subject groups (t = 0.10, df = 38, P med by plotting the intensity and area of each immuno- = 0.92; Table 1; Figure 3). Of the normal subject group, positive band, and calculating the area under the curve. three individuals showed considerably higher immu- To control for inter-blot variation, data from each noreactivity levels compared to other subjects (Table 1, immunoblot were standardized using immunopositive Figure 3). Importantly, although exclusion of the three bands of two reference internal controls. Also, to con- outliers from an analysis of data resulted in lower firm equal loading of total protein and thus control for mean immunoreactivity levels in tissue from normal intra-gel variation, Ponceau S Red staining of immuno- subjects compared to schizophrenic patients (622 ± 48 blots was performed. and 788 ± 115 respectively; mean ± s.e.), this was not statistically significant (t = 1.25, df = 35, P = 0.22). Statistical analysis The mean age (t = 0.32, df = 38, P = 0.75), PMI (t = Due to the relatively robust size of the sample popu- 1.7, df = 38, P = 0.10), FT (t = 0.33, df = 38, P = 0.74), lations, together with a greater sensitivity (ie, greater and pH of tissue (t = 0.04, df = 38, P = 0.97), did not statistical power) of normal distribution-based tests, differ between the groups studied (Table 1). For the parametric methods of analysis were considered appro- schizophrenic cohort, no relationship was found priate for the present studies. Thus, an unpaired t-test between immunoreactivity and age at death (r = 0.04, was used to compare immunolabeling, age, PMI, FT, P = 0.85), PMI (r =−0.41, P = 0.07), and FT (−0.09, P = and pH between the schizophrenic and normal groups. 0.69). There was a significant positive correlation Pearson Product-Moment correlation coefficients, between immunoreactivity in tissue of schizophrenic derived using a staight-line fit, were used to assess the patients and tissue pH (r = 0.46, P = 0.04). For the con- relationships between immunolabeling and age, PMI, trol subjects, no relationship was found between FT, and pH, and where relevant, DOI, final recorded immunoreactivity and PMI (r =−0.15, P = 0.53), FT (r antipsychotic drug dose, and average daily antipsy- =−0.31, P = 0.19), and tissue pH (r = 0.17, P = 0.47). chotic drug dose. Analysis of covariance (ANCOVA) Unlike schizophrenic patients, a significant negative was performed to determine whether age at death, PMI, correlation existed between immunoreactivity in tissue FT, or tissue pH were confounding variables in the from control subjects and age at death (r =−0.55, P = comparison between immunoreactivity of schizo- 0.01). ANCOVA showed no effect of age at death, PMI, phrenic and normal subjects. FT, or tissue pH on the comparison between immunoreactivity in tissue from schizophrenic and normal subjects (f = 0.20; df = 1, 34; P = 0.66). Finally, no relation- Results ship was demonstrated between immunoreactivity in Immunocytochemistry tissue from schizophrenic subjects and DOI (r = 0.35, Light microscopy showed similar patterns of P = 0.14), final recorded antipsychotic drug dose (r = mGluR2/3 immunoreactivity in BA 46 for patients with −0.32, P = 0.18), or average daily antipsychotic drug schizophrenia and normal controls. Punctate dose (r =−0.17, P = 0.47). immunostaining outlined unlabeled cell bodies throughout the thickness of cortex (Figure 1a), with no Discussion labeling in the white matter. A band of highest-density immunolabeled puncta corresponded to cortical lam- Using an antibody selective for mGluR2/3, we exam- ina III (Figure 1a). Occasionally, thin, axon-like immu- ined the localization and levels of immunoreactivity of nolabeled fibers were seen in superficial laminae. group II mGluRs in BA 46 of DLPFC from patients with Immunolabeled puncta were observed surrounding schizophrenia and normal subjects. Importantly, stud- predominantly pyramidal and occasionally non-pyr- ies to characterize the specificity of the antiserum used amidal neuronal and glial profiles (Figure 1b–c). A pro- for both slide-based immunocytochemistry and West- portion of pyramidal neurons exhibited darkly stained ern immunoblotting are consistent with the recognition puncta concentrated around the basal aspect of cell of group II mGluR proteins, as predicted from pre- somata (Figure 1b). Omission of the primary (Figure vious reports.41,46,47 1d) or secondary (Figure 1e) antibody resulted in the The immunocytochemical localization of group II absence of or minimal immunolabeling respectively. mGluRs in BA 46 is consistent with the role of these receptors in the modulation of both excitatory and Western immunoblotting with scanning densitometry inhibitory neurotransmission, deemed critical for nor- Western immunoblotting detected a major band of pro- mal cortical function. mGluR2/3 immunolabeled ter- tein at ෂ 100 kDa in isolated protein from BA 46 of nor- minals were closely associated with pyramidal and to mal human DLPFC, consistent with the predicted mass a lesser extent non-pyramidal somata. Presynaptic

Molecular Psychiatry Metabotropic glutamate receptors in schizophrenia JM Crook et al 162

Figure 1 Brightfield photomicrographs of mGluR2/3 immunolabeling in BA 46 of DLPFC from a normal subject. Panel (a) shows high-density punctate immunostaining in lamina III. Note the unlabeled profiles presumed to be somata of pyramidal (asterisks) and non-pyramidal (arrowheads) neurons, and glia (arrows). Panel (b) is a high magnification photomicrograph showing dense punctate immunolabeling at the base of a presumed pyramidal neuron (note the large triangular soma with the initial portion of an apical dendrite). Panel (c) shows an immunolabeled tissue section that has been counterstained with the modified Giemsa method. Immunolabeled puncta are black and Nissl substance is blue. Panel (c) confirms that the unlabeled shadows in panels (a) and (b) are indeed cell bodies of pyramidal neurons (asterisks), non-pyramidal neurons (arrowheads), and glia (arrows). Panels (d) and (e) show tissue sections processed with the exclusion of the primary or secondary antibody respectively. Scale bars = 50 ␮m (panel a), and 25 ␮m (panels b–e).

Figure 2 Western immunoblot of isolated mGluR2/3 proteins (arrow: ෂ 100 kD) in BA 46 of DLPFC from a normal subject. Lanes a–c illustrate immunoreactivity at 1:40, 1:80, 1:160 dilution of primary antibody respectively. Lanes d–f demonstrate specificity of primary and secondary antibodies. Immunoreactivity was abolished when a 1:80 dilution of primary antibody was preabsorbed with synthetic mGluR2/3 peptide (preabsorbed primary antibody (lane d); positive con- Figure 3 Levels of mGluR2/3 immunoreactivity in DLPFC trol for preabsorbtion control (lane e)), or excluded (lane f). from schizophrenic patients and normal subjects. Horizontal bars represent mean levels of immunoreactivity for each receptor localization was supported by labeling of study group. punctate elements at the periphery of cell bodies, including intense staining at the basal aspect of pyramidal somata. Taken together, our findings are con- mGluR2/3 protein expression is not appreciably alt- sistent with significant evidence30,34,41 for group II ered in BA 46 of patients with schizophrenia. mGluRs acting as autoreceptors at glutamatergic syn- While the present research represents the first post- apses and heteroreceptors at inhibitory synapses in BA mortem tissue study of mGluR2/3 protein expression 46 of the DLPFC. in schizophrenia, previous studies have focused on Although receptor localization studies demonstrate possible changes to mGluR2 and mGluR3 mRNA that group II mGluRs are anatomically positioned to expression in the disorder. Consistent with our find- underlie aberrant DLPFC function, this is not substan- ings, Richardson-Burns et al28 found no change to tiated by quantitative studies of immunoreactivity. mGluR2 and mGluR3 mRNA expression in the thala- Thus, Western immunoblot data indicate that mus of patients with schizophrenia. Similarly, Oh-

Molecular Psychiatry Metabotropic glutamate receptors in schizophrenia JM Crook et al 163 numa et al29 found no change to mGluR3 mRNA Acknowledgements expression in PFC of schizophrenic patients. The authors would like to acknowledge the contri- There are limitations to the present study, which bution made to the present study by Dr Mary M Her- should be considered when interpreting the data. First, man, Nikki Cohn, and Eva Tomaskovic-Crook. We are while the functional and anatomical characteristics of particularly indebted to the families of the individuals group II mGluRs support a similar role for both included in this study, and the Medical Examiners’ mGluR2 and mGluR3 in glutamate neurotransmission Office of the District of Columbia (Washington DC, and hence schizophrenia neurobiology, we cannot rule USA). out subtype specific receptor changes due to receptor selective mechanisms. Accordingly, subtype receptor selective antibodies will be necessary to determine if References each receptor is differentially altered. Similar to immu- 1 Greenamyre JT, Porter RHP. Anatomy and physiology of glutamate nochemical studies, the future development of subtype in the CNS. Neurology 1994; 44 (Suppl 8): S7–S13. receptor selective compounds for radioligand binding 2 Farber NB, Newcomer JW, Olney JW. The glutamate synapse in may prove useful to unmasking possible differences neuropsychiatric disorders: focus on schizophrenia and Alzhei- between mGluR2 and mGluR3 expression in schizo- mer’s disease. Prog Brain Res 1998; 116:421–437. 3 Cooper JR. Amino acid transmitters. In: The Biochemical Basis of phrenia. A second issue relates to the lack of anatom- Neuropharmacology. Oxford University Press: New York, 1996, pp ical resolution inherent with homogenized tissue stud- 116–193. ies. Using quantitative Western immunoblotting, 4 Fonnum F. Glutamate: a neurotransmitter in mammalian brain. J Neurochem 1984; 42:1–11. laminae- and or cellular-specific differences in 5 Bartha R, Williamson PC, Drost DJ, Malla A, Carr TJ, Cortese L et al. mGluR2/3 expression between the study groups would Measurement of glutamate and glutamine in the medial prefrontal not be detectable. Finally, as all schizophrenic patients cortex of never-treated schizophrenic patients and healthy controls studied were treated with antipsychotic drugs prior to by proton magnetic resonance spectroscopy. Arch Gen Psychiatry 1997; 54:959–965. death, we cannot exclude an effect of drug treatment on 6 Glantz LA, Lewis DA. Decreased dendritic spine density on pre- mGluR2/3 expression. Drug treatment may primarily or frontal cortical pyramidal neurons in schizophrenia. Arch Gen Psy- secondarily affect the expression of mGluR2 and/or chiatry 2000; 57:65–73. mGluR3 and mask possible changes associated with 7 Selemon LD, Rajkowska G, Goldman-Rakic PS. Elevated neuronal density in prefrontal area 46 in brains from schizophrenic patients: disease pathology. Interestingly, earlier animal studies application of a three-dimensional stereologic counting method. J indicate that levels of extracellular glutamate are not Comp Neurol 1998; 392:402–412. altered in PFC following chronic treatment with halo- 8 Lewis DA. Development of the prefrontal cortex during ado- peridol.48,49 Moreover, behavioral studies do not sup- lescence: insights into vulnerable neural circuits in schizophrenia. Neurophychopharmacology 1997; 16: 385–398. port the involvement of group II mGluRs in the activity 9 Karson CN, Mrak RE, Schluterman KO, Sturner WQ, Sheng JG, Grif- 50 of conventional antipsychotic drugs. fin WST. Alterations in synaptic proteins and their encoding In conclusion, it is clear that group II mGluRs are mRNAs in prefrontal cortex in schizophrenia: a possible neuro- abundant in BA 46 of human DLPFC, and anatomically chemical basis for ‘hypofrontality’. Mol Psychiatry 1999; 4:39–45. 10 Selemon LD, Goldman-Rakic PS. The reduced neuropil hypothesis: positioned to underlie cortical dysfunction. However, a circuit based model of schizophrenia. Biol Psychiatry 1999; 45: we were unable to find any appreciable change in the 17–25. localization or level of mGluR2/3 immunoreactivity in 11 Weinberger DR, Berman KF, Zec RF. Physiological dysfunction of BA 46 of schizophrenic patients. Alterations specificto dorsolateral prefrontal cortex in schizophrenia. I Regional cerebral blood flow (rCBF) evidence. Arch Gen Psychiatry 1986; 45: 609– mGluR2 or mGluR3, restricted to certain cortical lami- 615. nae, and/or associated with particular cell populations 12 Carlsson A, Hansson LO, Waters N, Carlsson ML. A glutamatergic may have been overlooked in the present study. The deficiency model of schizophrenia. Br J Psychiatry 1999; 174 present shortfalls may, in part, be overcome with the (Suppl 37): 2–6. 13 Bartha R, Williamson PC, Drost DJ, Malla A, Carr TJ, Cortese L et al. development of subtype receptor selective antibodies Measurement of glutamate and glutamine in the medial prefrontal for immunochemistry and pharmacological com- cortex of never-treated schizophrenic patients and healthy controls pounds for radioligand binding. Moreover, future stud- by proton magnetic resonance spectroscopy. Arch Gen Psychiatry ies using quantitative methods of immunohistochemis- 1997; 54:959–965. 14 Moghaddam B, Adams BW, Verma A, Daly D. Activation of gluta- try may prove more useful for identifying matergic neurotransmission by ketamine: a novel step in the path- ultrastructural changes in receptor levels and distri- way from NMDA receptor blockade to dopaminergic and cognitive bution. Finally, group II mGluRs continue to be of disruptions associated with the prefrontal cortex. J Neuroscience interest as targets for novel pharmacological agents for 1997; 17:2921–2927. 15 Moghaddam B, Adams BW. Reversal of phencyclidine effects by a the treatment of schizophrenia. This is supported by group II metabotropic glutamate receptor agonist in rats. Science the actions of recently developed mGluR2/3 selective 1998; 281: 1349–1352. agonists that show antispychotic activity, including 16 Javitt DC, Zukin SR. Recent advances in the phencyclidine model activation of dopaminergic and serotonergic brain of schizophrenia. Am J Psychiatry 1991; 148: 1301–1308. 17 Malhotra AK, Pinals DA, Weingartner H, Sirocco K, Missar CD, pathways previously associated with atypical antipsy- Pickar D et al. NMDA receptor function and human cognition: the chotics, and with low propensity for extrapyramidal effects of ketamine in healthy volunteers. Neuropsychopharmacol- side effects.51,52 ogy 1996; 14:301–307. 18 Newcomer JW, Farber NB, Jevtovic-Tedorovic V, Selke G, Melson AK, Hershey T et al. Ketamine-induced NMDA receptor hypofunc-

Molecular Psychiatry Metabotropic glutamate receptors in schizophrenia JM Crook et al 164 tion as a model of memory impairment and psychosis. Neuropsy- 36 Schoepp DD, Jane DD, Monn JA. Pharmacological agents acting at chopharmacology 1999; 20:106–118. subtypes of metabotropic glutamate receptors. Neuropharmacology 19 Lahti AC, Koffel B, LaPorte D, Tamminga CA. Subanesthetic doses 1999; 38: 1431–1476. of ketamine stimulate psychosis in schizophrenia. Neuropsycho- 37 Schoepp DD, Johnson BG, Monn JA. Inhibition of cyclic AMP for- pharmacology 1995; 13:9–19. mation by a selective metabotropic glutamate receptor agonist. J 20 Malhotra AK, Pinals DA, Adler CM, Elman I, Clifton A, Pickar D Neurochem 1992; 58: 1184–1186. et al. Ketamine-induced exacerbation of psychotic symptoms and 38 Chavis P, Shinozaki H, Bockaert J, Fagni L. The metabotropic gluta- cognitive impairment in neuroleptic-free schizophrenics. Neuro- mate receptor types 2/3 inhibit L-type calcium channels via a per- psychopharmacology 1997; 17: 141–150. tussis toxin-sensitive G-protein in cultured cerebellar granule cells. 21 Shroeder U, Schroeder H, Schwegler H, Sable BA. Neuroleptics J Neurosci 1994; 14: 7067–7076. ameliorate phencyclidine-induced impairments of short-term 39 Battaglia G, Bruno V, Ngomba RT, Di Grezia R, Copani A, Nicoletti memory. Br J Pharmacol 2000; 130:33–40. F. Selective activation of group-II metabotropic glutamate receptors 22 Verma A, Moghaddam B. The role of excitatory amino acids in pre- is protective against excitotoxic neuronal death. Eur J Pharmacol frontal cortex function as assessed by spatial delayed alternation 1998; 256:271–274. performance in rats: modulation by dopamine. J Neurosci 1996; 16: 40 Bond A, O’Neill MJ, Hicks CA, Monn JA, Lodge D. Neuroprotective 373–379. effects of a systemically active group II metabotropic glutamate 23 Parada-Turska J, Turski W. Excitatory amino acid antagonists and receptor agonist LY354740 in a gerbil model of global ischemia. memory: effect of drugs acting at N-methyl-D-aspartate receptors Neuroreport 1998; 9:1191–1193. in learning and memory tasks. Neuropharmacology 1990; 29: 41 Petralia RS, Wang YX, Niedzielski S, Wenthold RJ. The meta- 1111–1116. botropic glutamate receptors, mGluR2 and mGluR3, show unique 24 Krystal J, D’Souza DC, Karper LP, Bennett A, Abi-Dargham A, Abi- postsynaptic, presynaptic and glial localizations. Neuroscience Saab D et al. Interactive effects of subanesthetic ketamine and halo- 1996; 71: 949–976. peridol in healthy humans. Psychopharmacol (Berl) 1999; 145: 42 American Psychiatric Association. Diagnostic and Statistical Man- 193–204. ual of Mental Disorders, 4th edn. American Psychiatric Press: 25 Wang JK, Thukral V. Presynaptic NMDA receptors display physio- Washington, DC, 1994. logical characteristics of homomeric complexes of NRI subunits 43 Rajkowska G, Goldman-Rakic PS. Cytoarchitectonic deﬁnition of that contain the exon 5 insert in the N-terminal domain. J Neuro- prefrontal areas in the normal human cortex: II. Variability in chem 1996; 66: 865–868. locations of areas 9 and 46 and relationship to the talairach coordi- 26 Huntly GW, Vickers JC, Morrison JH. Cellular and synaptic localiz- nate system. Cereb Cortex 1995; 5:323–337. ation of NMDA and non-NMDA receptor subunits in neocortex: 44 Iniguez C, Gayoso M J, Carreres J. A versatile and simple method organization features related to cortical circuitry, function and dis- for staining nervous tissues using Giemsa dye. J Neurosci Meth ease. Trends Neurosci 1994; 17:536–542. 1985; 13:77–86. 27 Smirnova T, Stinnakre J, Mallet J. Characterization of a presynaptic 45 Laemmli UK. Cleavage of structural proteins during the assembly glutamate receptor. Science 1993; 262:430–433. of the head of bacteriophage T4. Nature 1979; 227:680–685. 28 Richardson-Burns SM, Haroutunian V, Davis KL, Watson SJ, 46 Testa CM, Friberg IK, Weiss SW, Standaert DG. Immunohistochem- Meador-Woodruff JH. Metabotropic glutamate receptor mRNA ical localization of metabotropic glutamate receptors mGluR1a and expression in the schizophrenic thalamus. Biol Psychiatry 2000; mGluR2/3 in the rat basal ganglia. J Comp Neurol 1998; 390:5–19. 47:22–28. 47 Yung KK. Localization of glutamate receptors in dorsal horn of rat 29 Ohnuma T, Augood SJ, Arai H, McKenna PJ, Emson PC. Expression spinal cord. Neuroreport 1998; 9: 1639–1644. of the human excitatory amino acid transporter 2 and metabotropic glutamate receptors 3 and 5 in the prefrontal cortex from normal 48 Daly DA, Moghaddam B. Actions of clozapine and haloperidol on individuals and patients with schizophrenia. Brain Res Mol Brain the extracellular levels of excitatory amino acids in the prefrontal Res 1998; 56:207–217. cortex and striatum of conscious rats. Neurosci Lett 1993; 152: 30 Pin JP, Duvoisin R. The metabotropic glutamate receptors: structure 61–64. and functions. Neuropharmacology 1995; 34:1–26. 49 Yamamoto BK, Cooperman MA. Differential effects of chronic anti- 31 Salt TE, Eaton SA. Function of metabotropic excitatory amino acid psychotic drug treatment on extracellular glutamate and dopamine receptors in sensory transmission in the thalamus: studies with concentrations. J Neurosci 1994; 14:4159–4166. novel phenyglycine antagonists. Neurochemistry 1994; 24: 451– 50 Ossowska K, Pietraszek M, Wardas J, Nowak G, Zajaczkowski W, 458. Wolfarth S et al. The role of glutamate receptors in antipsychotic 32 Anwyl R. Metabotropic glutamate receptors: electrophysiological drug action. Amino Acids 2000; 19:87–94. properties and role in plasticity. Brain Res Rev 1999; 29:83–120. 51 Cartmell J, Monn JA, Schoepp DD. The mGlu(2/3) receptor agonist 33 Holscher C, Gigg J, O’Mara S. Metabotropic glutamate receptor acti- LY379268 selectively blocks amphetamine ambulations and rear- vation and blockade: their role in long-term potentiation, learning, ing. Eur J Pharmacol 2000; 400:221–224. and neurotoxicity. Neurosci Biobehav Rev 1999; 23:399–410. 52 Cartmell J, Perry KW, Salhoff CR, Monn JA, Schoepp DD. The 34 Conn PJ, Pin JP. Pharmacology and functions of metabotropic gluta- potent, selective mGlu2/3 receptor agonist LY379268 increases mate receptors. Ann Rev Pharmacol Toxicol 1997; 37: 205–237. extracellular levels of dopamine, 3,4-dihydroxyphenylacetic acid, 35 Pellicciari R, Costantino G. Metabotropic G-protein-coupled gluta- homovanillic acid, and 5-hydroxyindole-3-acetic acid in the medial mate receptors as therapeutic targets. Curr Opin Chemical Biol prefrontal cortex of the freely moving rat. J Neurochem 2000; 75: 1999; 3:433–440. 1147–1154.

Molecular Psychiatry