0306-4522/90 163LKJ + 0.00 PergamonPresspk IQ 1990 IBRO
NeuroscienceVol. 39, NO. i, pp. 25-32, 1990 Print& in Great Britain
QUANTITATIVE AUTO~DIOGRAPHIC ANALYSIS OF GLUTAMATE BINDING SITES IN THE HIPPOCAMPAL FORMATION IN NORMAL AND SCHIZOPHRENIC BRAIN POS?’ ~~~T~~ R. Institute
KERWIN,*
S. PATEL and B.
MELDRUM
of Psychiatry, Departments of Neuroscience, Neurology and Psychiatry, De Crespigny Park, Denmark Hill, London SE5 SAF, U.K.
Abstract-Using quantitative autoradiography, the anatomical distribution of the binding sites (kainate, N-methyl-o-aspartate and q~~ualate) for the excitatory neurotransmitter @tamate has been established in the hjp~rnp~ fo~ation from control and ~~~phrenic brains, post martem.There is a loss of the kainate subtype particularly in ~~zoph~n~c ~pp~rnpi mainly from the CA&A, mossy fibre termination zone of the cornn ammonis (CA., and CA3; control and schizophrenic left hippocampus, respectively, 54.2 and 66.6 pmol/g; 18.3 and 17.9 pmol/g), as well as bilateral losses in the dentate gyrus (left 14.2 pmol/g and right 28.0 pmoljg; leR 9.5 pmol/g and right 7.9 pmol/g, control and schizophrenic, respectively) and parabippocampal gyrus (left 50.8pmoI/g and right 41.7 pmol/g, left 27,7pmol/g and right 25.3 pmoi/g, control and schizophrenic, respectively). There is complete preservation of N-methy-& aspartate sites in schizophrenic hippocampi, and a marginally sign&ant loss of the q~~u~a~ binding site in CA&A, regions (left 249 fmolfg and right 306 fmol/g, left 157 fmol/g and right 148 fmol/g, control and schizophrenic, respectively). These findings reflect the possible importance of glutamate in the ~~ophysiology of sc~zophren~a and represent novel targets for thera~~tic manipulation in ~hi~pbrenia.
A resurgence of interest in the neuropathology of psychiatric disorder has led to a surprising finding in schizophrenia research. The dopamine rich fimbic areas of traditional interest in ~h~ophrenia are free of ~tholo~~~ abno~ality2 whereas wjdes~r~d abnormality can be found in a variety of temporal lobe structures, notably the hippocampal formation (for general review see Ref. 26. For instance, Bogerts ea ~1.~have shown gross shrinkage in the ~p~carn~~ fo~ation and Brown et al.’ have demonstrated thinning of the parahippocampal gyrus.5*7Other studies have pointed to pyramidal cell loss in the cornu ammonjs of the ~p~arnpus’~*’ and some authors suggest that the degree of cellular disarray in this region is proportional to the severity of psychosis.1 Pharmacotherapeutic advances in schizophrenia are hampered by the inability to demonstrate a consistent patholo~cal role for any transmitter or ne~omodulator other than dopamine.‘O The hippocampus is reIatively devoid of dopamine and anatomical connections to dopam~ne systems.21 At best post corded autora~ography indicates a Dz receptor density in the CA1 and CA3 region of the hippocampus, some 20% of those found in striatum and nucleus accumhens.%* Unravel&g the neurochemical correlates of
hippocampal neuropatbolo~ may indicate other neurochemical substrates of schizophrenic amenable to pharmacological manipulation. One such candidate is the excitatory ne~otrans~tter L-~u~rnate which provides the major afferent input to the hippocampus, .I3 it is the neurotransmitter of many hippocampal intemeuronal circuitsj5 and the hippocampus has the highest content of glutamate in the mammalian CNS.35 Glutamate acts on several types of receptor according to their affinity for exogenous ligands, namely N-methyl-n-aspartate (the NMDA receptor), kainic acid [the kainate (KA) receptor] and q~squalic acid (the quisqualate ~ptor).~ We have previously shown a possible reduction in the number of KA receptors in the left hippocampus in schizophrenia in synaptic membranes from homogenized tissue. 22We have now performed a quantitative autoradio~aphic study of the anatomical distribution of glutamate binding sites in the human ~pp~rn~ formation in control and schizophrenic patients. Such an approach as well as quantif~ng receptor changes allows more precise ~croanatomical localization of receptor abnormalities in brain disorder, hitherto not possible with homogenate studies.
*To whom correspondence should be addressed. Abbreviation: CA,_+cornu ammo& (regions l-4); CNQX,
EXPERIMENTAL PROCEINRES Parims
~y~o”?-nitr~~noxa~n~Z,3~ione; EEct, ekxtroe~~phalo~~; KA, kainate; NMDA, N-methyl-naspartate; RDC, Research Diagnostic Criteria.
and methats
For details of patients, age, sex, posr ~orrern delay and
causes of deaths see Table 1. 2s
26
R.
KERWIN et al.
Table I. Patient characteristics Post mortem delay
Age
Sex
(h)
Cause of death
39
Male
12
40 65
Male Female Female
18 36 48
71
Female
70
75 83 85
Male Male Male
24 25 23
Sudden death (myocardial infarction) Myocardial infarction Mvocardial infarction Sudden death (cause uncertain) Post-ooerative acute renal f’ailure following bladder stone removal Hypertensive heart failure Acute pulmonary oedema Bronchopneumonia
Male
49
32
Male
37
52 56 63 74 78
Male Female Female Female Male
32 30 70 52 37
Control
63
Schizophrenic 21
Suicide (self-immolation) Suicide (jumped from altitude) Myocardial infarction Carcinoma of the bronchus Diabetic ketoacidosis Meningitis Disseminated carcinoma
All controls and the 21-year-old male schizophrenic had never received any psychotropic medication. All other patients had received either chlorpralmarine, trifluoperazine, haloperidol or thioridazine.
In the schizophrenic cases diagnoses were made according to Research Diagnostic Criteria (RDC).” Control cases were free of any neuropsychiatric disorder or psychotropic drug intake. Temporal lobes in the hippocampal region at the anterior thalamic level were cut in coronal sections at 5-mm intervals on a domestic meat slicer and stored at - 70°C until further sectioning. Cryostat sections were cut at 2Oqm intervals at - 18°C and thaw-mounted onto gelatin chromealum coated microscope slides and further stored at -70°C. Materials ~[~H]Glutamate ([-‘Hle;lutamate, specific activity 50 Ci/ mmol) was obtained from Amenham (Amersham, U.K.). [‘H]KA ([‘H]KA, specific activity 44Ci/mmol) was a gift from Dr J. Collins. Citv Po1vtechnic. London. I’Hl6-Cvano7nitroquinoxaline~2,3~dione ([)HaQX, sp&i& activity 5.47 Ci/mmol) was a gift from Ferrosan Research Division, Denmark. All other chemical reagents were purchased from Sigma (Poole, U.K.). Tritium sensitive LKB ultratilm was purchased from Pharmacia.
Autoradiography
Modifications of the method of Greenamyre et a1.l6 were used for glutamate receptor autoradiography. To remove endogenous glutamate all sections were preincubated in appropriate Ca*+-free. (to prevent uptake) assay buffer (see later for details) at 4°C for 15min, followed by 30°C for 15 min. In all cases three sections/case _ per ligand where studied. For [‘H]glutamate autoradiography, sections were incubated at 2°C for 45 min with 50 mM T&-acetate bulTer (pH 7.4) containing 100 nM [‘HIglutamate. Non-specific binding was determined by incubating adjacent sections as above in the presence of 1 mM glutamate. Reactions were terminated by 2 x 4-s washes in ice-cold assay buffer followed by a 2-s
dip in distilled water and a 2-s dip in a 2.5% solution of glutaraldehyde in acetone. For rwKA binding incubations were performed in 0.31 M T&citrate @H 7.2) containing 40 nM rI=IlKA for 2 h at 2°C. Non-specific binding was determined in adjacent sections by the presence of 1 mM glutamate. Reactions were terminated by 2 x I-min washes in ice-cold assay buffer followed by a dip in distilled water. To determine NMDA binding, incubations were performed in 50mM Tris-acetate buffer (pH 7.4) containing 100 nM 13Hlplutamate. 10 nM KA and 2.5 uM auisaualic acid for 45 r&r at Z”C.‘Nod-specific binding &as d&r&ined in adjacent sections by adding 1 mM cold NMDA to the incubation. Reactions were terminated as for [3H]glutamate autoradiography. For q&q&& binding site autoradiography, tissue was incubated with 50 mM Tris-HCl IpH 7.2) containinn 50 nM [smNQX for 45 min at 2°C. Re&ions were tern&ted by three rapid dips in ice-cold buffer only. Non-specific binding in adjacent sections was determined in the presence of 1 mM glutamate. In all cases sections were dried for 1 min on a hotplate and further dried for 10min in a stream of warm dry air. Dried sections were apposed against tritium-sensitive photographic film (LKB ultrafilm) placed in X-ray cassettes and exposed along with calibrated microscales (Amersham, U.K.) in the dark for 28 days and developed in Kodak D19 developer, fixed, rinsed and dried. Quantification and image analysis was performed on an IBAS II (Kontron-Zeiss) image analysis system which compares film densities of the tissue sections with those of a standard curve produced from polynomial regression analysis of calibrated radioactive standards. Data is presented as pmol/g or fmol/g tissue. Black and white photographs were prepared from digitized images produced by the image analysis system. Comparisons were performed using Tukey’s multivariate analysis of variance.
27
Glutamate and schizophrenia Table 2. Specific [3H&@amate binding to various regions from control and schizophrenic hippoearnpi
Dentate Area Control Right Left Schizophmnic Right Left
CA,
CA,
112*34 131 f 64.2
70&27 49.7 p 10.8
94 + 5.2 89.1 f 32
106f 19 43.6 j, 12.3
75.9 f 19 69.9 f 12.7
105 122 93.6 f 23
CA,
parahippocampal gyrus
106f37 136&45
104*37 119If136
78 f 37.7 109.2 f 47
169 f 30.4 132 f 19.6
181 rf: 24 102f31
124.6 f 30.7 149 f 23.6
CA,
Byrus
Results am expressed as pmol/g tissue. Each point is the mean j, S.E.M. from eight controls and seven schizophrenic patients.
RJBULTS
Total [3HJg&znrate
b&ding
A
representative auto~dio~ph for control is shown in Fig. 2A. Non-specific binding was approximately 30% of total. Highest levels of binding were found in the CA&A, region and throughout the dentate gyrus with the molecular layer being indistinguishable from the hilus. There were no significant differences between hippocampal areas in any of the tissue groups. Details of these results are shown in Table 2. material
[‘HJKainate binding Black and white representative autoradiographs (taken bilaterally) for control and schizophrenic material are shown in Fig. 1. Non-specific binding was approximately lo-20% of total. The binding site density was greatest in the CA, and CA, regions with intermediate levels in the parahippocampal gyrus.
There was a significant reduction of KA receptors in the mossy fibre termination, CA&A, region in schiiophrenic tissue. Significant losses could also be detected bilate~ly in the dentate gyrus and cortex and on the left in CAr and CA, in schizophrenic tissue. These results are detailed in Table 3. N-Methyl-o-aspartate displaceable [‘HIglutamate binding in the presence of 10 PM kainate and 2.5 FM quilcqualate (the N-methyl-D-aspartate receptor) A representative autoradiograph is shown in Fig. 2B. Binding to the NMDA receptor was highly specific with barely detectable (~5%) non-specific binding. Binding sites were particularly enriched in a single layer of cells (probably molecular layer) in the dentate gyrus and CA&A, region, with intermediate levels in the parahippocampal gyrus and low levels in the C&/CA, region. There were no significant differences between schizophrenic and control tissue. These results are shown in Table 4.
Table 3. Specific rH]kainate binding to various regions from control and schizophrenic hippocampi Area Control Right Left
~~p~nic Right Left
Dentate Byrus
CA.
CA,
CA,
CA,
Parahippocampal Byrus
28 f 4.0 14.2 f 2.7
57.2 rfr7.6 54.2 f 12.7
88 f 6.7 66.6 f 13.7
33k6.1 26.4 f 5.6
27.9 f 5.5 23.9 f 8.6
41.7 f 6.4 50.8 2 9.7
29 + 8.2* 18.3 f 2.1*
27.7 f 4.0* 17.9 * 4.3*
38.4 f 1.25 10.9 * 3.0+
25.3 f 11.0 9.0 f 1.2*+
25.3 & 3.1** 27.7 -+ 7.9**
7.9 f 2.9* 9.5 f 1.1+
Results am expressed as pmol/g tissue. Each point is the mean f S.E.M. from eight controls and seven schizophrenic patients. *Main effect of diagnosis at P < 0.01 or **P < 0.05 using MANOVA, i.e. significantly different from respective control area. Table 4. Specific ~-methyl-D-aviate
Area Control Right Left Schizophrenic Right Left
binding (~-me~yl-amputate sensitive ~3H]glutamate) to various regions from control and ~~ophren~ hip~mpi
Dentate gYrus
CA,
CA,
CA,
CA,
48.8 f 20 62k 11.4
33 f 16 37 If: 9.3
33+11 38.4 _+12.6
78 f 1.2 76.8 f 6.1
93.3 f 5.6 88 rf: 8.0
71 + 7.7 58.8 f 17.8
58.4 f 12.5 67 + 4.4
24.0 f 6.6 25.1 + 6.7
22 & 8.2 15.6 f 2.8
74.9 f 19.3 88 f 26.1
78.0 f 16.8 90.3 If: 10.6
61 f 26.6 45.4& 11.7
Parahippocampal gYrus
Results are expressed as pmolig tissue. Each point is the mean f S.E.M. from eight controls and seven schizophrenic patients.
R. ORWIN et ai,
Fig. 1. Rigitized reversed black and white autoradiographs of [“)I]&% binding in hippocampal sections from control right (A); control left (B); schizophrenic right (C) and schizophrenic Ieft (D). Due to constraints of film speed required for the dirtied data on this system, quantitative changes cannot be reliably inferred and figures are to illustrate mo~holo~ca1 localization (see tables for grey scale quantification data). White represents high density of binding. Ca,-Ca, cornu ammonis areas; SUB, subiculum; PHG, parahippocampal gyrus; DG, dentate gyrus.
[3H]6-Cyano-7-nitroquinuxalone-2,3-dione
binding
A representative autoradiograph is shown in Fig. 2C. Non-specific binding was approximaiely 3040% of control. The dist~bution of [3HtCbIQX was similar to that for NMDA displaceable glutamate binding. However, there was a just significant loss of receptors in the CA, region bilaterally and on the CA3 region on the ieft. These results are shown in Table 5.
DISCUSSION
Localization of glutamate receptors in human hippocampus In view of the paucity of ~h~zophrenic material it is not possible to undertake full saturation analysis to accurately determine Ko and B,,. In limited saturation studies with KA and glutamate in control material, saturable binding above 4OnM was
Glutamate and schizophrenia
29
Fig. 2. Digitized reversed black and white representative autoradiographs of (A) glutamate binding sites (B) putative NMDA binding sites and (C) quisqualate binding sites. See legend to Fig. 1 for further details.
observed for KA and above 100 nM for ghttamate. Half-maximal binding occurred at about 10 nM for [3HJKA and 60 nM for [3H~Iutamate. Because of the small amounts of materid, it was not possible to test whether the alterations seen reflect primary changes in KD or B,_. Although at saturation such changes are likely to be B_. Animal studies (S. Patek unpublished observations) have shown that chronic neuroleptic treatment in rats does not affect the kinetic parameters of binding for K.A. Thus the main value of this report is that it provides an anatomic localization for the abno~ality previously reported in
homogenates. Studies on the hippocampal distribution of gtutamate receptors in animals show localization of the KA receptor to the mossy fibre te~ina~ CA&A, region9 and NMDA to the CA1/CA2 region and dentate gyrus.’ More recxmtly, the quisqualate receptor has been shown to colocalize with the NMDA receptor mainly in the CA&A, region.32 PH]CNQX is a novel ligand that binds spe&cally to the putative quisqualate receptoti3 and some workers argue that a low afhnity IL4 receptor, which this hgand also recognizxs, is in fact the quisqu~ate receptor. ” Individual studies confirming
30
R.
KERWlN et d
Table 5. Specific [3H)dcyano-7-nj~r~uinoxaionc-Z,3~one and ~~p~~ Area Control Right Left Schizophrenic Right Left
(q~~~late
receptor) binding to various regions from control
hippeeampi
Dentate SWs
CA,
CA3
CA,
CA,
Parahippocampal Wus
3&t&52 288&23
262 & 89 249 f 35
254f22 306 i 92
417*66 492 + 101
485f63 403&41
293 4 53 337 f 85
398 & 78 269 f 97
174* 46* 207 4 48 467 & 247 419&66 157422* 148jI 24* 408f130 3OS&# Results are expressed as fmolia tissue. *P < 0.05 main e&ct of diagnosis (MANOVA), i.e. si~i~~~y respective Eontrol area. -
the anatomical localization of the kainate and NMDA receptors in human material have appeared.19cs However, the present study represents the first systematic localization of all three glutamate receptor subtypes in the human ~p~mpus. The specificity of the technique to localize KA receptors and NMDA displaceable [3H)glutamate binding to localize NMDA receptors in rats is now well established.3’ Although CNQX is a novel ligand for the quisqualate receptor and el~~ophysiolo~~ studies suggest high specificity of this ligand,” others have claimed that it may also bind to both KA and NMDA receptors also, 1753but these effects occur only in the micromolar range. The tissue was matched for age, sex and post mortem delay, as much as is possible with such scarce material. The multifactorial analyses of variances performed showed no significant interaction with these variables. In general post mortem delays of up to 48 h are aeceptable4’ although transient increases are noted at 24 h. Three patients in the schizophrenic group had post mortem delays of > 48 h. However, the dire&ion of change in the schizophrenic group is in the opposite direction to that produced by prolonged delay.‘” As for agonal status, we have no information on the duration of coma (if any) in these patients, Glutamate receptors and schizophrenia
The major finding of this study is the greater loss of IL4 binding sites in the CA&A, region, and dentate gurus and entorhinal cortex in schizophrenic tissue. There is already evidence of several kinds for a glutamate abnormality in schizophrenia. A decrease in cerebrospinal fluid glutamate levels in schizophrenic patients has been found.z3 Abno~alities regarding excitatory neurotrans~tt~r function have been reported in discrete brain regions: Nishikawa et al.” have shown an increase in KA receptors in frontal cortex (eye movement areas) and Deakin et at.” have shown abnormalities in f’H]o-aspartate binding (presynaptic sites) in left polar temporal cortex in schizophrenic tissue. The present study also describes glutamate receptor abnormalities, and localizes the abnormality to a particular group of pyramidal ceils in the hippocampus. It is not likely that these changes simply reelect generalized non-specific cell loss as NMDA receptors,
329 f 87 305 f 82.5 dit%rent from
and the majority of the quisqualate receptors are preserved in areas of kainate receptor loss. Such abnormalities have led some:’ including ourselves,M to propose a glutamate hypothesis for ~~op~e~a. Indeed, it can be argued from the reciprocal relationship between glutamate and dopamine in controlling each others’ release, that a primary abnormality in glutamate neurotransmission can account for dopamine overactivity.W The NMDA receptor has recently been a focus of interest in schizophrenic research as agents which bind to the complex phencyclidine/sigma sites may be psychoto~me~c (for a review see Ref. 19). In addition all traditional antipsychotics displace sigma binding with high potency.& However, our study does not demonstrate any abnormalities at least in the agonist (glutamate) recognition site of the NMDA receptor. Lateral@
and schizophrenia
Several studies support the notion of schizophrenia as an abno~~ty of the left temporal 10be.~ Thus electroencephalogram (EEG) abnormalities are principally left-sided. I4 Left cerebral hypodensity is a common CT scan appearancej7 and there is an increase in left amygdala dopamine levels” in schizophrenia as well as a reduction in [3H~-asp~~ binding in left amygdala.” In addition, some authors also claim that the neuropathological abnormalities are localized to the left,l*R althou~ recent magnetic resonance imaging studies do not support this.43 Consistent with our previous study,” KA binding was persistently lower throughout the left in and t-testing shows significant schizophrenics, left-right differences; multiv~ate analysis of variance, however, did not show a significant side by diagnosis interaction. Therefore, this study cannot then reliably contribute to this debate. The hippocampus and schizophrenia
The precise roles of the hippocampus, in general, and the KA receptor, in particular, in schizophrenia are unknown. It has long been known that hippo~ampal pathology such as temporal lobe epilepsy,” hippocampal infarction” and temporal lobectomyW4’ is associated with schizophreniform states. Indeed the mossy fibre pathway and the KA receptors in its
Glutamate and schizophrenia CA&A, terminal have been the. focus of intense study in pathogenetic mechanisms in temporal lobe epilepsy (for a review see Ref. 3). There are several other plausible explanations for hippocampal pathology and schizophrenia, all of them purely theoretical.*’ In this regard, of particular note as far as the IL4 receptor is concerned, is that these receptors develop late 4*36during early post-natal development at the same time as a proposed neurodevelopmental anomaly in hip~~pi in ~hizophrenics may also be occurring.27~39 CONCLUSION The above relationships all remain theoretical. Of tangible and important significance is that a glutamate hypothesis for schizophrenia opens up an avenue for the design of novel therapeutic agents, acting on the glutamate receptor system. Other lines of research have shown a neuroprotective effect against ischaemic damage29 by a variety of pharmacological agents acting at the glutamate receptor. Some of these
31
agents, currently in development for limitation of cerebral infar~ons in stroke and cardiac arrest,29 may be suitable for clinical use in schizophrenia. It should be noted that although a broad base of glutamate receptor pharmacology has been studied, from a cli~~pa~olo~~ standpoint due to the paucity in this type of material, the study is constrained by the relatively imprecise matching of age, sex, post mortem delay, agonal status and anatomical level. However, althou~ the numbers are small by standards of membrane binding studies, this study is the largest receptor autoradiographic series for schizophrenia to date. A long-term replication study with larger numbers of closely matched prospectively studied patients is underway. work was supported by the MRC and the Psvchiatrv Research Trust. We thank the Wellcome Trust for &anciai support for S. Patel. The study could not have been completed without the invaluable gift of brain material from Rosemary Brown and Dr Clive Bruton of
Acknowledgements-This
Runwell Hospital, Essex and Dr Gavin Reynolds, University of Nottingham.
REFERENCES
~~~~1 cell o~entation in ~~ophrenia. Arch. 1, Altschuler L., Kovelman A. and Scheibel A. (1987) Hip~mpal Gen. Psych&. 44, 1094-1098. 2. Bogerts B., Mema E. and Schonfeid-Bausch R. (1985) Basal ganglia and limbic system pathology in schizophrenia. A morphometric study of brain volume and shrinkage. Arclr. Gen. Psych&. 42, 784-791. 3. Ben Ati Y. (1985) Limbic seizures and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy. Neuroscience 14, 375-403. 4. Ben Ari Y., Tremblay E., Berger M. and Nitecka L. (1984) Kainic acid seizure syndrome and binding sites in developing rats. &VI Brain Res. 14,284-288. 5. Brown R., Colter N., Corsellis J. A. N., Crow T. J., Frith C. D., Jagoe R., Johnstone E. C. and Marsh L. (1988) Postmortem evidence of structural brain changes in schizophrenia: differences in brain weight; temporal horn area and ~a~p~~‘~~ compared with affective disorder, Arch. Gen. Psych&. 43, 36-42. 6. Camps M., Cortez R., Gueye B., Probst A. and Palacios J. M. (1989) Dopamine receptors in human brain: autoradiographic distribution of D* sites. Neuroscience 2t3, 274-296. I. Colter N., Battal S., Crow T. J. et al. (1987) White matter reduction in the parahippocampal gyrus of patients with schixophrenia. Arch. Gen. Psychiat. 44, 1023. 8. Cortes R., Gueye B., Paxos A., Probst A. and Palacios J. (1989) Dopamine receptors in human brain: autoradiographic distribution of D, sites. Neuroscience 28, 263-273. J. S. (1987) Anatomical organization of excitatory 9. Cotman C. W., Monaghan D. T., Ottersen 0. P. and Stow-Ma~isen ammo acid receptors and their pathways. Trenab Neurosci. 10, 273-280. 10. Crow T. J. (1987) The dopamine hypothesis survives: but there must be a way ahead. Br. J. Psyctdat. lSf,460-461. 11. De&in J. F. W., Slater P., Simpson M. D. C., Gilchrist A. C., Skan W. J., Royston M. C., Reynolds G. P. and Cross A. J. (1989) Frontal cortical and left temporal glutamatergic dysfunction in ~~zoph~nia. J. Neurochem. 52,1781-1786. 12. Falkai P. and Bogerts 8. (1986) Cell loss in the hippocampus of schizophrenics. Eur. Arch. Psychiat. Neurot. Sci. 236, 1.54-161. 13. Fonnum F. (1984) Glutamate: a neurotransmitter in mammalian brain. 1. Neurochem. 42, l-l 1. 14. Flor-Henry P. (1983) Commentary and synthesis. In Luterality and Psychopathology (eds Flor-Henry P. and Grureher J.), pp. l-18. Elsevier, Amsterdam. 15. Ror-Henry P. (1969) Psychosis and epilepsy a controlled investigation. E~ile~~ 10, 363-396. 16. Geenamyre J. T., Young A. B. and Penney J. B. (1984) Quantitative autoradiographic distribution of L-)H glutamate binding sites in rat central nervous system. J. Neurosci. 4, 2133-2144. 17. Honore T., Davies S. N., Drejer J. et 4i. (1988) Quinoxalinediones: potent competitive non NMDA glutamate receptor antagonists. Science 241, 701-703. 18. Jakob H. and Beckman H. (1986) Prenatal developmental disturbances in the limbic allocortex in schizophrenics, J. Neural Transm. 65, 303-326. 19. Kemp J., Foster A. C. and Wong E. H. F. (1987) Noncompetitive antagonists of excitatory amino acid receptors, Trenak Neurosci. IO, 294-298. 20. Ketwin R. W. (1988) Letter: pathology, phenomenology and the dopamine hypothesis of schizophrenia. or. J. Psychiat. 152, 141. 21. Kerwin R. W. (1989) How do the neuropathological changes of schixophrenia relate to pm-existing neurotransmitter and aetiological hypotheses? Psychal. Med. 19, 563-567. 22. Kerwin R. W., Pate1S., Meldrum B. et al. (1988) Asymmetrical loss of a glutamate receptor subtype in left hippocampus in ~~~rnor~ern~~zophrenic brain. Lancet 583-584. NSC 39/b-s
32
R. KERWIN
et ai.
23. Kim J. S., Komhuber H., S&mid-Burgk W. and Holzmuiier B. (1980) Low cerebrospinal fluid glutamate in ~h~zophren~c patients and a new h~~esis on schizophrenia. Newosci. Lett. 20, 379-382. 24. Komhuber J. and Fischer E. G. (1982) Glutamic acid diethyl ester induces catalepsy in rats: a new mode1 for schizophrenia? Neurosci. Left. 24, 93-96. 25. Kovelman J. A. and Scheibel A. B. (1984) A neurohistological correlate of schizophrenia. Biof. Psychiat. 19, 160-1621. 26. Lantos P. L. (1987) The neuropathology of schizophrenia: a critical review of recent work. In Schizophrenia: The Major Issues (eds McGuthn P. and Bebbington P.). Mental Health Foundation, London. 27. Lewis S. W. and Murray R. M. (1987) Is schizophrenia a neurodevelopmental disorder? Br. med. J. 295, 681682. 28. Mednick S. A. (1970) Breakdown in individuals at high risk for schizophrenia: possible predispositional perinatal factors. J. Ment. Hyg. 54, 5063. 29. Meldrum B. (1985) Possible therapeutic applications of antagonists of excitatory amino acid neurotransmitters. Clin. S&i. 68, 113-122. 30. Meldrum B. S. and Kerwin R. W. (1987) Glutamate receptors and ~~ophrenia. J. Psychopkarmac. 1, 217-221. 31. Monaghan J. B. and Michel J. (1987) Identification and characte~~tion of an ~-methyl-~as~rtate specific L3~]-glutamate recognition site in synaptic plasma membranes. J. Neurochem. 48, 1699-1708. _ 32. Monarrhan D. T.. Yao D. and Cotman C. W. (1984) Distribution of ‘H-AMPA binding sites in rat brain as determined by qu&titative autoradiography. Brain Res. j24, iSO-164. 33. Neuman R. S., Ben Ari Y., Gho M. and Cherubini E. (1988) Blockade of excitatory synaptic transmission by 6-cyano-7-nitroquinoxaline-2-3,dione (CNQX) in the hippocampus in vitro. Neurosci. Left. 92, 64-68. 34. Nishikawa T., Takashima M. and Toru M. (1983) Increased ‘H-kainic acid binding in the prefrontal cortex in schizophrenia. Neurosci. L&t. 49, 245-250. 35. Otterson 0. P. and Storm-Math&n J. (1984) Neurons containing or accumulating transmitter amino acids. In Hark of ~hern~ai Nearo~atomy (eds Bjorkl~d A., Hokfelt T, and Kubar M. J.), Vol. 3. Classical Tr~sm~tters and Tr~mitter Receptors in the CNS, part II, pp. 141-245. Elsevier, Amsterdam. 36. Represa A., Trembley E. and Ben Ari Y. (1987) Kainate binding sites in the hippocampal mossy fibres: localization and plasticity. Neuroscience 20, 739-748. 37. Reveley M., Revely A. and Baldy R. (1987) Left cerebral hemisphere hypodensity in discordant schizophrenic twins. A controlled study. Arch. Gen. Psychiat. 44, 625632. 38. Reynolds G. P. (1983) Increased concentrations and lateral asymmetry of amygdala dopamine in schizophrenia. Nature 305, 527-529. 39. Roberts G. W., Colter N., Lofthouse R. et al. (1987) Is there gliosis in schizophrenia? Investigation in the temporal lobe. Biol. Psych&t. 22, 1459-1468. 40. Roberts G. W., Done D. J., Bruton C. and Crow T. J. (1990) A ‘mock up’ of schizophrenia: temporal lobe epilepsy and ~hizophrenia like psychoses. Biol. Psychjat. (in press). 41. Slater E. and Beard A. W. (1963) The schizophrenia like psychoses of epilepsy. &. J. Psych&. 109, 95-150. 42. Spitzer R. L., Endicott J. J. and Robins E. (1975) Research Diagnostic Criteria. Biometrics Research, New York State
Department of Mental Hygiene. 43. Suddath R. L., Casanova M. F., Goldberg T. E., Daniel D. G., Kelsoe J. R. and Weinberger D. R. (1989) Temporal lobe pathology in schizophrenia: a quantitative magnetic resonance imaging study. Am. J. Psychiat. 146,464-472. 44. Tam S. W. and Cook L. (1984) Sigma opiates and certain antipsychotic drugs mutually inhibit (+)-‘H-haloperidol binding in guinea pig brain membranes. Proc. natn. Acad. Sci. 81, 5618-523. 45. Tremblay E., R$esa A. and Ben Ari Y. (1985) Autoradiographic localization of kainic acid binding sites in the human hippocampus. Bruin Res. 343, 378-382. 46. Watkins J. C. and Evans R. H. (1981) Excitatory amino acid transmitter. A. ROD.Pkar~c. Toxicol. 21, 165-204. (Accepted 30 May 1990)
NeuroscienceVol. 39, NO. i, pp. 25-32, 1990 Print& in Great Britain
QUANTITATIVE AUTO~DIOGRAPHIC ANALYSIS OF GLUTAMATE BINDING SITES IN THE HIPPOCAMPAL FORMATION IN NORMAL AND SCHIZOPHRENIC BRAIN POS?’ ~~~T~~ R. Institute
KERWIN,*
S. PATEL and B.
MELDRUM
of Psychiatry, Departments of Neuroscience, Neurology and Psychiatry, De Crespigny Park, Denmark Hill, London SE5 SAF, U.K.
Abstract-Using quantitative autoradiography, the anatomical distribution of the binding sites (kainate, N-methyl-o-aspartate and q~~ualate) for the excitatory neurotransmitter @tamate has been established in the hjp~rnp~ fo~ation from control and ~~~phrenic brains, post martem.There is a loss of the kainate subtype particularly in ~~zoph~n~c ~pp~rnpi mainly from the CA&A, mossy fibre termination zone of the cornn ammonis (CA., and CA3; control and schizophrenic left hippocampus, respectively, 54.2 and 66.6 pmol/g; 18.3 and 17.9 pmol/g), as well as bilateral losses in the dentate gyrus (left 14.2 pmol/g and right 28.0 pmoljg; leR 9.5 pmol/g and right 7.9 pmol/g, control and schizophrenic, respectively) and parabippocampal gyrus (left 50.8pmoI/g and right 41.7 pmol/g, left 27,7pmol/g and right 25.3 pmoi/g, control and schizophrenic, respectively). There is complete preservation of N-methy-& aspartate sites in schizophrenic hippocampi, and a marginally sign&ant loss of the q~~u~a~ binding site in CA&A, regions (left 249 fmolfg and right 306 fmol/g, left 157 fmol/g and right 148 fmol/g, control and schizophrenic, respectively). These findings reflect the possible importance of glutamate in the ~~ophysiology of sc~zophren~a and represent novel targets for thera~~tic manipulation in ~hi~pbrenia.
A resurgence of interest in the neuropathology of psychiatric disorder has led to a surprising finding in schizophrenia research. The dopamine rich fimbic areas of traditional interest in ~h~ophrenia are free of ~tholo~~~ abno~ality2 whereas wjdes~r~d abnormality can be found in a variety of temporal lobe structures, notably the hippocampal formation (for general review see Ref. 26. For instance, Bogerts ea ~1.~have shown gross shrinkage in the ~p~carn~~ fo~ation and Brown et al.’ have demonstrated thinning of the parahippocampal gyrus.5*7Other studies have pointed to pyramidal cell loss in the cornu ammonjs of the ~p~arnpus’~*’ and some authors suggest that the degree of cellular disarray in this region is proportional to the severity of psychosis.1 Pharmacotherapeutic advances in schizophrenia are hampered by the inability to demonstrate a consistent patholo~cal role for any transmitter or ne~omodulator other than dopamine.‘O The hippocampus is reIatively devoid of dopamine and anatomical connections to dopam~ne systems.21 At best post corded autora~ography indicates a Dz receptor density in the CA1 and CA3 region of the hippocampus, some 20% of those found in striatum and nucleus accumhens.%* Unravel&g the neurochemical correlates of
hippocampal neuropatbolo~ may indicate other neurochemical substrates of schizophrenic amenable to pharmacological manipulation. One such candidate is the excitatory ne~otrans~tter L-~u~rnate which provides the major afferent input to the hippocampus, .I3 it is the neurotransmitter of many hippocampal intemeuronal circuitsj5 and the hippocampus has the highest content of glutamate in the mammalian CNS.35 Glutamate acts on several types of receptor according to their affinity for exogenous ligands, namely N-methyl-n-aspartate (the NMDA receptor), kainic acid [the kainate (KA) receptor] and q~squalic acid (the quisqualate ~ptor).~ We have previously shown a possible reduction in the number of KA receptors in the left hippocampus in schizophrenia in synaptic membranes from homogenized tissue. 22We have now performed a quantitative autoradio~aphic study of the anatomical distribution of glutamate binding sites in the human ~pp~rn~ formation in control and schizophrenic patients. Such an approach as well as quantif~ng receptor changes allows more precise ~croanatomical localization of receptor abnormalities in brain disorder, hitherto not possible with homogenate studies.
*To whom correspondence should be addressed. Abbreviation: CA,_+cornu ammo& (regions l-4); CNQX,
EXPERIMENTAL PROCEINRES Parims
~y~o”?-nitr~~noxa~n~Z,3~ione; EEct, ekxtroe~~phalo~~; KA, kainate; NMDA, N-methyl-naspartate; RDC, Research Diagnostic Criteria.
and methats
For details of patients, age, sex, posr ~orrern delay and
causes of deaths see Table 1. 2s
26
R.
KERWIN et al.
Table I. Patient characteristics Post mortem delay
Age
Sex
(h)
Cause of death
39
Male
12
40 65
Male Female Female
18 36 48
71
Female
70
75 83 85
Male Male Male
24 25 23
Sudden death (myocardial infarction) Myocardial infarction Mvocardial infarction Sudden death (cause uncertain) Post-ooerative acute renal f’ailure following bladder stone removal Hypertensive heart failure Acute pulmonary oedema Bronchopneumonia
Male
49
32
Male
37
52 56 63 74 78
Male Female Female Female Male
32 30 70 52 37
Control
63
Schizophrenic 21
Suicide (self-immolation) Suicide (jumped from altitude) Myocardial infarction Carcinoma of the bronchus Diabetic ketoacidosis Meningitis Disseminated carcinoma
All controls and the 21-year-old male schizophrenic had never received any psychotropic medication. All other patients had received either chlorpralmarine, trifluoperazine, haloperidol or thioridazine.
In the schizophrenic cases diagnoses were made according to Research Diagnostic Criteria (RDC).” Control cases were free of any neuropsychiatric disorder or psychotropic drug intake. Temporal lobes in the hippocampal region at the anterior thalamic level were cut in coronal sections at 5-mm intervals on a domestic meat slicer and stored at - 70°C until further sectioning. Cryostat sections were cut at 2Oqm intervals at - 18°C and thaw-mounted onto gelatin chromealum coated microscope slides and further stored at -70°C. Materials ~[~H]Glutamate ([-‘Hle;lutamate, specific activity 50 Ci/ mmol) was obtained from Amenham (Amersham, U.K.). [‘H]KA ([‘H]KA, specific activity 44Ci/mmol) was a gift from Dr J. Collins. Citv Po1vtechnic. London. I’Hl6-Cvano7nitroquinoxaline~2,3~dione ([)HaQX, sp&i& activity 5.47 Ci/mmol) was a gift from Ferrosan Research Division, Denmark. All other chemical reagents were purchased from Sigma (Poole, U.K.). Tritium sensitive LKB ultratilm was purchased from Pharmacia.
Autoradiography
Modifications of the method of Greenamyre et a1.l6 were used for glutamate receptor autoradiography. To remove endogenous glutamate all sections were preincubated in appropriate Ca*+-free. (to prevent uptake) assay buffer (see later for details) at 4°C for 15min, followed by 30°C for 15 min. In all cases three sections/case _ per ligand where studied. For [‘H]glutamate autoradiography, sections were incubated at 2°C for 45 min with 50 mM T&-acetate bulTer (pH 7.4) containing 100 nM [‘HIglutamate. Non-specific binding was determined by incubating adjacent sections as above in the presence of 1 mM glutamate. Reactions were terminated by 2 x 4-s washes in ice-cold assay buffer followed by a 2-s
dip in distilled water and a 2-s dip in a 2.5% solution of glutaraldehyde in acetone. For rwKA binding incubations were performed in 0.31 M T&citrate @H 7.2) containing 40 nM rI=IlKA for 2 h at 2°C. Non-specific binding was determined in adjacent sections by the presence of 1 mM glutamate. Reactions were terminated by 2 x I-min washes in ice-cold assay buffer followed by a dip in distilled water. To determine NMDA binding, incubations were performed in 50mM Tris-acetate buffer (pH 7.4) containing 100 nM 13Hlplutamate. 10 nM KA and 2.5 uM auisaualic acid for 45 r&r at Z”C.‘Nod-specific binding &as d&r&ined in adjacent sections by adding 1 mM cold NMDA to the incubation. Reactions were terminated as for [3H]glutamate autoradiography. For q&q&& binding site autoradiography, tissue was incubated with 50 mM Tris-HCl IpH 7.2) containinn 50 nM [smNQX for 45 min at 2°C. Re&ions were tern&ted by three rapid dips in ice-cold buffer only. Non-specific binding in adjacent sections was determined in the presence of 1 mM glutamate. In all cases sections were dried for 1 min on a hotplate and further dried for 10min in a stream of warm dry air. Dried sections were apposed against tritium-sensitive photographic film (LKB ultrafilm) placed in X-ray cassettes and exposed along with calibrated microscales (Amersham, U.K.) in the dark for 28 days and developed in Kodak D19 developer, fixed, rinsed and dried. Quantification and image analysis was performed on an IBAS II (Kontron-Zeiss) image analysis system which compares film densities of the tissue sections with those of a standard curve produced from polynomial regression analysis of calibrated radioactive standards. Data is presented as pmol/g or fmol/g tissue. Black and white photographs were prepared from digitized images produced by the image analysis system. Comparisons were performed using Tukey’s multivariate analysis of variance.
27
Glutamate and schizophrenia Table 2. Specific [3H&@amate binding to various regions from control and schizophrenic hippoearnpi
Dentate Area Control Right Left Schizophmnic Right Left
CA,
CA,
112*34 131 f 64.2
70&27 49.7 p 10.8
94 + 5.2 89.1 f 32
106f 19 43.6 j, 12.3
75.9 f 19 69.9 f 12.7
105 122 93.6 f 23
CA,
parahippocampal gyrus
106f37 136&45
104*37 119If136
78 f 37.7 109.2 f 47
169 f 30.4 132 f 19.6
181 rf: 24 102f31
124.6 f 30.7 149 f 23.6
CA,
Byrus
Results am expressed as pmol/g tissue. Each point is the mean j, S.E.M. from eight controls and seven schizophrenic patients.
RJBULTS
Total [3HJg&znrate
b&ding
A
representative auto~dio~ph for control is shown in Fig. 2A. Non-specific binding was approximately 30% of total. Highest levels of binding were found in the CA&A, region and throughout the dentate gyrus with the molecular layer being indistinguishable from the hilus. There were no significant differences between hippocampal areas in any of the tissue groups. Details of these results are shown in Table 2. material
[‘HJKainate binding Black and white representative autoradiographs (taken bilaterally) for control and schizophrenic material are shown in Fig. 1. Non-specific binding was approximately lo-20% of total. The binding site density was greatest in the CA, and CA, regions with intermediate levels in the parahippocampal gyrus.
There was a significant reduction of KA receptors in the mossy fibre termination, CA&A, region in schiiophrenic tissue. Significant losses could also be detected bilate~ly in the dentate gyrus and cortex and on the left in CAr and CA, in schizophrenic tissue. These results are detailed in Table 3. N-Methyl-o-aspartate displaceable [‘HIglutamate binding in the presence of 10 PM kainate and 2.5 FM quilcqualate (the N-methyl-D-aspartate receptor) A representative autoradiograph is shown in Fig. 2B. Binding to the NMDA receptor was highly specific with barely detectable (~5%) non-specific binding. Binding sites were particularly enriched in a single layer of cells (probably molecular layer) in the dentate gyrus and CA&A, region, with intermediate levels in the parahippocampal gyrus and low levels in the C&/CA, region. There were no significant differences between schizophrenic and control tissue. These results are shown in Table 4.
Table 3. Specific rH]kainate binding to various regions from control and schizophrenic hippocampi Area Control Right Left
~~p~nic Right Left
Dentate Byrus
CA.
CA,
CA,
CA,
Parahippocampal Byrus
28 f 4.0 14.2 f 2.7
57.2 rfr7.6 54.2 f 12.7
88 f 6.7 66.6 f 13.7
33k6.1 26.4 f 5.6
27.9 f 5.5 23.9 f 8.6
41.7 f 6.4 50.8 2 9.7
29 + 8.2* 18.3 f 2.1*
27.7 f 4.0* 17.9 * 4.3*
38.4 f 1.25 10.9 * 3.0+
25.3 f 11.0 9.0 f 1.2*+
25.3 & 3.1** 27.7 -+ 7.9**
7.9 f 2.9* 9.5 f 1.1+
Results am expressed as pmol/g tissue. Each point is the mean f S.E.M. from eight controls and seven schizophrenic patients. *Main effect of diagnosis at P < 0.01 or **P < 0.05 using MANOVA, i.e. significantly different from respective control area. Table 4. Specific ~-methyl-D-aviate
Area Control Right Left Schizophrenic Right Left
binding (~-me~yl-amputate sensitive ~3H]glutamate) to various regions from control and ~~ophren~ hip~mpi
Dentate gYrus
CA,
CA,
CA,
CA,
48.8 f 20 62k 11.4
33 f 16 37 If: 9.3
33+11 38.4 _+12.6
78 f 1.2 76.8 f 6.1
93.3 f 5.6 88 rf: 8.0
71 + 7.7 58.8 f 17.8
58.4 f 12.5 67 + 4.4
24.0 f 6.6 25.1 + 6.7
22 & 8.2 15.6 f 2.8
74.9 f 19.3 88 f 26.1
78.0 f 16.8 90.3 If: 10.6
61 f 26.6 45.4& 11.7
Parahippocampal gYrus
Results are expressed as pmolig tissue. Each point is the mean f S.E.M. from eight controls and seven schizophrenic patients.
R. ORWIN et ai,
Fig. 1. Rigitized reversed black and white autoradiographs of [“)I]&% binding in hippocampal sections from control right (A); control left (B); schizophrenic right (C) and schizophrenic Ieft (D). Due to constraints of film speed required for the dirtied data on this system, quantitative changes cannot be reliably inferred and figures are to illustrate mo~holo~ca1 localization (see tables for grey scale quantification data). White represents high density of binding. Ca,-Ca, cornu ammonis areas; SUB, subiculum; PHG, parahippocampal gyrus; DG, dentate gyrus.
[3H]6-Cyano-7-nitroquinuxalone-2,3-dione
binding
A representative autoradiograph is shown in Fig. 2C. Non-specific binding was approximaiely 3040% of control. The dist~bution of [3HtCbIQX was similar to that for NMDA displaceable glutamate binding. However, there was a just significant loss of receptors in the CA, region bilaterally and on the CA3 region on the ieft. These results are shown in Table 5.
DISCUSSION
Localization of glutamate receptors in human hippocampus In view of the paucity of ~h~zophrenic material it is not possible to undertake full saturation analysis to accurately determine Ko and B,,. In limited saturation studies with KA and glutamate in control material, saturable binding above 4OnM was
Glutamate and schizophrenia
29
Fig. 2. Digitized reversed black and white representative autoradiographs of (A) glutamate binding sites (B) putative NMDA binding sites and (C) quisqualate binding sites. See legend to Fig. 1 for further details.
observed for KA and above 100 nM for ghttamate. Half-maximal binding occurred at about 10 nM for [3HJKA and 60 nM for [3H~Iutamate. Because of the small amounts of materid, it was not possible to test whether the alterations seen reflect primary changes in KD or B,_. Although at saturation such changes are likely to be B_. Animal studies (S. Patek unpublished observations) have shown that chronic neuroleptic treatment in rats does not affect the kinetic parameters of binding for K.A. Thus the main value of this report is that it provides an anatomic localization for the abno~ality previously reported in
homogenates. Studies on the hippocampal distribution of gtutamate receptors in animals show localization of the KA receptor to the mossy fibre te~ina~ CA&A, region9 and NMDA to the CA1/CA2 region and dentate gyrus.’ More recxmtly, the quisqualate receptor has been shown to colocalize with the NMDA receptor mainly in the CA&A, region.32 PH]CNQX is a novel ligand that binds spe&cally to the putative quisqualate receptoti3 and some workers argue that a low afhnity IL4 receptor, which this hgand also recognizxs, is in fact the quisqu~ate receptor. ” Individual studies confirming
30
R.
KERWlN et d
Table 5. Specific [3H)dcyano-7-nj~r~uinoxaionc-Z,3~one and ~~p~~ Area Control Right Left Schizophrenic Right Left
(q~~~late
receptor) binding to various regions from control
hippeeampi
Dentate SWs
CA,
CA3
CA,
CA,
Parahippocampal Wus
3&t&52 288&23
262 & 89 249 f 35
254f22 306 i 92
417*66 492 + 101
485f63 403&41
293 4 53 337 f 85
398 & 78 269 f 97
174* 46* 207 4 48 467 & 247 419&66 157422* 148jI 24* 408f130 3OS&# Results are expressed as fmolia tissue. *P < 0.05 main e&ct of diagnosis (MANOVA), i.e. si~i~~~y respective Eontrol area. -
the anatomical localization of the kainate and NMDA receptors in human material have appeared.19cs However, the present study represents the first systematic localization of all three glutamate receptor subtypes in the human ~p~mpus. The specificity of the technique to localize KA receptors and NMDA displaceable [3H)glutamate binding to localize NMDA receptors in rats is now well established.3’ Although CNQX is a novel ligand for the quisqualate receptor and el~~ophysiolo~~ studies suggest high specificity of this ligand,” others have claimed that it may also bind to both KA and NMDA receptors also, 1753but these effects occur only in the micromolar range. The tissue was matched for age, sex and post mortem delay, as much as is possible with such scarce material. The multifactorial analyses of variances performed showed no significant interaction with these variables. In general post mortem delays of up to 48 h are aeceptable4’ although transient increases are noted at 24 h. Three patients in the schizophrenic group had post mortem delays of > 48 h. However, the dire&ion of change in the schizophrenic group is in the opposite direction to that produced by prolonged delay.‘” As for agonal status, we have no information on the duration of coma (if any) in these patients, Glutamate receptors and schizophrenia
The major finding of this study is the greater loss of IL4 binding sites in the CA&A, region, and dentate gurus and entorhinal cortex in schizophrenic tissue. There is already evidence of several kinds for a glutamate abnormality in schizophrenia. A decrease in cerebrospinal fluid glutamate levels in schizophrenic patients has been found.z3 Abno~alities regarding excitatory neurotrans~tt~r function have been reported in discrete brain regions: Nishikawa et al.” have shown an increase in KA receptors in frontal cortex (eye movement areas) and Deakin et at.” have shown abnormalities in f’H]o-aspartate binding (presynaptic sites) in left polar temporal cortex in schizophrenic tissue. The present study also describes glutamate receptor abnormalities, and localizes the abnormality to a particular group of pyramidal ceils in the hippocampus. It is not likely that these changes simply reelect generalized non-specific cell loss as NMDA receptors,
329 f 87 305 f 82.5 dit%rent from
and the majority of the quisqualate receptors are preserved in areas of kainate receptor loss. Such abnormalities have led some:’ including ourselves,M to propose a glutamate hypothesis for ~~op~e~a. Indeed, it can be argued from the reciprocal relationship between glutamate and dopamine in controlling each others’ release, that a primary abnormality in glutamate neurotransmission can account for dopamine overactivity.W The NMDA receptor has recently been a focus of interest in schizophrenic research as agents which bind to the complex phencyclidine/sigma sites may be psychoto~me~c (for a review see Ref. 19). In addition all traditional antipsychotics displace sigma binding with high potency.& However, our study does not demonstrate any abnormalities at least in the agonist (glutamate) recognition site of the NMDA receptor. Lateral@
and schizophrenia
Several studies support the notion of schizophrenia as an abno~~ty of the left temporal 10be.~ Thus electroencephalogram (EEG) abnormalities are principally left-sided. I4 Left cerebral hypodensity is a common CT scan appearancej7 and there is an increase in left amygdala dopamine levels” in schizophrenia as well as a reduction in [3H~-asp~~ binding in left amygdala.” In addition, some authors also claim that the neuropathological abnormalities are localized to the left,l*R althou~ recent magnetic resonance imaging studies do not support this.43 Consistent with our previous study,” KA binding was persistently lower throughout the left in and t-testing shows significant schizophrenics, left-right differences; multiv~ate analysis of variance, however, did not show a significant side by diagnosis interaction. Therefore, this study cannot then reliably contribute to this debate. The hippocampus and schizophrenia
The precise roles of the hippocampus, in general, and the KA receptor, in particular, in schizophrenia are unknown. It has long been known that hippo~ampal pathology such as temporal lobe epilepsy,” hippocampal infarction” and temporal lobectomyW4’ is associated with schizophreniform states. Indeed the mossy fibre pathway and the KA receptors in its
Glutamate and schizophrenia CA&A, terminal have been the. focus of intense study in pathogenetic mechanisms in temporal lobe epilepsy (for a review see Ref. 3). There are several other plausible explanations for hippocampal pathology and schizophrenia, all of them purely theoretical.*’ In this regard, of particular note as far as the IL4 receptor is concerned, is that these receptors develop late 4*36during early post-natal development at the same time as a proposed neurodevelopmental anomaly in hip~~pi in ~hizophrenics may also be occurring.27~39 CONCLUSION The above relationships all remain theoretical. Of tangible and important significance is that a glutamate hypothesis for schizophrenia opens up an avenue for the design of novel therapeutic agents, acting on the glutamate receptor system. Other lines of research have shown a neuroprotective effect against ischaemic damage29 by a variety of pharmacological agents acting at the glutamate receptor. Some of these
31
agents, currently in development for limitation of cerebral infar~ons in stroke and cardiac arrest,29 may be suitable for clinical use in schizophrenia. It should be noted that although a broad base of glutamate receptor pharmacology has been studied, from a cli~~pa~olo~~ standpoint due to the paucity in this type of material, the study is constrained by the relatively imprecise matching of age, sex, post mortem delay, agonal status and anatomical level. However, althou~ the numbers are small by standards of membrane binding studies, this study is the largest receptor autoradiographic series for schizophrenia to date. A long-term replication study with larger numbers of closely matched prospectively studied patients is underway. work was supported by the MRC and the Psvchiatrv Research Trust. We thank the Wellcome Trust for &anciai support for S. Patel. The study could not have been completed without the invaluable gift of brain material from Rosemary Brown and Dr Clive Bruton of
Acknowledgements-This
Runwell Hospital, Essex and Dr Gavin Reynolds, University of Nottingham.
REFERENCES
~~~~1 cell o~entation in ~~ophrenia. Arch. 1, Altschuler L., Kovelman A. and Scheibel A. (1987) Hip~mpal Gen. Psych&. 44, 1094-1098. 2. Bogerts B., Mema E. and Schonfeid-Bausch R. (1985) Basal ganglia and limbic system pathology in schizophrenia. A morphometric study of brain volume and shrinkage. Arclr. Gen. Psych&. 42, 784-791. 3. Ben Ati Y. (1985) Limbic seizures and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy. Neuroscience 14, 375-403. 4. Ben Ari Y., Tremblay E., Berger M. and Nitecka L. (1984) Kainic acid seizure syndrome and binding sites in developing rats. &VI Brain Res. 14,284-288. 5. Brown R., Colter N., Corsellis J. A. N., Crow T. J., Frith C. D., Jagoe R., Johnstone E. C. and Marsh L. (1988) Postmortem evidence of structural brain changes in schizophrenia: differences in brain weight; temporal horn area and ~a~p~~‘~~ compared with affective disorder, Arch. Gen. Psych&. 43, 36-42. 6. Camps M., Cortez R., Gueye B., Probst A. and Palacios J. M. (1989) Dopamine receptors in human brain: autoradiographic distribution of D* sites. Neuroscience 2t3, 274-296. I. Colter N., Battal S., Crow T. J. et al. (1987) White matter reduction in the parahippocampal gyrus of patients with schixophrenia. Arch. Gen. Psychiat. 44, 1023. 8. Cortes R., Gueye B., Paxos A., Probst A. and Palacios J. (1989) Dopamine receptors in human brain: autoradiographic distribution of D, sites. Neuroscience 28, 263-273. J. S. (1987) Anatomical organization of excitatory 9. Cotman C. W., Monaghan D. T., Ottersen 0. P. and Stow-Ma~isen ammo acid receptors and their pathways. Trenab Neurosci. 10, 273-280. 10. Crow T. J. (1987) The dopamine hypothesis survives: but there must be a way ahead. Br. J. Psyctdat. lSf,460-461. 11. De&in J. F. W., Slater P., Simpson M. D. C., Gilchrist A. C., Skan W. J., Royston M. C., Reynolds G. P. and Cross A. J. (1989) Frontal cortical and left temporal glutamatergic dysfunction in ~~zoph~nia. J. Neurochem. 52,1781-1786. 12. Falkai P. and Bogerts 8. (1986) Cell loss in the hippocampus of schizophrenics. Eur. Arch. Psychiat. Neurot. Sci. 236, 1.54-161. 13. Fonnum F. (1984) Glutamate: a neurotransmitter in mammalian brain. 1. Neurochem. 42, l-l 1. 14. Flor-Henry P. (1983) Commentary and synthesis. In Luterality and Psychopathology (eds Flor-Henry P. and Grureher J.), pp. l-18. Elsevier, Amsterdam. 15. Ror-Henry P. (1969) Psychosis and epilepsy a controlled investigation. E~ile~~ 10, 363-396. 16. Geenamyre J. T., Young A. B. and Penney J. B. (1984) Quantitative autoradiographic distribution of L-)H glutamate binding sites in rat central nervous system. J. Neurosci. 4, 2133-2144. 17. Honore T., Davies S. N., Drejer J. et 4i. (1988) Quinoxalinediones: potent competitive non NMDA glutamate receptor antagonists. Science 241, 701-703. 18. Jakob H. and Beckman H. (1986) Prenatal developmental disturbances in the limbic allocortex in schizophrenics, J. Neural Transm. 65, 303-326. 19. Kemp J., Foster A. C. and Wong E. H. F. (1987) Noncompetitive antagonists of excitatory amino acid receptors, Trenak Neurosci. IO, 294-298. 20. Ketwin R. W. (1988) Letter: pathology, phenomenology and the dopamine hypothesis of schizophrenia. or. J. Psychiat. 152, 141. 21. Kerwin R. W. (1989) How do the neuropathological changes of schixophrenia relate to pm-existing neurotransmitter and aetiological hypotheses? Psychal. Med. 19, 563-567. 22. Kerwin R. W., Pate1S., Meldrum B. et al. (1988) Asymmetrical loss of a glutamate receptor subtype in left hippocampus in ~~~rnor~ern~~zophrenic brain. Lancet 583-584. NSC 39/b-s
32
R. KERWIN
et ai.
23. Kim J. S., Komhuber H., S&mid-Burgk W. and Holzmuiier B. (1980) Low cerebrospinal fluid glutamate in ~h~zophren~c patients and a new h~~esis on schizophrenia. Newosci. Lett. 20, 379-382. 24. Komhuber J. and Fischer E. G. (1982) Glutamic acid diethyl ester induces catalepsy in rats: a new mode1 for schizophrenia? Neurosci. Left. 24, 93-96. 25. Kovelman J. A. and Scheibel A. B. (1984) A neurohistological correlate of schizophrenia. Biof. Psychiat. 19, 160-1621. 26. Lantos P. L. (1987) The neuropathology of schizophrenia: a critical review of recent work. In Schizophrenia: The Major Issues (eds McGuthn P. and Bebbington P.). Mental Health Foundation, London. 27. Lewis S. W. and Murray R. M. (1987) Is schizophrenia a neurodevelopmental disorder? Br. med. J. 295, 681682. 28. Mednick S. A. (1970) Breakdown in individuals at high risk for schizophrenia: possible predispositional perinatal factors. J. Ment. Hyg. 54, 5063. 29. Meldrum B. (1985) Possible therapeutic applications of antagonists of excitatory amino acid neurotransmitters. Clin. S&i. 68, 113-122. 30. Meldrum B. S. and Kerwin R. W. (1987) Glutamate receptors and ~~ophrenia. J. Psychopkarmac. 1, 217-221. 31. Monaghan J. B. and Michel J. (1987) Identification and characte~~tion of an ~-methyl-~as~rtate specific L3~]-glutamate recognition site in synaptic plasma membranes. J. Neurochem. 48, 1699-1708. _ 32. Monarrhan D. T.. Yao D. and Cotman C. W. (1984) Distribution of ‘H-AMPA binding sites in rat brain as determined by qu&titative autoradiography. Brain Res. j24, iSO-164. 33. Neuman R. S., Ben Ari Y., Gho M. and Cherubini E. (1988) Blockade of excitatory synaptic transmission by 6-cyano-7-nitroquinoxaline-2-3,dione (CNQX) in the hippocampus in vitro. Neurosci. Left. 92, 64-68. 34. Nishikawa T., Takashima M. and Toru M. (1983) Increased ‘H-kainic acid binding in the prefrontal cortex in schizophrenia. Neurosci. L&t. 49, 245-250. 35. Otterson 0. P. and Storm-Math&n J. (1984) Neurons containing or accumulating transmitter amino acids. In Hark of ~hern~ai Nearo~atomy (eds Bjorkl~d A., Hokfelt T, and Kubar M. J.), Vol. 3. Classical Tr~sm~tters and Tr~mitter Receptors in the CNS, part II, pp. 141-245. Elsevier, Amsterdam. 36. Represa A., Trembley E. and Ben Ari Y. (1987) Kainate binding sites in the hippocampal mossy fibres: localization and plasticity. Neuroscience 20, 739-748. 37. Reveley M., Revely A. and Baldy R. (1987) Left cerebral hemisphere hypodensity in discordant schizophrenic twins. A controlled study. Arch. Gen. Psychiat. 44, 625632. 38. Reynolds G. P. (1983) Increased concentrations and lateral asymmetry of amygdala dopamine in schizophrenia. Nature 305, 527-529. 39. Roberts G. W., Colter N., Lofthouse R. et al. (1987) Is there gliosis in schizophrenia? Investigation in the temporal lobe. Biol. Psych&t. 22, 1459-1468. 40. Roberts G. W., Done D. J., Bruton C. and Crow T. J. (1990) A ‘mock up’ of schizophrenia: temporal lobe epilepsy and ~hizophrenia like psychoses. Biol. Psychjat. (in press). 41. Slater E. and Beard A. W. (1963) The schizophrenia like psychoses of epilepsy. &. J. Psych&. 109, 95-150. 42. Spitzer R. L., Endicott J. J. and Robins E. (1975) Research Diagnostic Criteria. Biometrics Research, New York State
Department of Mental Hygiene. 43. Suddath R. L., Casanova M. F., Goldberg T. E., Daniel D. G., Kelsoe J. R. and Weinberger D. R. (1989) Temporal lobe pathology in schizophrenia: a quantitative magnetic resonance imaging study. Am. J. Psychiat. 146,464-472. 44. Tam S. W. and Cook L. (1984) Sigma opiates and certain antipsychotic drugs mutually inhibit (+)-‘H-haloperidol binding in guinea pig brain membranes. Proc. natn. Acad. Sci. 81, 5618-523. 45. Tremblay E., R$esa A. and Ben Ari Y. (1985) Autoradiographic localization of kainic acid binding sites in the human hippocampus. Bruin Res. 343, 378-382. 46. Watkins J. C. and Evans R. H. (1981) Excitatory amino acid transmitter. A. ROD.Pkar~c. Toxicol. 21, 165-204. (Accepted 30 May 1990)
