- Email: [email protected]
Comparison of glutamate and gamma-aminobutyric acid uptake binding sites in frontal and temporal lobes in schizophrenia
Comparison of Glutamate and Gamma-Aminobutyric Acid Uptake Binding Sites in Frontal and Temporal Lobes in Schizophrenia M.D.C. Simpson, Paul Slater, and J.F.W. Deakin Background: Theories of schizophrenia proposing deficiences of amino acid [glutamate, gamma-aminobutyric acid (GABA)] neurons are in accord with the observed temporal lobe pathology of the disease rather than with the newer theory of glutamate hyperinnervation and hyperfunction in areas of prefrontal cortex. This study addresses the issue by measuring specific uptake sites as indices of glutamatergic and GABAergic neuron densities in frontal and temporal lobes. Methods: Frontal cortex (six areas) and temporal lobe (six areas of cortex, amygdala, and hippocampus) were dissected from 19 control autopsy brains and 12 brains from neuroleptic drug-treated schizophrenic patients. Groups had similar ages, postmortem intervals, and storage times. Membranes, prepared from tissue homogenates, were incubated with D-[3H]aspartate to measure neuronal and glial glutamate uptake site binding in 14 areas and with [3H]nipecotic acid to measure neuronal GABA uptake site binding in 11 areas. Results: Glutamate and GABA uptake sites were not reduced in prefrontal and temporal areas. Instead, we found small increases in glutamate uptake sites in prefrontal areas. Some tendency toward increased GABA uptake sites were not disease-related. Conclusions: Our findings concur with other studies that propose locally overabundant glutamate systems in prefrontal cortex in schizophrenia. Losses of amino acid neurons do not accompany the temporal lobe pathology. Biol Psychiatry 1998;44:423– 427 © 1998 Society of Biological Psychiatry Key Words: Schizophrenia, D-[3H]aspartate binding, [3H]nipecotic acid binding, prefrontal cortex, temporal lobes, postmortem brain
From the School of Biological Sciences, Neuroscience Division (MDCS, PS) and Department of Psychiatry (JFWD), University of Manchester, Manchester, United Kingdom. Address reprint requests to Dr. Paul Slater, School of Biological Sciences, Room 1.124 Stopford Building, University of Manchester, Oxford Road, Manchester M13 9PT, United Kingdom. Received June 9, 1997; revised November 7, 1997; accepted February 3, 1998.
© 1998 Society of Biological Psychiatry
Introduction
P
ostmortem brains from schizophrenic patients have reduced neocortical grey matter volumes and neuron disarray (Bogerts 1993). Scanning shows reduced cortical volumes consistent with mild atrophy (Lim et al 1996). Unanswered questions are whether the cortical atrophy is static or slowly progressive and localized or diffuse, and whether it relates to symptoms. Functional brain imaging on patients showing subnormal blood flow responses to tests that activate dorsolateral prefrontal cortex (Kirkby et al 1996) suggests that malfunction of prefrontal cortex and interconnected temporal lobe structures accounts for some symptoms. Finding abnormal neurons and neurotransmitters in frontal neocortex and temporal lobes will more closely link schizophrenia pathophysiology with symptoms. Many cortical pyramidal neurons and hippocampal neurons are glutamatergic, whereas many intrinsic cortical neurons are gamma-aminobutyric acid (GABA)ergic. The abundance of amino acid neurons strongly suggests they are affected by mild neocortical atrophy in schizophrenia. Normal brain glutamate (Toru et al 1988) excludes a widespread loss of glutamatergic neurons. In prefrontal cortex there is evidence of excess markers, mainly receptors and uptake sites, for glutamate systems implying hyperinnervation by terminals (Toru et al 1988; Deakin et al 1989; Ishimaru et al 1994). The presence of normal AMPA receptors (Kurumani et al 1992; Longson et al 1994) implies no excess glutamate synapses. The possibility of having excess terminals in prefrontal cortex is supported by elevated glutamate levels in orbital frontal cortex (Reynolds 1991) and excess glutamate-immunoreactive axons in cingulate cortex (Benes et al 1992); however, excess glutamatergic innervation may downregulate AMPA receptors and thus mask an increase in excitatory amino acid-releasing synapses. There may be failure to eliminate excess N-methyl-D-aspartate (NMDA) receptors in childhood (Slater et al 1993). GABAergic interneurons in prefrontal cortex may be deficient in schizophrenia and mildly up-regulate GABAA receptors in anterior cingulate cortex (Benes et al 1992, 0006-3223/98/$19.00 PII S0006-3223(98)00077-8
424
M.D.C. Simpson et al
BIOL PSYCHIATRY 1998;44:423– 427
Table 1. Details of Subjects and Autopsy Brains Control n Ages (years) Postmortem interval (PMI) (hours) Freezer storage (years)
Schizophrenia
19: 14 men, 5 women 63 6 2 46 6 5
12: 9 men, 3 women 59 6 6 34 6 9
1.0 6 0.12
1.4 6 0.28
Data are mean 6 SEM. There were no differences between control and schizophrenia brains for age, PMI, or freezer storage.
1996). While the presence of normal GABA uptake sites (Kiuchi et al 1989; Simpson et al 1989) argues against large losses of prefrontal GABAergic systems, problems may be confined to a few cortical laminae. Schizophrenia pathology is most apparent in temporal lobes with reduced volumes of some structures and abnormal laterality of hemispheres. Reduced glutamate uptake site binding and normal kainate receptors occur in left polar temporal cortex (Deakin et al 1989). While consistent changes in the several glutamate receptor subtypes may be best seen in individual schizophrenic brains (Ulas and Cotman 1993), in groups of brains there were kainate receptor deficiencies in hippocampus (left), dentate gyrus, and parahippocampus, with little or no change in AMPA or NMDA receptor binding (Kerwin et al 1990). Temporal lobes in schizophrenia may have some reductions in GABA neurons. Below-normal GABA uptake site binding was recorded in amygdala, hippocampus, and left polar cortex (Simpson et al 1989; Reynolds et al 1990). Some reports suggest that schizophrenia entails increases in glutamate system markers in prefrontal cortex with reductions in temporal lobe and inconsistent alterations in GABA systems. Ligand binding to glutamate and GABA transporter sites is preserved in postmortem brain and widely used in pathological studies. We have measured both uptake sites in frontal and temporal lobes in schizophrenia.
10 min, 4°C), the supernatants were discarded, and the pellets were resuspended in buffer, recentrifuged, and stored frozen for 14 days. Pellets were incubated with [3H]nipecotic acid to measure GABA uptake site binding and D-[3H]aspartate to measure glutamate uptake site binding (Simpson et al 1988). Brain area differences between control and schizophrenia were tested by analysis of covariance (ANCOVA) with factors for diagnosis, side of brain, and gender, and positive interactions resulted in further analyses of data from the majority of male subjects. ANCOVA also controlled for age, postmortem interval (PMI), and freezer storage time.
Results [3H]Nipecotic acid binding, measured in 11 brain areas, found no overall reduction in GABA uptake sites in the schizophrenic brains (Table 2). Statistically significant data analyses pinpointed left hemispheres and included cingulate gyrus, hippocampus, and temporal cortex (polar and superior temporal gyrus), with some interactions between diagnosis and gender; however, it is probable that the data are influenced by the three brains from female control subjects having lower GABA uptake site binding (not shown) compared to male subjects. The significant data from the hippocampus produced large standard errors, perhaps due to variations in the sampling of this complex structure. Some caution is therefore required in interpreting hippocampal [3H]nipecotic acid binding. D-[3H]Aspartate binding, measured in 14 prefrontal and temporal areas, also did not find a systematic reduction in glutamate uptake site binding in schizophrenia (Table 3). In fact, the data revealed more of a tendency in the schizophrenia brains to increased D-[3H]aspartate binding in the majority of frontal/prefrontal cortical areas; however, a significant difference between the two diagnostic groups occurred only with data from anterior cingulate gyrus (Table 3), with borderline significance for medial orbital gyrus (p 5 .07 for diagnosis). An increase in glutamate uptake sites in medial orbital gyrus was related to gender in the majority of subjects (men).
Methods and Materials Brains were taken during autopsies on neuroleptic drug-treated patients with confirmed chronic schizophrenia and control subjects with no histories of psychiatric or neurological illness (Table 1). Case notes excluded alcohol and substance abuse, neurodegenerative disease, and antemortem anoxia or coma. Brains were stored at 270°C after initial snap-freezing. After warming to 210°C, brains were sliced coronally (10 mm) before prefrontal cortex and temporal lobe areas were dissected. Frozen tissues were finely chopped for homogeneous sampling, and more than one part of large structures was sampled. Tissues were thawed to 4°C and homogenized in 100 volumes of 50 mmol/L Tris–acetate, pH 7.4. After centrifugation (19,000 g,
Discussion Previously we reported increased glutamate system markers (uptake sites and receptors) in orbital frontal cortex in schizophrenia (Deakin et al 1989; Simpson et al 1992). Although this study examined many more prefrontal areas, we found no real evidence of widespread increases in glutamate uptake site binding beyond some small increases of borderline statistical significance, mostly in left hemispheres. Because the disease-related increase in glutamate markers in ventral frontal cortex is replicated in two studies, it is probably not an artifact related to the
Amino Acid Uptake Sites in Schizophrenia
BIOL PSYCHIATRY 1998;44:423– 427
425
Table 2. GABA Uptake Site Binding in Prefrontal and Temporal Brain Areas from Control (C) and Schizophrenic (SZ) Subjects [3H]Nipecotic acid bound Area Anterior cingulate gyrus Mid cingulate gyrus Anterior precentral gyrus Medial polar temporal cortex Lateral polar temporal cortex Amygdala Hippocampus Anterior parahippocampal gyrus Posterior parahippocampal gyrus Anterior superior temporal gyrus Posterior superior temporal gyrus
Diagnosis
Left
Right
C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ
218 6 21 (19) 231 6 23 (11)a 655 6 55 (19) 774 6 102 (11)a 1205 6 121 (19) 1159 6 161 (11) 808 6 57 (19) 686 6 57 (11) 680 6 72 (18) 578 6 74 (11)a 636 6 66 (18) 655 6 47 (11) 891 6 238 (19) 1274 6 341 (12)b 605 6 62 (19) 544 6 66 (12) 457 6 54 (18) 515 6 80 (10) 978 6 89 (19) 993 6 111 (11)a 167 6 32 (19) 214 6 61 (12)
219 6 30 (19) 228 6 20 (11) 827 6 79 (19) 1006 6 252 (12) 1039 6 127 (18) 964 6 155 (12) 862 6 77 (19) 784 6 79 (9) 456 6 41 (19) 528 6 54 (10) 688 6 61 (19) 588 6 66 (10) 2682 6 582 (18) 2694 6 599 (12) 671 6 64 (19) 680 6 81 (12) 479 6 60 (19) 464 6 53 (12) 749 6 62 (19) 805 6 65 (12) 341 6 48 (19) 381 6 55 (12)
Data are mean 6 SEM fmol/mg protein (n). a Significant effect of diagnosis 3 gender (p , .05). b Significant effect of diagnosis 3 gender 3 side (p , .05).
demographic features of the patient groups. Intersubject variations and methodological differences may account for differences in the scale of increased glutamatergic markers reported in different studies. Antemortem exposure to neuroleptic drugs cannot account for increased glutamate uptake sites being localized to frontal cortex, nor explain why increased serotonin (5-HT)1A receptor binding associated with glutamatergic pyramidal neurons occurs mainly in frontal cortex in schizophrenia (Simpson et al 1996). Furthermore, some GABAA receptors are localized on pyramidal neurons, and increases in GABAA receptors in frontal cortex (Hanada et al 1987) are compatible with either increased pyramidal cell densities or GABAA receptor up-regulation caused by GABA cell loss (Benes et al 1996). Greater numbers of structural elements of pyramidal cells is probably the more convincing explanation for the increased density of 5-HT1A receptors, especially since we find normal GABA uptake sites, synonymous with GABAergic neurons and terminals, in frontal cortex in schizophrenia. The conclusion that pyramidal cells and associated synapses are somewhat overabundant in schizophrenic frontal cortex (Deakin et al 1989) is supported by reports of excess glutamate immunoreactive fibers in cingulate cortex and neurons in frontal cortex (Benes et al 1992). Earlier, reduced glutamate uptake sites in temporal
cortex were limited to patients who were off neuroleptics when they died (Deakin et al 1989). We had 2 such patients in the present study, and separate analysis of these data is not feasible. We find no overall reduction in glutamate uptake sites in schizophrenia; however, D-[3H]aspartate binding probably labels neuronal and glial transporters, with many of the latter in brain. Reduced neuronal uptake might be compensated by more expression of glial transporters or somehow normalized by neuroleptics; however, male subjects may have deficient glutamate uptake sites, especially in anterior parahippocampal gyrus, which is anatomically close to polar temporal cortex, where loss of uptake sites was previously found (Deakin et al 1989). Based upon D-[3H]aspartate binding data, there is scant evidence of reduced glutamatergic innervation of temporal cortex in schizophrenia. No consistent reductions in cortical glutamate receptors are reported, implying that many intrinsic cortical neurons are intact. We speculate that any relatively small faults in glutamatergic innervation of anterior temporal cortex may be caused by a primary problem in pathways from ventral frontal cortex. Some GABAergic interneurons in cortical circuits may be lost in schizophrenia, thus affecting GABAA receptors (Benes et al 1996). GABA uptake sites were measured previously to investigate reductions in GABA neurons in
426
M.D.C. Simpson et al
BIOL PSYCHIATRY 1998;44:423– 427
Table 3. Glutamate Uptake Site Binding in Prefrontal/Temporal Brain Areas from Control (C) and Schizophrenic (SZ) Subjects D-[3H]Aspartate bound Area
Diagnosis
Left
Right
C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ
1180 6 107 (19) 1431 6 140 (11)a,b 4687 6 379 (19) 5612 6 464 (11) 731 6 55 (18) 702 6 58 (11)c 2329 6 250 (19) 2311 6 317 (12) 697 6 58 (18) 765 6 63 (11)b 4302 6 722 (19) 4083 6 913 (11) 2988 6 223 (19) 2679 6 189 (11) 2196 6 127 (18) 2141 6 232 (11) 3023 6 346 (18) 3404 6 317 (11) 4179 6 766 (19) 7252 6 2102 (12) 4970 6 545 (19) 3792 6 324 (12) 1941 6 167 (18) 2231 6 269 (10) 2897 6 185 (19) 3025 6 268 (11) 795 6 160 (19) 1172 6 279 (12)
1246 6 148 (19) 1342 6 118 (11) 5194 6 527 (16) 5464 6 370 (11) 522 6 52 (19) 637 6 56 (10) 2455 6 195 (18) 3237 6 393 (10) 608 6 61 (18) 656 6 87 (10) 5398 6 686 (18) 4660 6 477 (12) 3742 6 322 (19) 2950 6 258 (10) 1409 6 129 (19) 1623 6 89 (10) 4137 6 324 (19) 3501 6 247 (10) 9886 6 1939 (18) 10753 6 2075 (12) 4857 6 565 (19) 4608 6 501 (12) 1607 6 144 (19) 1644 6 112 (12) 1833 6 136 (19) 1864 6 125 (12) 1529 6 199 (19) 1688 6 215 (12)
Anterior cingulate gyrus Mid cingulate gyrus Anterior straight gyrus Posterior straight gyrus Anterior medial orbital gyrus Anterior precentral gyrus Medial polar temporal cortex Lateral polar temporal cortex Amygdala Hippocampus Anterior parahippocampal gyrus Posterior parahippocampal gyrus Anterior superior temporal gyrus Posterior superior temporal gyrus Data are mean 6 SEM fmol/mg protein (n). a Significant effect of diagnosis (p , .05). b Significant effect of gender (p , .05). c Significant effect of diagnosis 3 side (p , .05).
schizophrenia, which occurred in amygdala, hippocampus, and left polar temporal cortex, but only to a small extent in ventral frontal cortex (Simpson et al 1989). Because there was no reduction in [3H]nipecotic acid binding in frontal cortex in the present study, there is no significant deficit in schizophrenia of prefrontal GABA neurons and no GABAA receptor up-regulation (unpublished data). GABAergic neuron counts were not altered in prefrontal cortex in schizophrenia (Akbarian et al 1995). Therefore it is not possible to link prefrontal hypofunction in schizophrenic patients (Weinberger 1995) with reductions in inhibitory GABAergic neurons; however, there is further evidence from the present study that GABAergic neurons in parts of temporal cortex are affected by localized pathology in schizophrenia. The present findings give further evidence of abnormal amino acid neurons in schizophrenia. An emerging pattern is increased markers for glutamatergic neurons in ventral prefrontal cortex, which is compatible with the hypothesis that within parts of prefrontal cortex there is hyperinner-
vation by glutamatergic neurons, which may be genetically determined. Thus impaired performance in frontal tests occurred in both schizophrenics and their close normal relatives, suggesting a genetic contribution (Hellewell and Deakin 1994). But only the patients had memory impairments, which suggests that neurodegeneration, or other pathology, may affect temporal lobes and produce psychosis after late adolescence. There may be reduced extrinsic innervation by glutamatergic nerve terminals, perhaps from prefrontal cortex, because of abnormal postnatal remodeling of glutamatergic elements within the frontal lobes. For this reason, frontotemporal connections may be vulnerable to neuropathologic processes in schizophrenia, and slow neurotoxicity may remove more afferent terminals from left anterior temporal lobe.
M.D.C.S. was supported by a Wellcome Trust project grant.
Amino Acid Uptake Sites in Schizophrenia
References Akbarian S, Kim JJ, Potkin SG, Hagman JO, Tafazzoli A, Bunney WE, et al (1995): Gene expression for glutamic acid decarboxylase is reduced without loss of neurons in prefrontal cortex of schizophrenics. Arch Gen Psychiatry 52:258 –266. Benes FM, Sorensen I, Vincent SL, Bird ED, Sathi M (1992): Increased density of glutamate-immunoreactive vertical processes in superficial laminae in cingulate cortex of schizophrenic brain. Cereb Cortex 2:503–512. Benes FM, Vincent SL, Marie A, Khan Y (1996): Up-regulation of GABAA receptor binding on neurons of the prefrontal cortex in schizophrenic subjects. Neuroscience 75:1021– 1031. Bogerts B (1993): Recent advances in the neuropathology of schizophrenia. Schizophr Bull 19:431– 445. Deakin JFW, Slater P, Simpson MDC, Gilchrist AC, Skan WJ, Royston MC, et al (1989): Frontal cortical and left temporal glutamatergic dysfunction in schizophrenia. J Neurochem 52:1781–1786. Hanada S, Mita T, Nishino N, Tanaka C (1987): [3H]Muscimol binding sites increased in autopsied brains of chronic schizophrenics. Life Sci 40:259 –266. Hellewell JSE, Deakin JFW (1994): Impaired frontal lobe function in schizophrenia is familial. Schizophr Res 11:149. Ishimaru M, Kurumaji A, Toru M (1994): Increases in strychnine-insensitive glycine binding sites in cerebral cortex of chronic schizophrenics: Evidence for glutamate hypothesis. Biol Psychiatry 35:84 –95. Kerwin RW, Patel S, Meldrum BS (1990): Quantitative autoradiographic analysis of glutamate binding sites in the hippocampal formation in normal and schizophrenic brain post mortem. Neuroscience 39:25–32. Kirkby BS, Berman KF, Weinberger DR (1996): Neurodevelopment and schizophrenia: The contribution of functional brain imaging. In: Waddington JL, Buckley PF, editors. The Neurodevelopmental Basis of Schizophrenia. Austin, TX: RG Landes, pp 119 –133. Lim KO, Tew W, Kushner M, Chow K, Matsumoto B, DeLisi LE (1996): Cortical gray matter volume deficit in patients
BIOL PSYCHIATRY 1998;44:423– 427
427
with first-episode schizophrenia. Am J Psychiatry 153:1548 – 1553. Longson D, Simpson MDC, Deakin JFW (1994): Autoradiographic study of [3H]AMPA binding in frontal cortex in schizophrenia. Schizophr Res 11:121. Reynolds GP (1991): Amino acid transmitters and dopamine function in the brain in schizophrenia. In: Racagni G, Brunello N, Fokuda T, editors. Biological Psychiatry, vol 1. Amsterdam: Excerpta Medica, pp 491– 494. Reynolds GP, Czudek C, Andrews HB (1990): Deficit and hemispheric asymmetry of GABA uptake sites in the hippocampus in schizophrenia. Biol Psychiatry 27:1038 –1044. Simpson MDC, Cross AJ, Slater P, Deakin JFW (1988): GABA uptake sites in Alzheimer’s disease. J Neural Transm Gen Sect 71:219 –226. Simpson MDC, Slater P, Deakin JFW, Royston MC, Skan WJ (1989): Reduced GABA uptake sites in the temporal lobe in schizophrenia. Neurosci Lett 107:211–215. Simpson MDC, Royston MC, Slater P, Deakin JFW (1992): Neurochemical abnormalities of the cerebral cortex in schizophrenia. Schizophr Res 6:133–134. Simpson MDC, Lubman D, Slater P, Deakin JFW (1996): Autoradiography with [3H]8-OH-DPAT reveals increases in 5-HT1A receptors in ventral prefrontal cortex in schizophrenia. Biol Psychiatry 39:919 –928. Slater P, McConnell SE, D’Souza SW, Barson AJ (1993): Postnatal changes in N-methyl-D-aspartate receptor binding and stimulation by glutamate and glycine of [3H]MK-801 binding in human temporal cortex. Br J Pharmacol 108: 1143–1149. Toru M, Watanabe S, Shibuya H, Nishikawa T, Noda K, Mitsushio H, et al (1988): Neurotransmitters, receptors and neuropeptides in post-mortem brains of chronic schizophrenic patients. Acta Psychiatr Scand 78:121–137. Ulas J, Cotman CW (1993): Excitatory amino acid receptors in schizophrenia. Schizophr Bull 19:105–117. Weinberger DR (1995): Schizophrenia as a neurodevelopmental disorder. In: Hirsch SR, Weinberger DR, editors. Schizophrenia. Oxford: Blackwell, pp 293–323.
From the School of Biological Sciences, Neuroscience Division (MDCS, PS) and Department of Psychiatry (JFWD), University of Manchester, Manchester, United Kingdom. Address reprint requests to Dr. Paul Slater, School of Biological Sciences, Room 1.124 Stopford Building, University of Manchester, Oxford Road, Manchester M13 9PT, United Kingdom. Received June 9, 1997; revised November 7, 1997; accepted February 3, 1998.
© 1998 Society of Biological Psychiatry
Introduction
P
ostmortem brains from schizophrenic patients have reduced neocortical grey matter volumes and neuron disarray (Bogerts 1993). Scanning shows reduced cortical volumes consistent with mild atrophy (Lim et al 1996). Unanswered questions are whether the cortical atrophy is static or slowly progressive and localized or diffuse, and whether it relates to symptoms. Functional brain imaging on patients showing subnormal blood flow responses to tests that activate dorsolateral prefrontal cortex (Kirkby et al 1996) suggests that malfunction of prefrontal cortex and interconnected temporal lobe structures accounts for some symptoms. Finding abnormal neurons and neurotransmitters in frontal neocortex and temporal lobes will more closely link schizophrenia pathophysiology with symptoms. Many cortical pyramidal neurons and hippocampal neurons are glutamatergic, whereas many intrinsic cortical neurons are gamma-aminobutyric acid (GABA)ergic. The abundance of amino acid neurons strongly suggests they are affected by mild neocortical atrophy in schizophrenia. Normal brain glutamate (Toru et al 1988) excludes a widespread loss of glutamatergic neurons. In prefrontal cortex there is evidence of excess markers, mainly receptors and uptake sites, for glutamate systems implying hyperinnervation by terminals (Toru et al 1988; Deakin et al 1989; Ishimaru et al 1994). The presence of normal AMPA receptors (Kurumani et al 1992; Longson et al 1994) implies no excess glutamate synapses. The possibility of having excess terminals in prefrontal cortex is supported by elevated glutamate levels in orbital frontal cortex (Reynolds 1991) and excess glutamate-immunoreactive axons in cingulate cortex (Benes et al 1992); however, excess glutamatergic innervation may downregulate AMPA receptors and thus mask an increase in excitatory amino acid-releasing synapses. There may be failure to eliminate excess N-methyl-D-aspartate (NMDA) receptors in childhood (Slater et al 1993). GABAergic interneurons in prefrontal cortex may be deficient in schizophrenia and mildly up-regulate GABAA receptors in anterior cingulate cortex (Benes et al 1992, 0006-3223/98/$19.00 PII S0006-3223(98)00077-8
424
M.D.C. Simpson et al
BIOL PSYCHIATRY 1998;44:423– 427
Table 1. Details of Subjects and Autopsy Brains Control n Ages (years) Postmortem interval (PMI) (hours) Freezer storage (years)
Schizophrenia
19: 14 men, 5 women 63 6 2 46 6 5
12: 9 men, 3 women 59 6 6 34 6 9
1.0 6 0.12
1.4 6 0.28
Data are mean 6 SEM. There were no differences between control and schizophrenia brains for age, PMI, or freezer storage.
1996). While the presence of normal GABA uptake sites (Kiuchi et al 1989; Simpson et al 1989) argues against large losses of prefrontal GABAergic systems, problems may be confined to a few cortical laminae. Schizophrenia pathology is most apparent in temporal lobes with reduced volumes of some structures and abnormal laterality of hemispheres. Reduced glutamate uptake site binding and normal kainate receptors occur in left polar temporal cortex (Deakin et al 1989). While consistent changes in the several glutamate receptor subtypes may be best seen in individual schizophrenic brains (Ulas and Cotman 1993), in groups of brains there were kainate receptor deficiencies in hippocampus (left), dentate gyrus, and parahippocampus, with little or no change in AMPA or NMDA receptor binding (Kerwin et al 1990). Temporal lobes in schizophrenia may have some reductions in GABA neurons. Below-normal GABA uptake site binding was recorded in amygdala, hippocampus, and left polar cortex (Simpson et al 1989; Reynolds et al 1990). Some reports suggest that schizophrenia entails increases in glutamate system markers in prefrontal cortex with reductions in temporal lobe and inconsistent alterations in GABA systems. Ligand binding to glutamate and GABA transporter sites is preserved in postmortem brain and widely used in pathological studies. We have measured both uptake sites in frontal and temporal lobes in schizophrenia.
10 min, 4°C), the supernatants were discarded, and the pellets were resuspended in buffer, recentrifuged, and stored frozen for 14 days. Pellets were incubated with [3H]nipecotic acid to measure GABA uptake site binding and D-[3H]aspartate to measure glutamate uptake site binding (Simpson et al 1988). Brain area differences between control and schizophrenia were tested by analysis of covariance (ANCOVA) with factors for diagnosis, side of brain, and gender, and positive interactions resulted in further analyses of data from the majority of male subjects. ANCOVA also controlled for age, postmortem interval (PMI), and freezer storage time.
Results [3H]Nipecotic acid binding, measured in 11 brain areas, found no overall reduction in GABA uptake sites in the schizophrenic brains (Table 2). Statistically significant data analyses pinpointed left hemispheres and included cingulate gyrus, hippocampus, and temporal cortex (polar and superior temporal gyrus), with some interactions between diagnosis and gender; however, it is probable that the data are influenced by the three brains from female control subjects having lower GABA uptake site binding (not shown) compared to male subjects. The significant data from the hippocampus produced large standard errors, perhaps due to variations in the sampling of this complex structure. Some caution is therefore required in interpreting hippocampal [3H]nipecotic acid binding. D-[3H]Aspartate binding, measured in 14 prefrontal and temporal areas, also did not find a systematic reduction in glutamate uptake site binding in schizophrenia (Table 3). In fact, the data revealed more of a tendency in the schizophrenia brains to increased D-[3H]aspartate binding in the majority of frontal/prefrontal cortical areas; however, a significant difference between the two diagnostic groups occurred only with data from anterior cingulate gyrus (Table 3), with borderline significance for medial orbital gyrus (p 5 .07 for diagnosis). An increase in glutamate uptake sites in medial orbital gyrus was related to gender in the majority of subjects (men).
Methods and Materials Brains were taken during autopsies on neuroleptic drug-treated patients with confirmed chronic schizophrenia and control subjects with no histories of psychiatric or neurological illness (Table 1). Case notes excluded alcohol and substance abuse, neurodegenerative disease, and antemortem anoxia or coma. Brains were stored at 270°C after initial snap-freezing. After warming to 210°C, brains were sliced coronally (10 mm) before prefrontal cortex and temporal lobe areas were dissected. Frozen tissues were finely chopped for homogeneous sampling, and more than one part of large structures was sampled. Tissues were thawed to 4°C and homogenized in 100 volumes of 50 mmol/L Tris–acetate, pH 7.4. After centrifugation (19,000 g,
Discussion Previously we reported increased glutamate system markers (uptake sites and receptors) in orbital frontal cortex in schizophrenia (Deakin et al 1989; Simpson et al 1992). Although this study examined many more prefrontal areas, we found no real evidence of widespread increases in glutamate uptake site binding beyond some small increases of borderline statistical significance, mostly in left hemispheres. Because the disease-related increase in glutamate markers in ventral frontal cortex is replicated in two studies, it is probably not an artifact related to the
Amino Acid Uptake Sites in Schizophrenia
BIOL PSYCHIATRY 1998;44:423– 427
425
Table 2. GABA Uptake Site Binding in Prefrontal and Temporal Brain Areas from Control (C) and Schizophrenic (SZ) Subjects [3H]Nipecotic acid bound Area Anterior cingulate gyrus Mid cingulate gyrus Anterior precentral gyrus Medial polar temporal cortex Lateral polar temporal cortex Amygdala Hippocampus Anterior parahippocampal gyrus Posterior parahippocampal gyrus Anterior superior temporal gyrus Posterior superior temporal gyrus
Diagnosis
Left
Right
C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ
218 6 21 (19) 231 6 23 (11)a 655 6 55 (19) 774 6 102 (11)a 1205 6 121 (19) 1159 6 161 (11) 808 6 57 (19) 686 6 57 (11) 680 6 72 (18) 578 6 74 (11)a 636 6 66 (18) 655 6 47 (11) 891 6 238 (19) 1274 6 341 (12)b 605 6 62 (19) 544 6 66 (12) 457 6 54 (18) 515 6 80 (10) 978 6 89 (19) 993 6 111 (11)a 167 6 32 (19) 214 6 61 (12)
219 6 30 (19) 228 6 20 (11) 827 6 79 (19) 1006 6 252 (12) 1039 6 127 (18) 964 6 155 (12) 862 6 77 (19) 784 6 79 (9) 456 6 41 (19) 528 6 54 (10) 688 6 61 (19) 588 6 66 (10) 2682 6 582 (18) 2694 6 599 (12) 671 6 64 (19) 680 6 81 (12) 479 6 60 (19) 464 6 53 (12) 749 6 62 (19) 805 6 65 (12) 341 6 48 (19) 381 6 55 (12)
Data are mean 6 SEM fmol/mg protein (n). a Significant effect of diagnosis 3 gender (p , .05). b Significant effect of diagnosis 3 gender 3 side (p , .05).
demographic features of the patient groups. Intersubject variations and methodological differences may account for differences in the scale of increased glutamatergic markers reported in different studies. Antemortem exposure to neuroleptic drugs cannot account for increased glutamate uptake sites being localized to frontal cortex, nor explain why increased serotonin (5-HT)1A receptor binding associated with glutamatergic pyramidal neurons occurs mainly in frontal cortex in schizophrenia (Simpson et al 1996). Furthermore, some GABAA receptors are localized on pyramidal neurons, and increases in GABAA receptors in frontal cortex (Hanada et al 1987) are compatible with either increased pyramidal cell densities or GABAA receptor up-regulation caused by GABA cell loss (Benes et al 1996). Greater numbers of structural elements of pyramidal cells is probably the more convincing explanation for the increased density of 5-HT1A receptors, especially since we find normal GABA uptake sites, synonymous with GABAergic neurons and terminals, in frontal cortex in schizophrenia. The conclusion that pyramidal cells and associated synapses are somewhat overabundant in schizophrenic frontal cortex (Deakin et al 1989) is supported by reports of excess glutamate immunoreactive fibers in cingulate cortex and neurons in frontal cortex (Benes et al 1992). Earlier, reduced glutamate uptake sites in temporal
cortex were limited to patients who were off neuroleptics when they died (Deakin et al 1989). We had 2 such patients in the present study, and separate analysis of these data is not feasible. We find no overall reduction in glutamate uptake sites in schizophrenia; however, D-[3H]aspartate binding probably labels neuronal and glial transporters, with many of the latter in brain. Reduced neuronal uptake might be compensated by more expression of glial transporters or somehow normalized by neuroleptics; however, male subjects may have deficient glutamate uptake sites, especially in anterior parahippocampal gyrus, which is anatomically close to polar temporal cortex, where loss of uptake sites was previously found (Deakin et al 1989). Based upon D-[3H]aspartate binding data, there is scant evidence of reduced glutamatergic innervation of temporal cortex in schizophrenia. No consistent reductions in cortical glutamate receptors are reported, implying that many intrinsic cortical neurons are intact. We speculate that any relatively small faults in glutamatergic innervation of anterior temporal cortex may be caused by a primary problem in pathways from ventral frontal cortex. Some GABAergic interneurons in cortical circuits may be lost in schizophrenia, thus affecting GABAA receptors (Benes et al 1996). GABA uptake sites were measured previously to investigate reductions in GABA neurons in
426
M.D.C. Simpson et al
BIOL PSYCHIATRY 1998;44:423– 427
Table 3. Glutamate Uptake Site Binding in Prefrontal/Temporal Brain Areas from Control (C) and Schizophrenic (SZ) Subjects D-[3H]Aspartate bound Area
Diagnosis
Left
Right
C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ C SZ
1180 6 107 (19) 1431 6 140 (11)a,b 4687 6 379 (19) 5612 6 464 (11) 731 6 55 (18) 702 6 58 (11)c 2329 6 250 (19) 2311 6 317 (12) 697 6 58 (18) 765 6 63 (11)b 4302 6 722 (19) 4083 6 913 (11) 2988 6 223 (19) 2679 6 189 (11) 2196 6 127 (18) 2141 6 232 (11) 3023 6 346 (18) 3404 6 317 (11) 4179 6 766 (19) 7252 6 2102 (12) 4970 6 545 (19) 3792 6 324 (12) 1941 6 167 (18) 2231 6 269 (10) 2897 6 185 (19) 3025 6 268 (11) 795 6 160 (19) 1172 6 279 (12)
1246 6 148 (19) 1342 6 118 (11) 5194 6 527 (16) 5464 6 370 (11) 522 6 52 (19) 637 6 56 (10) 2455 6 195 (18) 3237 6 393 (10) 608 6 61 (18) 656 6 87 (10) 5398 6 686 (18) 4660 6 477 (12) 3742 6 322 (19) 2950 6 258 (10) 1409 6 129 (19) 1623 6 89 (10) 4137 6 324 (19) 3501 6 247 (10) 9886 6 1939 (18) 10753 6 2075 (12) 4857 6 565 (19) 4608 6 501 (12) 1607 6 144 (19) 1644 6 112 (12) 1833 6 136 (19) 1864 6 125 (12) 1529 6 199 (19) 1688 6 215 (12)
Anterior cingulate gyrus Mid cingulate gyrus Anterior straight gyrus Posterior straight gyrus Anterior medial orbital gyrus Anterior precentral gyrus Medial polar temporal cortex Lateral polar temporal cortex Amygdala Hippocampus Anterior parahippocampal gyrus Posterior parahippocampal gyrus Anterior superior temporal gyrus Posterior superior temporal gyrus Data are mean 6 SEM fmol/mg protein (n). a Significant effect of diagnosis (p , .05). b Significant effect of gender (p , .05). c Significant effect of diagnosis 3 side (p , .05).
schizophrenia, which occurred in amygdala, hippocampus, and left polar temporal cortex, but only to a small extent in ventral frontal cortex (Simpson et al 1989). Because there was no reduction in [3H]nipecotic acid binding in frontal cortex in the present study, there is no significant deficit in schizophrenia of prefrontal GABA neurons and no GABAA receptor up-regulation (unpublished data). GABAergic neuron counts were not altered in prefrontal cortex in schizophrenia (Akbarian et al 1995). Therefore it is not possible to link prefrontal hypofunction in schizophrenic patients (Weinberger 1995) with reductions in inhibitory GABAergic neurons; however, there is further evidence from the present study that GABAergic neurons in parts of temporal cortex are affected by localized pathology in schizophrenia. The present findings give further evidence of abnormal amino acid neurons in schizophrenia. An emerging pattern is increased markers for glutamatergic neurons in ventral prefrontal cortex, which is compatible with the hypothesis that within parts of prefrontal cortex there is hyperinner-
vation by glutamatergic neurons, which may be genetically determined. Thus impaired performance in frontal tests occurred in both schizophrenics and their close normal relatives, suggesting a genetic contribution (Hellewell and Deakin 1994). But only the patients had memory impairments, which suggests that neurodegeneration, or other pathology, may affect temporal lobes and produce psychosis after late adolescence. There may be reduced extrinsic innervation by glutamatergic nerve terminals, perhaps from prefrontal cortex, because of abnormal postnatal remodeling of glutamatergic elements within the frontal lobes. For this reason, frontotemporal connections may be vulnerable to neuropathologic processes in schizophrenia, and slow neurotoxicity may remove more afferent terminals from left anterior temporal lobe.
M.D.C.S. was supported by a Wellcome Trust project grant.
Amino Acid Uptake Sites in Schizophrenia
References Akbarian S, Kim JJ, Potkin SG, Hagman JO, Tafazzoli A, Bunney WE, et al (1995): Gene expression for glutamic acid decarboxylase is reduced without loss of neurons in prefrontal cortex of schizophrenics. Arch Gen Psychiatry 52:258 –266. Benes FM, Sorensen I, Vincent SL, Bird ED, Sathi M (1992): Increased density of glutamate-immunoreactive vertical processes in superficial laminae in cingulate cortex of schizophrenic brain. Cereb Cortex 2:503–512. Benes FM, Vincent SL, Marie A, Khan Y (1996): Up-regulation of GABAA receptor binding on neurons of the prefrontal cortex in schizophrenic subjects. Neuroscience 75:1021– 1031. Bogerts B (1993): Recent advances in the neuropathology of schizophrenia. Schizophr Bull 19:431– 445. Deakin JFW, Slater P, Simpson MDC, Gilchrist AC, Skan WJ, Royston MC, et al (1989): Frontal cortical and left temporal glutamatergic dysfunction in schizophrenia. J Neurochem 52:1781–1786. Hanada S, Mita T, Nishino N, Tanaka C (1987): [3H]Muscimol binding sites increased in autopsied brains of chronic schizophrenics. Life Sci 40:259 –266. Hellewell JSE, Deakin JFW (1994): Impaired frontal lobe function in schizophrenia is familial. Schizophr Res 11:149. Ishimaru M, Kurumaji A, Toru M (1994): Increases in strychnine-insensitive glycine binding sites in cerebral cortex of chronic schizophrenics: Evidence for glutamate hypothesis. Biol Psychiatry 35:84 –95. Kerwin RW, Patel S, Meldrum BS (1990): Quantitative autoradiographic analysis of glutamate binding sites in the hippocampal formation in normal and schizophrenic brain post mortem. Neuroscience 39:25–32. Kirkby BS, Berman KF, Weinberger DR (1996): Neurodevelopment and schizophrenia: The contribution of functional brain imaging. In: Waddington JL, Buckley PF, editors. The Neurodevelopmental Basis of Schizophrenia. Austin, TX: RG Landes, pp 119 –133. Lim KO, Tew W, Kushner M, Chow K, Matsumoto B, DeLisi LE (1996): Cortical gray matter volume deficit in patients
BIOL PSYCHIATRY 1998;44:423– 427
427
with first-episode schizophrenia. Am J Psychiatry 153:1548 – 1553. Longson D, Simpson MDC, Deakin JFW (1994): Autoradiographic study of [3H]AMPA binding in frontal cortex in schizophrenia. Schizophr Res 11:121. Reynolds GP (1991): Amino acid transmitters and dopamine function in the brain in schizophrenia. In: Racagni G, Brunello N, Fokuda T, editors. Biological Psychiatry, vol 1. Amsterdam: Excerpta Medica, pp 491– 494. Reynolds GP, Czudek C, Andrews HB (1990): Deficit and hemispheric asymmetry of GABA uptake sites in the hippocampus in schizophrenia. Biol Psychiatry 27:1038 –1044. Simpson MDC, Cross AJ, Slater P, Deakin JFW (1988): GABA uptake sites in Alzheimer’s disease. J Neural Transm Gen Sect 71:219 –226. Simpson MDC, Slater P, Deakin JFW, Royston MC, Skan WJ (1989): Reduced GABA uptake sites in the temporal lobe in schizophrenia. Neurosci Lett 107:211–215. Simpson MDC, Royston MC, Slater P, Deakin JFW (1992): Neurochemical abnormalities of the cerebral cortex in schizophrenia. Schizophr Res 6:133–134. Simpson MDC, Lubman D, Slater P, Deakin JFW (1996): Autoradiography with [3H]8-OH-DPAT reveals increases in 5-HT1A receptors in ventral prefrontal cortex in schizophrenia. Biol Psychiatry 39:919 –928. Slater P, McConnell SE, D’Souza SW, Barson AJ (1993): Postnatal changes in N-methyl-D-aspartate receptor binding and stimulation by glutamate and glycine of [3H]MK-801 binding in human temporal cortex. Br J Pharmacol 108: 1143–1149. Toru M, Watanabe S, Shibuya H, Nishikawa T, Noda K, Mitsushio H, et al (1988): Neurotransmitters, receptors and neuropeptides in post-mortem brains of chronic schizophrenic patients. Acta Psychiatr Scand 78:121–137. Ulas J, Cotman CW (1993): Excitatory amino acid receptors in schizophrenia. Schizophr Bull 19:105–117. Weinberger DR (1995): Schizophrenia as a neurodevelopmental disorder. In: Hirsch SR, Weinberger DR, editors. Schizophrenia. Oxford: Blackwell, pp 293–323.