Protection against ischemic hippocampal CAl damage in the rat with a n e w non-NMDA antagonist, NBQX Diemer NH, Jmgensen MB, Johansen FF, Sheardown M, Honor&T. Protection against ischemic hippocampal CAI damage in the rat with a new non-NMDA antagonist, NBQX. Acta Neurol Scand 1992: 86: 45-49. Two glutamate antagonists were tested in a rat model of complete, transient cerebral ischemia. Six days after 10 min ischemia the mean loss of hippocampal CA 1 pyramidal neurones was 73 ”/, . Administration of the proprionic acid) antagonist AMPA (a-amino-3-hydroxy-5-methyl-4-isoxazole reduced the NBQX (2,3-dihydro-6-nitro-7-sulfamoyl-benzo(F)quinoxal~ne) when given before, pyramidal neurone loss to 1 11 and 15 immediately after or 1 h after ischemia, respectively. MK-801 (dizocilpine), a competitive NMDA antagonist gave no protection in this model. We suggest that the AMPA receptor transduction mechanisms are sensitized by ischemia and that the postischemic blockade of the main glutamatergic input to the CA1 cells with NBQX impairs the deleterious effect of “normal” postischemic excitatory transmission.
z,
I
’
N. H. Diemer ’, M. B. Jirgensen F. F. Johansen’, M. Sheardown’: T. Honore’
’
Cerebral Ischemia Research Group, Institute of Neuropathology, University of Copenhagen, NOVD-Nordisk. CNS Division, S ~ b o r g , Denmark
z,
After more than a few min of cerebral ischemia in the rat the hippocampal CA1 pyramidal cells are irreversibly damaged (1,2). The main excitatory input to these cells comes from the CA3 pyramidal cells (3). Transsection of the Schaffer collaterals and commisurals from CA3 before (4-6) or immediately after ischemia (4) has a protective effect. Autoradiographically the CA1 region is characterized by a high density of glutamate receptors of the NMDA (N-methyl-D-aspartate) and AMPA (a-amino-3hydroxy-5-Methyl-4-isoxazole proprionic acid) subtype (formerly named quisqualate receptor). The normal fast excitation in the SchafTer collateral synapse is mediated by glutamate and AMPA-receptors whereas the NMDA receptor seems to be involved in more excessive stimulation such as long-term potentiation (7). A number of both competitive and non-competitive antagonists to the NMDA receptor are available (8), but in vivo tests of their possible ischemia protective effect have given contradictory results (9-1 l), seemingly dependent on the intraischemic brain temperature and the amount of residual blood flow (12, 13). Blockade of the AMPA receptor with systemically injected antagonists has been made possible recently, with the development of the quiiioxaline dione NBQX (14). In gerbils this compound offered protection after administration up to 4 h after ischemia. The present paper report on the ability of NBQX to ameliorate CA1
Nils Henrik Dierner, institute of Neuropathology, University of Copenhagen, 11 Frederik V’s vej. DK-2100 Copenhagen, Denmark Accepted for publication December 3, 1991
pyramidal cell loss in a rat model of 10 rnin global ischemia. Material and methods
The study comprises the following experiments with age and weight matched groups of male wistar rats (a total of 49 rats) weighing 300-350 g. 1) 6 rats given 30 mg/kg NBQX i.p. 15 min before (5 rnin before ischemia), 5 min before and at start of recirculation. 6 untreated 10 min ischemia rats served as controls; 2) 8 rats given 30 mg/kg NBQX i.p. at start of recirculation, 10 min after and 25 min after start of recirculation, 5 untreated 10 rnin ischemia rats served as controls; 3 ) 12 rats given 30 mg/kg NBQX i.p. at 60, 70 and 85 rnin after start of recirculation, 12 untreated 10min ischemia rats served as controls; 4) 6 rats given 5 mg/kg MK-801 i.v. at start of recirculation, 6 untreated 10 min ischemia rats served as controls. Ischemia
The day before ischemia the rats were anaesthetized with methohexital(50 mg/kg i.p.) and their vertebral arteries were electrocauterized (setting 4-6; Monopolar Martin Elektrotom) (15). They were fasted with free access to water until the next day when they were anaesthetized with a mixture of 1 Halothane,
45
Diemer et al. 33% 0, and 66% N,O. After gentle exposure of the carotids, a femoral artery and vein were cannulated with Polystan pp 25 catheters filled with Heparinized saline (100 IU/ml). 1 % Lidocaine was applied to the incisions. Blood pressure, rectal and temporal muscle temperature were monitored throughout the experiment. By means of a heating lamp the body temperature was kept at 37.5”C and head temperature (temporal muscle probe) at 36.5 “C. The halothane anesthesia was turned off and as the rats began to wake up, the carotid arteries were clamped and blood withdrawn in a 10ml syringe flushed with heparinized saline (100 IU/ml) in order to reduce the MABP to between 50 and 60 mm Hg. After 10 min the clamps were removed and the blood reinfused. Before the animals woke up from the postischemic coma, all wounds were sutured. Administration of drugs
2,3-Dihydro-6-nitro-7-sulfamoylbenzo(F)quinoxaline, NBQX ( = FG 9202, NOVO Nordisk, CNS division), was dissolved in Krebs’ buffer, 30 mg/ml, pH = 7.2 and injected intraperitoneally (30 mg/kgx 3, see below). In all NBQX groups, the compound was given as 3 injections, with the second and third 10 and 25 min later, respectively. Control rats received only buffer. MK-801 ( = dizocilpine MSD, ( + )-5-methyl-10,lldihydro-5H-dibenzo[ a,d]cyclohepten-5,10-imine maleate) was dissolved in Krebs’ buffer and given intraperitoneally ( 5 mg/kg). Control animals received Krebs’ buffer. Histological procedures
Six days after ischemia the animals were anesthetized with halothane, the thorax quickly opened and a cannula inserted into the ascending aorta. After clamping of the descending aorta perfusion with 300 ml Lillies phosphate buffered formalin was performed using a roller pump. After gentle removal of the brain it was postfixed, dehydrated, blocked and paraffin embedded using a routine schedule. From the blocks 4-5 pm semiserial frontal sections were cut and stained with hematoxylin eosine. The sections covered a 2 mm span of the dorsal hippocampus.
thickness or nuclear size were performed. Thus in a control animal 12 fields, each containing about 50 neurons were counted ( = approx 600 neurons per animal). Results
Table 1 shows the physiological parameters in the various groups. In the group of animals which received NBQX before ischemia, a slight decrease of blood pressure from 95 mm Hg to 60 mm Hg was detected. Otherwise there were no differences in the physiological parameters measured in the experimental and control groups, and especially the brain and body temperature did not differ between the treated and untreated groups. The rats which received NBQX before or immediately after ischemia showed no other deviations from control ischemic rats. Histology
In all control groups most of the CA1 pyramidal cells were eosinophilic with a pyknotic or karyorhexic nucleus. This neuronal necrosis was accompanied by a pronounced glial reaction: i.e. presence of many macrophages with rod-shaped nuclei and a slight increase in the number of astrocyte nuclei. A few mitoses, most arrested in metaphase, were seen. Hippocampus of the MK-801 treated animals did not deviate from this pattern of neurone loss and accompanying glial changes. In the brains from the NBQX treated animals hippocampus showed only damage to a fraction of the pyramidal cells which showed qualitatively the same changes as the controls. The interneurons in stratum oriens, stratum radiatum and stratum lacunosum moleculare showed in any group no eosinophilic cell changes and their numbers appeared to be within normal range. The present ischemia model does not induce consistent damage to cerebellar cortex, neocortex or striatum, and in accordance with this only a few Table 1. MABP before, during and after cerebral ischemia, plasma glucose concentration and CAI pyramidal cell loss. (Values are meanfs.d.). One way ANOVA with Bonferroni correction. * p
z,
I
’
N. H. Diemer ’, M. B. Jirgensen F. F. Johansen’, M. Sheardown’: T. Honore’
’
Cerebral Ischemia Research Group, Institute of Neuropathology, University of Copenhagen, NOVD-Nordisk. CNS Division, S ~ b o r g , Denmark
z,
After more than a few min of cerebral ischemia in the rat the hippocampal CA1 pyramidal cells are irreversibly damaged (1,2). The main excitatory input to these cells comes from the CA3 pyramidal cells (3). Transsection of the Schaffer collaterals and commisurals from CA3 before (4-6) or immediately after ischemia (4) has a protective effect. Autoradiographically the CA1 region is characterized by a high density of glutamate receptors of the NMDA (N-methyl-D-aspartate) and AMPA (a-amino-3hydroxy-5-Methyl-4-isoxazole proprionic acid) subtype (formerly named quisqualate receptor). The normal fast excitation in the SchafTer collateral synapse is mediated by glutamate and AMPA-receptors whereas the NMDA receptor seems to be involved in more excessive stimulation such as long-term potentiation (7). A number of both competitive and non-competitive antagonists to the NMDA receptor are available (8), but in vivo tests of their possible ischemia protective effect have given contradictory results (9-1 l), seemingly dependent on the intraischemic brain temperature and the amount of residual blood flow (12, 13). Blockade of the AMPA receptor with systemically injected antagonists has been made possible recently, with the development of the quiiioxaline dione NBQX (14). In gerbils this compound offered protection after administration up to 4 h after ischemia. The present paper report on the ability of NBQX to ameliorate CA1
Nils Henrik Dierner, institute of Neuropathology, University of Copenhagen, 11 Frederik V’s vej. DK-2100 Copenhagen, Denmark Accepted for publication December 3, 1991
pyramidal cell loss in a rat model of 10 rnin global ischemia. Material and methods
The study comprises the following experiments with age and weight matched groups of male wistar rats (a total of 49 rats) weighing 300-350 g. 1) 6 rats given 30 mg/kg NBQX i.p. 15 min before (5 rnin before ischemia), 5 min before and at start of recirculation. 6 untreated 10 min ischemia rats served as controls; 2) 8 rats given 30 mg/kg NBQX i.p. at start of recirculation, 10 min after and 25 min after start of recirculation, 5 untreated 10 rnin ischemia rats served as controls; 3 ) 12 rats given 30 mg/kg NBQX i.p. at 60, 70 and 85 rnin after start of recirculation, 12 untreated 10min ischemia rats served as controls; 4) 6 rats given 5 mg/kg MK-801 i.v. at start of recirculation, 6 untreated 10 min ischemia rats served as controls. Ischemia
The day before ischemia the rats were anaesthetized with methohexital(50 mg/kg i.p.) and their vertebral arteries were electrocauterized (setting 4-6; Monopolar Martin Elektrotom) (15). They were fasted with free access to water until the next day when they were anaesthetized with a mixture of 1 Halothane,
45
Diemer et al. 33% 0, and 66% N,O. After gentle exposure of the carotids, a femoral artery and vein were cannulated with Polystan pp 25 catheters filled with Heparinized saline (100 IU/ml). 1 % Lidocaine was applied to the incisions. Blood pressure, rectal and temporal muscle temperature were monitored throughout the experiment. By means of a heating lamp the body temperature was kept at 37.5”C and head temperature (temporal muscle probe) at 36.5 “C. The halothane anesthesia was turned off and as the rats began to wake up, the carotid arteries were clamped and blood withdrawn in a 10ml syringe flushed with heparinized saline (100 IU/ml) in order to reduce the MABP to between 50 and 60 mm Hg. After 10 min the clamps were removed and the blood reinfused. Before the animals woke up from the postischemic coma, all wounds were sutured. Administration of drugs
2,3-Dihydro-6-nitro-7-sulfamoylbenzo(F)quinoxaline, NBQX ( = FG 9202, NOVO Nordisk, CNS division), was dissolved in Krebs’ buffer, 30 mg/ml, pH = 7.2 and injected intraperitoneally (30 mg/kgx 3, see below). In all NBQX groups, the compound was given as 3 injections, with the second and third 10 and 25 min later, respectively. Control rats received only buffer. MK-801 ( = dizocilpine MSD, ( + )-5-methyl-10,lldihydro-5H-dibenzo[ a,d]cyclohepten-5,10-imine maleate) was dissolved in Krebs’ buffer and given intraperitoneally ( 5 mg/kg). Control animals received Krebs’ buffer. Histological procedures
Six days after ischemia the animals were anesthetized with halothane, the thorax quickly opened and a cannula inserted into the ascending aorta. After clamping of the descending aorta perfusion with 300 ml Lillies phosphate buffered formalin was performed using a roller pump. After gentle removal of the brain it was postfixed, dehydrated, blocked and paraffin embedded using a routine schedule. From the blocks 4-5 pm semiserial frontal sections were cut and stained with hematoxylin eosine. The sections covered a 2 mm span of the dorsal hippocampus.
thickness or nuclear size were performed. Thus in a control animal 12 fields, each containing about 50 neurons were counted ( = approx 600 neurons per animal). Results
Table 1 shows the physiological parameters in the various groups. In the group of animals which received NBQX before ischemia, a slight decrease of blood pressure from 95 mm Hg to 60 mm Hg was detected. Otherwise there were no differences in the physiological parameters measured in the experimental and control groups, and especially the brain and body temperature did not differ between the treated and untreated groups. The rats which received NBQX before or immediately after ischemia showed no other deviations from control ischemic rats. Histology
In all control groups most of the CA1 pyramidal cells were eosinophilic with a pyknotic or karyorhexic nucleus. This neuronal necrosis was accompanied by a pronounced glial reaction: i.e. presence of many macrophages with rod-shaped nuclei and a slight increase in the number of astrocyte nuclei. A few mitoses, most arrested in metaphase, were seen. Hippocampus of the MK-801 treated animals did not deviate from this pattern of neurone loss and accompanying glial changes. In the brains from the NBQX treated animals hippocampus showed only damage to a fraction of the pyramidal cells which showed qualitatively the same changes as the controls. The interneurons in stratum oriens, stratum radiatum and stratum lacunosum moleculare showed in any group no eosinophilic cell changes and their numbers appeared to be within normal range. The present ischemia model does not induce consistent damage to cerebellar cortex, neocortex or striatum, and in accordance with this only a few Table 1. MABP before, during and after cerebral ischemia, plasma glucose concentration and CAI pyramidal cell loss. (Values are meanfs.d.). One way ANOVA with Bonferroni correction. * p
