The Effect of MK-801 on Cortical Spreading Depression in the Penumbral Zone Following Focal Ischaemia in the Rat

Journal of Cerebral Blood Flow and Metabolism 12:371-379 1992 The International Society of Cerebral Blood Flow and Metabolism Published by Raven Press, Ltd., New York The Effect of MK-801 on Cortical Spreading Depression in the Penumbral Zone Following Focal Ischaemia in the Rat *R. Gill, P. Andine, tl. Hillered, tl. Persson, and H. Hagberg Institute of Neurobiology and Department of Obstetrics and Gynaecology, University of G6teborg, G6teborg, Sweden; *Department of Veterinary Basic Sciences, Royal Veterinary College, Royal College Street, London, England; and tdepartments of Neurosurgery and Clinical Chemistry, University of Uppsala, Uppsala, Sweden Summary: Cortical spreading depression (CSD) is a transient depression of neuronal activity that spreads across the cortical surface. In the present studies, we have investigated CSD activity in the penumbral zone following permanent middle cerebral artery (MCA) occlusion in the rat (n = 16/group), using double-barreled Ca2+ -sensitive microelectrodes. Measurements of CSD activity were made for 3 h in each animal. During this time, a varying number of spontaneous CSDs were seen in the control group (total was 30, with a range of 0-7/rat). These CSDs were of varying duration: "small" (-1 min) and "big" (5-45 min) CSDs. During a CSD, the extracellular [Ca2+] decreased to 0.11 ± 0.07 mm (mean ± SD). After 3 h, the extracellular [Ca2+] in the cortex (penumbral zone) was either normal (10/16 rats) or lowered to 0.5 mm (2/16 rats) or to 0.1 mm (4/16 rats). In the caudate nucleus (ischaemic core area), all rats had an extracellular [Ca2 +] of -0.1 mm when measured after the 3 h recording period. Neuropathological evaluation of the brains of the animals, which had been allowed to survive for 24 h after MCA occlusion, revealed ischaemic damage in the dorsolateral cortex and caudate nucleus. Administration of the noncompetitive NMDA antagonist, MK-801 (3 mglkg i.p.), 30 min after MCA occlusion resulted in 24 and 29% reductions in the volume of hemispheric and cortical damage, respectively, which was highly significant (p < 0.0001); no protection was seen against caudate damage. MK-801 also significantly (p < 0.05) reduced the number of CSDs (total was IO with a range of 0-2/rat) and the degree of decrease in extracellular [Ca2+] during CSDs (low level was 0.19 ± 0.15 mm, mean ± SD). Both the number of "small CSDs" and significantly the "big CSDs" were reduced by MK-80l. The "small CSDs" were present in the border zone and also deeper within the infarct, whereas "big CSDs" and anoxic depolarisations only occurred closer to the infarct core. CSDs were not seen outside the border zone of the infarct. In conclusion, there was a good correlation between extracellular [Ca2+] changes and ischaemic damage in the infarcted region following MCA occlusion. MK-801 reduced the ischaemic damage in the penumbral zone in parallel with attentuation of the number and amplitude of CSDs and thereby a reduced intracellular Ca2 + overload. Key Words: Cerebral ischaemia-mk-801-spreading depression-glutamate-nmda. The excitatory amino acid neurotransmitter L-glutamate is thought to be involved in the neuronal degeneration that is seen following a period of cerebral ischaemia (Meldrum, 1985; Rothman and Olney, 1986; Choi, 1988). The first link between glutamate and ischaemia was proposed by Van Received September 18. 199 1; final revision received November 28, 199 1; accepted December 4, 199 1. Address correspondence and reprint requests to Dr. R. Gill at Department of Veterinary Basic Sciences. Royal Veterinary College, Royal College Street, London NWI OTU, England. Abbreviations used: AMPA, a-amino-3-hydroxy-5- methylisoxazole-4-proprionic acid; ANOV A. analysis of variance; CSD, cortical spreading depression; MCA, middle cerebral artery; NMDA, N-methyl-D-aspartate. Harreveld (1959), who showed that cortical spreading depression (CSD) could be produced in rabbits by applying glutamate to the cortical surface and proposed that this was related to neuronal hypoxia. Further evidence then came from studies using in vivo microdialysis techniques (Benveniste et ai., 1984; Hagberg et ai., 1985) in models of transient forebrain ischaemia, which showed that there was an excessive extracellular accumulation of glutamate and other excitatory amino acids during and shortly after the ischaemic insult. These studies have recently been supplemented with dialysis studies done in animal models of focal cerebral ischaemia (Hillered et ai., 1989; Butcher et ai., 1990), 371

372 R. GILL ET AL. which have shown that there is an increase in extracellular glutamate levels immediately after the induction of ischaemia and the high levels are sustained for up to 90 min postischaemically. The glutamate is thought to act on at least two postsynaptic subtypes of ionotropic excitatory amino acid receptors: N-methyl-D-aspartate (NMDA) and non-nmda receptors (Monaghan et ai., 1989). The NMDA receptor appears to play a crucial role in ischaemia-induced neuronal degeneration because the ion channel associated with the receptor is highly permeable to Ca2+ (MacDermott et ai., 1986). Thus, depolarisation of postsynaptic membranes may result in opening of NMDA and non-nmda receptor-linked ionic channels, resulting in an accumulation of intracellular Ca2 + that is detrimental to the cell (Siesjo, 1981). The excitatory amino acid glutamate and other selective agonists such as NMDA, u-amino-3-hydroxy-5-methylisoxazole-4-proprionic acid (AMPA), and kainate can initiate CSD (Van Harreveld, 1959; Curtis and Watkins, 1961; Bures et ai., 1974; Lauritzen et ai., 1988). CSD is associated with a transient depression of neuronal activity that spreads across the cortical surface (Leao, 1944; Hansen and Zeuthen, 1981; Nicholson and Kraig, 1981), and these waves are characterised by a large negative shift in the extracellular direct current potential accompanied by a marked increase in extracellular [K +] and a decrease in extracellular [Na +], [Cl-], and importantly [Ca2+] (Hansen and Zeuthen, 1981; Hansen, 1985). The Ca2+ is thought to enter the cell, with a subsequent rise in intracellular Ca2 + from the normal level of 100 nm. CSDs, measured as transient increases in extracellular [K +], have been shown to occur in the border zone of a focal ischaemic infarct in the rat (Nedergaard and Astrup, 1986; Hansen and Nedergaard, 1988). NMDA antagonists and more recently non NMDA antagonists have been shown to be potent neuroprotective agents in models of focal cerebral ischaemia (Park et ai., 1988a,b; Bielenburg, 1989; Bullock et ai., 1990; Gill et ai., 1991b; Gill and Lodge, 1991) in which the middle cerebral artery (MCA) is permanently occluded (Tamura et ai., 1981a). These compounds have been suggested to be neuroprotective in such models by blocking the glutamate-induced excitation in the penumbral zone and thus decreasing the amount of infarction in this area. MK-801, a noncompetitive NMDA antagonist acting at the ion channel site (Wong et ai., 1986; Kemp et ai., 1987), has been shown to protect against the neuronal degeneration that results from focal cerebral ischaemia in the rat and cat (Park et ai., 1988a,b; Gill et ai., 1991b). Mk-801 has also been shown to block almost completely CSDs, as have other NMDA antagonists (Goroleva et ai., 1987; Hansen et ai., 1988). Thus, it is of interest to investigate the effect of MK-801 on the changes in extracellular [Ca2+] and direct current potential associated with CSD in the penumbral zone of a focal ischaemic lesion in the rat. In the present studies, we have measured CSD in the specific penumbral zone where MK-801 had previously been shown to be neuroprotective (Park et ai., 1988b; Gill and Lodge, 1990; Gill et ai., 1991b). A preliminary report of this work has appeared as an abstract (Gill et ai., 1991a). MATERIALS AND METHODS Calcium-sensitive microelectrodes Calcium-sensitive microelectrodes were prepared as described by Andine et al. (1988). Briefly, doublebarreled borosillicate glass capillaries (inner diameter of 0.87 mm, outer diameter of 1.5 mm, Hilgenberg GmbH, Malsfeld, Germany) were pulled into fine-tipped micropipettes using an electrode puller (Siegenbeck van Heukelorn et ai., 1976). The tips were broken to a diameter of 3-7 J.Lm. One barrel was filled with 4% tri-n-butylchlorosilane in carbon tetrachloride and heated for 5 min at 500 C; it was then filled with a 300-400 J.Lm column of a neutral carrier ionophore for calcium (Fluka AG, Buchs, Switzerland) and CaCl 2 (0.1 M) was used as an inner reference solution. The reference barrel was filled with 0.15 M NaC!. Ag/AgCl wires were used as inner reference electrodes and connected to high-input impedance amplifiers. The extracellular [Ca2+] signal was obtained by differential recording between the calcium and the reference barrel. The reference potential was recorded between the reference barrel and an external ground Ag/ AgC!-agar electrode placed on the cortical surface. The potentials were recorded with a two-channel pen writer; the slope of the Ca2+ -sensitive microelectrodes was 26-30 m V per decade change of Ca2 + concentration. Surgical procedures Male Sprague-Dawley rats (330-370 g) were used for these studies and were maintained on a 12 h: 12 h light: dark cycle with access to food and water ad libitum. The animals were anaesthetised with a mixture of 3% halothane, in 70% nitrous oxide, and 30% oxygen. They were then intubated and ventilated with a mixture of 1.5% halothane, in 70% nitrous oxide, and 30% oxygen throughout the surgical period. The left femoral artery and vein were cannulated to enable continuous monitoring of blood pressure, blood sampling, and injections. All physiological variables were monitored throughout the study, and the animals were maintained normothermic (37 ± 0.8 C) using a rectal probe and thermostatically controlled heating blanket. The left MCA was exposed through a craniotomy of 2 mm that was made using a saline-cooled dental drill. The MCA (from its stem to the lenticulostriate branches) was permanently occluded using microbipolar coagulation, as described by Park et al. (l988b). Once the MCA had been occluded, the retracted temporalis muscle was allowed to fall back into place and sutured as was the skin covering it, using silk sutures. The animals were then J Cereb Blood Flow Me/ab, Vol. 12, No.3, 1992

MK-801 AND SPREADING DEPRESSION IN ISCHAEMIA 373 placed in a stereotactic frame, and a Ca2+ -sensitive microelectrode was lowered into the cortex at coordinates from bregma of anteroposterior (AP) 0.0 mm, lateral 4.5 mm, and ventral 1.3 mm (corresponding to the stereotactic level anterior 7.19 mm in Fig. 1, Konig and Klippel, 1963). The cranial window was covered with modified Ringer solution. The time from MCA occlusion to the start of electrode recording was 17 ± 6 and 18 ± 4 min in the saline- and MK-801-treated groups, respectively. The animals were subsequently ventilated with a mixture of 0.8% halothane, in 70% nitrous oxide, 30% oxygen. Recordings were made continuously from this area for 3 h, following which the electrode was lowered in 1.0 mm steps up to ventral 5.3 mm. The animals were then removed from the frame, the skin covering the scalp was sutured using silk sutures, the femoral artery cannula was removed, and the vessel coagulated. Following this, the halothane was turned off and the animals ventilated with air until they started to breathe spontaneously, at which point they were extubated. The animals were allowed to survive for a period of 24 h, during which time they were kept normothermic using an overhead heating lamp and a heated blanket placed underneath the cage. During the recovery period, they had free access to food and water. Perfusion and fixation The following day, the animals were deeply anaesthetised with sodium pentobarbitone and perfused transcardially with 0.9% heparinised saline at 100 mm Hg until the perfusate from the right atrium ran clear. The animals were then perfused with 200 ml of a mixture of 40% formaldehyde, glacial acetic acid, and methanol (FAM 1:1:8, vlvlv; Tamura et ai., 1981a) at 100 mm Hg. The brains were removed after 24 h, processed and embedded in paraffin wax, and 8 f.lm coronal sections stained with haematoxylin-eosin. These sections were examined by light microscopy and areas of ischaemic damage were delineated at eight preselected coronal levels from anterior 10.5 mm to anterior 1.0 mm (Konig and Klippel, 1963). This was done blindly with respect to the treatment of the animals. Quantification of ischaemic damage The areas of ischaemic brain damage were transposed onto scale diagrams (x4 the actual size of the brain) of FIG. 1. The stereotactic level, anterior 7.19 mm (Konig and Klippel, 1963), illustrates the position (asterisk) where the Ca2+ electrode was positioned. forebrain and measured in terms of hemispheric, cortical, and caudate damage using an image analyzer (Quantimet 760, Cambridge instruments, Cambridge, U.K.). The areas of ischaemic damage were used to determine the total volume of ischaemic tissue in each brain (Osborne et ai., 1987), by integration of areas with the distance between each level. Experimental groups MK-801 was administered at a dose of 3 mg/kg i.p. 30 min after MCA occlusion and control animals received saline (1 mukg i.p.) 30 min after MCA occlusion. There were 16 animals in each group. Statistical analysis The number of CSDs, the amplitude, depolarisation time, and minimum [Ca2+] level du 'ing the CSD for the control vs. MK-801-treated animals was tested using the unpaired, two-tailed Students t test. The difference between the two groups with regard to the number of small or big CSDs was compared using Fischer's exact test. The differences in the volume of ischaemic damage for each brain region (hemisphere, cortex, and caudate) for the control and MK-801-treated animals was tested using analysis of variance (ANOV A) with a Bonferroni correction. The area of ischaemic damage at the different stereotactic levels was compared for the control and MK- 801-treated group using BMDP 2v (BMDP Statistical Software, Cork, Ireland) with repeated measures. This was followed by a comparison of the area of damage at each coronal plane for the two groups using the BMDP 3D multivariate t test. All data are presented as the mean ± SD for N animals. RESULTS The histopathology of the brains revealed that permanent MCA occlusion resulted in ischaemic damage in the dorsolateral cortex and caudate nucleus, with mean (±SD) volume of ischaemic damage in the hemisphere and cortex of control animals of 159.8 ± 20.7 and 115.3 ± 17.8 mm3, respectively. The MK-801-treated animals showed significant protection against hemispheric (ANOY A, F = 28.87, p < 0.0001) and cortical (ANOYA, F = 27.09, p < 0.0001) ischaemic damage (Fig. 2); MK- 801 did not protect against caudate damage (ANOYA, F = 3.58, p > 0.07; Fig. 2). There was an overall significant (ANOY A, F = 27.59, p < 0.0001) difference between the control and MK-801- treated groups when a comparison was done for the area of cortical ischaemic damage at different stereotactic levels (Fig. 3). The multivariate t test showed that there was a significant difference between the groups at the individual stereotactic levels (Hotteling, 12 = 56.1, F = 5.38,p < 0.001). The area of ischaemic damage was significantly less in the MK-801 group compared to the control group, at each coronal plane studied (anterior 10.5 mm to anterior 1.0 mm; Fig. 3). In the control group, typical waves of CSD were associated with a large negative shift in the extra- J Cereb Blood Flow Metab, Vol. 12, No.3, 1992

.. 374 R. GILL ET AL. 200 '" OJ) " EM " E E '8 '":' ",::E "..c:uj u. ",C/J :.::+ c '" " E Q) ",::E o > 160 120 80 40 * FIG. 2. Volume of ischaemic damage in the cerebral hemisphere, cerebral cortex, and caudate nucleus for saline (open bars) and MK-801 (shaded bars) (3 mg/kg i.p.)-treated animals. There was significant protection against hemispheric (*p '" 0.01) and cortical (*p '" 0.01) ischaemic damage. However, no protection was seen against caudate damage. Each bar = mean ± SO for 16 animals. 0 Hemisphere Cortex Caudate Brain Region cellular direct current potential and a decrease in extracellular [Ca2+] levels (Fig. 4). These CSD-like phenomena are referred to as "CSDs" in the present paper although the use of one electrode gives no information concerning the propagation of the observed shifts. During the 3 h period of recording, we observed spontaneous CSD transients in the cortex of most animals; these varied in amplitude and duration. However, in some animals, a direct depolarisation with low [Ca2 +] levels was seen that did not recover to baseline levels. For the control group, following 3 h of MCA occlusion, the extracellular [Ca2+] in the cortex was normal 0-1.3 mm) in 10 of the 16 animals and lower in the remaining 6, of which 2 had extracellular [Ca2+]'s of approximately 0.5 mm and 4 below 0.1 mm. The MK-SOl-treated group of animals displayed a similar distribution in their cortical extracellular [Ca2+] 02/16 animals showed normal levels, 3/16 intermediate, and 1116 had low levels). Thus, there was a significant (p < 0.04, t test) difference in terms of the minimum [Ca2+] levels seen during the CSDs for the control (0.11 ± 0.07 mm, mean ± SD) and MK-SOI (0.19 ± 0.15 mm) groups of animals (Table 1). The extracellular [Ca2 +] in the caudate was approximately 0.1 mm or below for both groups of animals, at 3 h after induction of the ischaemia. The total number of CSDs in the MK-SOI group (10 with a range of 0-2 per rat) was significantly (p 0.05) less than that seen in the control group (30 with a range of 0-7 per rat, Table 1). MK-SOI also significantly (p 0.05) lowered the amplitude of the CSD from 0.S9 ± 0.07 to O.SI ± 0.15 mm (mean ± SD). The duration of CSDs was variable in different animals. In some animals, the CSDs lasted from seconds up to 5 min, and we termed these "small CSDs" (Fig. 4A). In other animals, CSDs lasted for up to 45 min (Fig. 4B) and we named any CSDs of longer than 5 min duration "big CSDs." A signifi- 20 FIG. 3. The area of cortical ischaemic damage at different stereotactic levels (from anterior, 10.5 mm to anterior, 1.0 mm) for the saline (squares) and MK-801 (circles) (3 mg/kg i.p.) treated animals. There was a significant (BMOP 2v, ANOVA; F = 27.59, P < 0.0001) difference in the area of ischaemic damage for the MK-801 group compared to saline-treated controls at the different stereotactic levels. There was a significant (*p < 0.05; p < 0.01) reduction in the ischaemic area for the MK-801 group at each stereotactic level compared to the control group. The arrow indicates the stereotactic level at which the Ca2+ electrode was positioned. Each point represents the mean ± SO for 16 animals. '" OJ) " EN " Cl E u '8 '" ci. "C/J -8 +1 '" c.. " '- 0 Q),, ::E '" -< 10 0 12 * * 10 + 8 6 7.19 Stereotactic Level 4 2 0 J Cereb Blood Flow Metab, Vol. 12, No.3, 1992

MK-80i AND SPREADING DEPRESSION in ischaemia 375 FIG. 4. Examples of typical changes in extracellular calcium concentration ([Ca2+1ec) and direct current potential seen after MCA occlusion. A: "Small" cortical spreading depressions (CSDs) characterised by a transient negative shift in the extracellular direct current potential accompanied with a decrease in [Ca2+1ec and a duration lasting between 30 s and 5 min. B: "Big CSDs" were of longer duration (5--45 min). C: In some animals, the shifts in direct current potential and [Ca2+1ec did not recover and therefore appeared to have the characteristics of an anoxic depolarisation. Note the different time scale in B. [AI [BI [C] «:: [V-'l 1// -1!l (2g; V) _ J\ tr--\ J\J --'- ---.J, - 5 min 10 min 5 min candy (p < 0.04, Fisher's two-tailed exact probability test, Table 2) higher proportion of control animals (seven) had "big CSDs" compared to the MK-801 (one)-treated group. These "big CSDs" were seen 1.25 to 2.5 mm within the infarct rim for the control group, and the one "big CSD" seen in the MK-801-treated animal was 1.75 mm within the infarct. There were also more "small CSDs" in the control (23) group compared to the MK-801 (9) treated group of animals. "Small CSDs" were seen from 0.25 mm outside the infarct to 2.25 mm within the infarct for both groups of animals. In seven rats from both groups of animals, the microelectrode tip was located outside the infarct and during the aggregate total 21 h period in which CSDs were measured for these animals, only one "small CSD" was detected 1 mm outside the infarct. Therefore, the remainder of CSDs, of which there were 39 measured over a period of 75 h, in the other animals were recorded within the infarct. The position of the calcium electrode was 0.4 ± 0.94 mm from the rim of the infarct in the MK-801 group of animals and it was 1.2 ± 0.82 mm from the infarct edge in the control group. Thus, there was a significant (p < 0.02, Student's unpaired t test) difference in the position of the electrode from the infarct rim for the two groups of animals. A few animals did show a depolarisation from which the ionic homeostasis did not recover (Fig. 4C). There was no significant (p > 0.4) difference between the MK-801 and control group in terms of the number of these anoxic depolarisations (Table 1). No significant (p > 0.08) difference was found between the groups for the direct current shifts during CSD (Table 1). Nor was there any difference (p > 0.6) between the groups with respect to % depolarisation time during the experiment, when compared with or excluding the animals that had anoxic depolarisations upon electrode penetration (Table 1). The % depolarisation represents the % of the experimental time (3 h) during which the tissue was depolarised. The tissue was regarded as depolarised after a rapid negative direct current shift (including CSDs and irreversible depolarisations). The physiological variables such as ph, blood glucose, P ac0 2' and P a0 2 (Table 3) did not differ significantly between the control and MK-801- treated animals. However, there was a significant difference between the MABP for the control group (100 mm Hg) and MK-801-treated animals (107 mm Hg) when compared prior to MCA occlusion (ANOVA, F = 7.8, p < 0.01); the MABP was 7 mm Hg higher in the MK-801 group (Table 3). This was not due to drug effects because the MK-801 was actually administered 30 min after MCA occlusion. The rectal temperature of the animals was maintained at 37 ± 0.8 C throughout the experimental period. TABLE 1. The different variables measured for the transient cortical spreading depression activity Numbers in Saline MK-S0 1 Statistics Parameter comparison (1 mllkg) (3 mg/kg) (P) Number of small CSDs 16 VS. 16 rats 23 9 p > 0.9 Number of big CSDs 16 VS. 16 rats 7 1 p 0.04 Total number of CSDs 16 VS. 16 rats 30 10 p < 0.05* Anoxic depolarisation 16 VS. 16 rats 6 4 p > 0.4 Depolarisation time (%) 16 VS. 16 rats 31.5 ± 39.5 25. 1 ± 37.0 p > 0.6 Depolarisation time %a 14 VS. 14 rats 21.7 ± 31.S 14.9 ± 25.S p > 0.5 Min. [Ca2+] during CSD (mm) 30 VS. 10 CSDs 0. 11 ± 0.07 0. 19 ± 0. 15 p < 0.04* Direct current shift during CSD (m V) 30 VS. 10 CSDs 23.2 ± S.36 IS.0 ± 7. 16 p > O.OS The depolarisation time is the % of the experimental time (3 h) during which the tissue was depolarised. * p < 0.05, unpaired, two-tailed, Student's t test; '*Fisher's exact probability test, two tailed. Each value represents the mean ± SD. a This is the depolarisation time (%) excluding the animals that had a final CSD upon electrode penetration. J Cereb Blood Flow Me/ab, Vol. 12, No.3, 1992

376 R. GILL ET AL. TABLE 2. Comparison of the control and MK-801-treated animals in terms of the "big" CSDs Number of rats with "big CSDs" Number of rats with no "big CSDs" MK-801 Control (3 mg/kg) (n = 16) (n = 16) 7 9 I 15 (p = 0.04) The number of animals that had "big CSDs" in each group were compared. n = number of animals in each group. The number in parentheses denotes the significance value when compared using Fisher's exact probability test (two tailed). DISCUSSION The present studies have shown that CSDs of varying amplitude and duration could be seen in the cortical infarct rim. These CSDs were characterised by a large negative shift in the extracellular direct current potential and a decrease in extracellular [Ca2 + ]. The noncompetitive NMDA antagonist MK-801 (3 mglkg i.p.), administered 30 min after MCA occlusion, resulted in a significant reduction in the number and amplitude of CSDs in parallel with a significant reduction in the area of cortical damage. There are now numerous studies showing the beneficial effects of NMDA antagonists in animal models of focal ischaemia (Park et ai., 1988a,b; Bielenburg, 1989; Bullock et ai., 1990; Gill and Lodge, 1990; Gill et ai., 1991b). The majority of the protection is seen in the infarct rim of the cerebral cortex, in a region that has the characteristics of a "penumbra" (Astrup et ai., 1981). This area has a regional blood flow of approximately 20% of normal (Tamura et ai., 1981b; Nedergaard et ai., 1986; Bolander et ai., 1989), which is the threshold level at which synaptic transmission is abolished, but normal ionic homeostasis is maintained. The penumbra displays an increased glucose utilization probably TABLE 3. Physiological variables for cortical spreading depression study Time of Physiological MK-801 measurement variables Control (3 mg/kg) Pre-MCA MABP (mm Hg) 99.6 ± 9.9 107.0 ± 5.0' Pre-MCA Glucose (mmoi/l) 11.l±3.5 10.9 ± 2.1 Pre-MCA ph 7.46 ± 0.05 7.47 ± 0.04 Pre-MCA PaC02 (kpa) 5.0 ± 0.8 5.3 ± 0.6 Pre-MCA Pa02 (kpa) 16.4 ± 3.4 15.7 ± 3.8 Post-MCA MABP(mm Hg) 84.7 ± 15.9 87.8 ± 12.1 Post-MCA Glucose (mmoi/l) 8.7 ± 2.0 7.8 ± 1.3 Post-MCA ph 7.44 ± 0.06 7.45 ± 0.04 Post-MCA PaC02 (kpa) 5.0 ± 0.6 5.4 ± 0.9 Post-MCA Pa02 (kpa) 17.6 ± 2.8 17.7 ± 2.5 The physiological variables were measured prior to occlusion of the MCA and again during the 3 h of recording CSD activity after MCA occlusion. Each value is the mean ± SD of 13-16 animals. p < 0.01, Bonferroni t distribution. due to an increased demand on ionic pumps to maintain ionic homeostasis (Nedergaard and Astrup, 1986). In contrast, no neuroprotection is seen with NMDA antagonists in the caudate nucleus, which could be described as the ischaemic core; blood flow here is reduced to < 10% of normal. Moreover, at the end of the 3 h recording period, the extracellular [Ca2 + ] in the caudate (0.13 mm) was lower than that seen in the cortex, and we saw no CSD activity when the electrode was lowered into the caudate. This can be explained by the fact that the blood flow is very low in this area, due to occlusion of the lenticulostriate branch (Tamura et ai., 1981b), which is an end artery for the caudate and there is not sufficient energy present to enable a return to normal ionic homeostasis, even in the presence of an NMDA antagonist (Siesjo and Bengtsson, 1989). The presence of CSD activity in the infarct rim, as measured by changes in extracellular direct current potential, has previously been reported by Nedergaard and Astrup (1986). They demonstrated that there was a clear correlation between CSD activity and increased glucose utilisation, possibly due to restoration of ionic homeostasis. They also found CSDs of varying amplitude and duration with the maximal duration being up to 15 min. During their 80 min period of measurement in each animal, they recorded a mean of five direct current potential deflections (3.8/h), i.e., considerably more than in the present study (0.53/h). This is surprising considering the similar blood glucose levels in the two studies. The different frequency of CSDs may be related to the more supemcial position of their electrodes or a different extent of the infarct. Another explanation may be that the recordings were started 17 min after MCA occlusion in the present study, compared to 2.5 min in the study by Nedergaard and Astrup (1986). However, the latter explanation seems unlikely since the frequency of CSD activity in each animal appears to be similar during the first hours after MCA occlusion. CSD activity as measured by increases in extracellular [K + ] and decreases in extracellular [Ca2 + ] have also been shown to occur in focal ischaemia induced in baboons (Branston et ai., 1977; Harris et ai., 1981) and cats (Strong et ai., 1983). Brain slice preparations have also been used to investigate spreading depression and the resulting ionic changes (Balestrino and Somjen, 1986; Somjen et ai., 1990). Glutamate and other excitatory amino acid agonists can evoke CSDs (Van Harreveld, 1959; Curtis and Watkins, 1961; Bures et ai., 1974; Lauritzen et ai., 1988) that can be blocked by competitive and noncompetitive NMDA antagonists (Goroleva et J Cereb Blood Flow Metab. Vol. 12, No.3. 1992

MK-801 AND SPREADING DEPRESSION IN ISCHAEMIA 377 ai., 1987; Hansen et ai., 1988; Lauritzen et ai., 1988). Therefore, the increased levels of glutamate in the cortex and striatum during focal ischaemia (Hillered et ai., 1989; Butcher et ai., 1990) could be initiating CSDs in the penumbra. Additionally, the penumbral zone is hypoperfused and hypoglycemic; these factors will lower the threshold for eliciting CSDs (Nedergaard, 1988). The microelectrode tip was situated deeper within the infarct in the control group of animals compared to the MK-801-treated group, which had significantly less cortical damage. The CSD activity may only be a phenomenon in certain layers within the developing infarct without any pathological importance. Then, the effect of MK-801 on CSDs could be explained by a different position of the electrode in relation to the infarct. However, there is now substantial evidence that suggests that penumbral CSD activity is detrimental to the cell. An important fact is the CSD is associated with a decrease in extracellular [Ca 2 +] and consequently an increase in intracellular [Ca 2 +]. CSDs per se do not actually cause damage to the normal brain. When waves of CSD were elicited repeatedly in nonischaemic animals, no signs of cortical neuronal injury were seen (Nedergaard and Hansen, 1988). Therefore, under normal conditions, the brain ATP levels are not altered during spreading depression (Lauritzen et ai., 1990), thus enabling neurones to extrude Ca 2 + and maintain normal ionic homeostasis. This is also reflected in the fact that in nonischaemic animals there is an increase in glucose utilisation prior to the ionic shifts characteristic of CSD. This is probably due to the increased demands on ion pumps because of the disturbed ionic homeostasis (Shinohara et al., 1979). Under ischaemic conditions such as in the penumbra, energy supply is severely compromised due to reduced blood flow during the first 6 h after MCA occlusion (Bolander et ai., 1989). Therefore, under such conditions, CSD activity may result in an accumulation of intracellular Ca 2 + and, consequently, cell death. In addition to reducing the penumbral damage, MK- 801 also reduced the number of CSDs and the degree of [Ca 2 +] shift during CSD without any effect on the direct current potential shifts. This suggests that MK-801 protected the penumbra by blocking NMDA receptors, thus limiting the elicitation of CSDs and reducing the amount of Ca 2 + uptake into cells during the CSDs that did appear. The CSDs observed in the present study displayed a large variation in both amplitude and duration. The "small CSDs" were the same as those previously described, i.e., the classical spreading depression lasting a minute or slightly longer. The "big CSDs" had different characteristics in terms of their duration and lasted up to 45 min. It could be speculated that these CSDs were prolonged by the low energy levels and hypoglycaemic state of the tissue (Nedergaard, 1988). These "big CSDs," although taking a long time, were always spontaneously reversible unlike an anoxic depolarisation. MK-801 significantly attenuated elicitation of the "big CSDs"; this also supports their being CSDs since MK-801 has previously been shown to block evoked CSDs but not the anoxic depolarisation seen following forebrain ischaemia (Hansen et ai., 1988). The complexity of CSDs is further exemplified by the fact that these CSDs that are evoked in the cortex by bipolar stimulation are completely blocked by MK-801 whereas only a partial effect was seen in the ischaemic penumbra in the present study. Also, under conditions of severe hypoglycaemia, MK-801 was unable to block the CSDs (Zhang et al., 1990) in spite of its ability to ameliorate the hypoglycaemia-induced neuronal degeneration (Westerberg et ai., 1988). In a hypoxic hippocampal slice preparation, Radar and Lanthorn (1989) demonstrated that persistent depolarisation could be blocked by MK-801 but not the initial depolarisation. Using a similar preparation, Somjen et al. (1990) suggested that selective vulnerability of CA 1 neurones of the hippocampus may be related to spreading depression-like depolarisation. The "small CSDs" appeared in the border zone and deeper within the infarct whereas the "big CSDs" and anoxic depolarisations only occurred deeper within the infarct. No CSD activity was found outside the infarct (except one "small CSD"). However, we can not exclude the possibility of high CSD activity in more superficial cortical layers (cf. Nedergaard and Astrup, 1986). During an aggregate total 21 h period of measurements (in seven animals) from outside the infarct, only one "small CSD" was found in comparison to the 39 CSDs recorded during the 75 h period of measurements (in 25 animals) within the developing infarct. Thus, CSD activity predicted that the tissue would degenerate. The great variability in the CSD activity between different rats may be explained by the variability of the MCA occlusion model with regard to neuropathological outcome. For example, in some animals, a final depolarisation without recovery was seen upon lowering the electrode into the cortex. This probable anoxic depolarisation could not be reversed by MK-801 (which was administered after the start of the depolarisation), and in these rats the cortex also displayed massive damage. In conclusion, CSD activity, as measured by J Cereb Blood Flow Metab, Vol. 12, No.3, 1992

378 R. GILL ET AL. changes in extracellular [Ca 2 +] and negative direct current potential, were seen in the penumbral area of the cerebral cortex of animals subjected to focal ischaemia. MK-801 attenuated the number and amplitude of the CSDs and also reduced the volume of cortical infarction. Thus, a reduction in the amount of CSD activity in the penumbral area, and consequently attenuation of the calcium overload upon the cells, may be one of the mechanisms by which MK-801 reduces ischaemic damage in the penumbral zone. Acknowledgment: This work was supported by grants from the Swedish Medical Research Council (12x-09455), Anna Ahrenberg Foundation, Magnus Bergwall Foundation, the 1987 Foundation for Stroke Research, Ragnar and Torsten SOderberg Foundation, the King Gustav V Foundation, Tore Nilsson Foundation, Ake Wiberg Foundation, Ahlen Foundation, the Swedish Medical Society, the Swedish Society for Medical Research, and the Odd Fellow Research Fund. The authors would like to thank Prof. A. Hamberger for his financial help and enthusiasm. We would also like to acknowledge the help and enthusiasm of Prof. D. Lodge from the Royal Veterinary College of London. R. Gill is being sponsored by Eli Lilly (Indianapolis, IN, U.S.A.). REFERENCES Andine P, Jacobson I, Hagberg H (1988) Calcium uptake evoked by electrical stimulation is enhanced post-ischaemically and precedes delayed neuronal death in CAl of rat hippocampus: Involvement of N-methyl-D-aspartate receptors. J Cereb Blood Flow Metab 8:799-807 Astrup J, Siesj6 BK, Symon L (198 1) Thresholds in cerebral ischemia-the ischemic penumbra. Stroke 12:723-725 Balestrino M, Somjen GG (1986) Chlorpromazine protects brain tissue in hypoxia by delaying spreading depression-mediated calcium influx. Brain Res 385:2 19-226 Benveniste H, Drejer J, Schousboe A, Diemer NH (1984) Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischaemia monitored by intracerebral microdialysis. J Neurochern 43: 1369-1374 Bielenburg GW (1989) Infarct reduction by NMDA antagonists in a rat stroke model. In: Pharmacology of Cerebral Ischaemia (Krieglstein J, ed), Boca Raton, FL, CRC Press Inc., pp 239-242 Bolander HG, Persson L., Hillered L, d'argy R, Ponten U, Olsson Y (1989) Regional cerebral blood flow and histopathologic changes after middle cerebral artery occlusion in rats. Stroke 20:930--937 Branston NM, Strong AJ, Symon L (1977) Extracellular potassium activity, evoked potential and tissue blood flow. Relationships during progressive ischaemia in baboon cerebral cortex. J Neurol Sci 32:305-32 1 Bullock R, Graham DI, Chen MH, Lowe D, McCulloch J (1990) Focal cerebral ischaemia in the cat: Pretreatment with a competitive NMDA receptor antagonist, D-CPPene. J Cereb Blood Flow Metab 10:668-674 Bures J, Buresova 0, Krivanek J (1974) The Mechanism and Applications of Leao's Spreading Depression of Electroencephalographic Activity. New York, Academic Press Butcher SP, Bullock R, Graham DI, McCulloch J (1990) Correlation between amino acid release and neuropathologic outcome in rat brain following middle cerebral artery occlusion. Stroke 21: 1727-1733 Choi DW (1988) Glutamate neurotoxicity and diseases of the nervous system. Neuron 1 :623--634 Curtis DR, Watkins JC (196 1) Analogues of glutamic and gammaaminobutyric acids having potent actions on mammalian neurones. Nature (Lond) 19 1: 10 10-- 10 11 Gill R, Andine P, Hillered L, Persson L, Hagberg H (199 Ia) The effect of MK-80 1 on cortical spreading depression in the penumbral zone following focal ischaemia in the rat. J Cereb Blood Flow Metab li(suppl 2):S286 Gill R, Brazell C, Woodruff GN, Kemp JA (l99 1b) The neuroprotective action of dizocilpine (MK-80 1) in the rat middle cerebral artery occlusion model of focal ischaemia. Br J Pharmacal 103:2030--2036 Gill R, Lodge D (1990) Investigation of the neuroprotective actions of MK-80 1 and kynurenic acid in a rat stroke model. Neurochem Int 16:40 Gill R, Lodge D (199 1) The neuroprotective action of 2,3- dihydroxy-6-nitro-7-sulfamoylbenzo(f)quinoxaline (NBQX) in a rat focal ischaemia model. J Cereb Blood Flow Metab II(suppI2):S224 Goroleva NA, Koroleva VI, Amemori T, Pavlik V, Bures J (1987) Ketamine blockade of cortical spreading depression in rats. Electroencephalogr Clin Neurophysiol 66:440-447 Hagberg H, Lehmann A, Sandberg M, Nystrom B, Jacobson I, Hamberger A (1985) Ischaemia-induced shift of inhibitory and excitatory amino acids from intra- to extracellular compartments. J Cereb Blood Flow Metab 5:413-419 Hansen AJ (1985) Effect of anoxia on ion distribution in the brain. Physiol Rev 65: 10 1-148 Hansen AJ, Lauritzen M, Wieloch T (1988) NMDA antagonists block cortical spreading depression but not anoxic de polarisation. In: Frontiers in Excitatory Amino Acid Research (Calvalheiro EA, Lehmann J, Turski L, eds), New York, Alan R. Liss, pp 66 1--666 Hansen AJ, Nedergaard M (1988) Brain ion homeostasis in cerebral ischemia. Neurochem Pathol 9: 195-209 Hansen AJ, Zeuthen T (198 1) Changes of brain extracellular ions during spreading depression and ischemia in rats. Acta Physiol Scand 113:437--445 Harris RJ, Symon L, Branston NM, Bayhan M (198 1) Changes in extracellular calcium activity in cerebral ischemia. J Cereb Blood Flow Metab 1:203-209 Hillered L, Hallstrom A, Segersvard S, Persson L, Ungerstedt U (1989) Dynamics of extracellular metabolites in the striatum after middle cerebral artery occlusion in the rat monitored by intracerebral microdialysis. J Cereb Blood Flow Metab 9:607--616 Kemp JA, Foster AC, Wong EHF (1987) Non-competitive antagonists of excitatory amino acid receptors. Trends Neurol Sci 10:294-298 Konig JFR, Klippel RA (1963) The Rat Brain: A Stereotaxic Atlas of the Forebrain and Lower Parts of the Brain Stem. New York, Krieger Lauritzen M, Rice ME, Okada Y, Nicholson C (1988) Quisqualate, kainate and NMDA can initiate spreading depression in the turtle cerebellum. Brain Res 475:317-327 Lauritzen M, Hansen AJ, Kronborg D, Wieloch T (1990) Cortical spreading depression is associated with arachidonic acid accumulation and preservation of energy charge. J Cereb Blood Flow Metab 10: 115-122 Leao AAP (1944) Spreading depression of activity in the cerebral cortex. J Neurophysiol 7:359-390 MacDermott AB, Mayer ML, Westbrook GL, Smith SJ, Barker JL (1986) NMDA-receptor activation increases cytoplasmic calcium concentration in cultured spinal cord neurons. Nature (Lond) 321:5 19-522 Meldrum BS (1985) Possible therapeutic applications of antagonists of excitatory amino acid neurotransmitters. Clin Sci 68: 1 13-122 Monaghan DT, Bridges RJ, Cotman CW (1989) The excitatory amino acid receptors: Their classes, pharmacology, and distinct properties in the function of the central nervous system. Annu Rev Pharmacol Toxicol 29:365-402 J Cereb Blood Flow Metab, Vol. 12, No.3, 1992

MK-801 AND SPREADING DEPRESSION IN ISCHAEMIA 379 Nedergaard M (1988) Mechanisms of brain damage in focal cerebral ischaemia. Acta Neurol Scand 77:81-101 Nedergaard M, Astrup J (1986) Infarct rim: Effect of hyperglycemia on direct current potential and [14C12-deoxyglucose phosphorylation. J Cereb Blood Flow Metab 6:607-615 Nedergaard M, Gjedde A, Diemer NH (1986) Focal cerebral ischaemia of the rat brain: Autoradiographic determination of cerebral glucose utilization, glucose content and blood flow. J Cereb Blood Flow Metab 6:414-424 Nedergaard M, Hansen AJ (1988) Spreading depression is not associated with neuronal injury in the normal brain. Brain Res 449:395-398 Nicholson C, Kraig RP (1981) The behavior of extracellular ions during spreading depression. In: The application of ion selective electrodes (Zeuthen T, ed) Amsterdam, Elsevier, pp 217-238 Osborne KA, Shigeno T, Balarsky AM, Ford I, McCulloch J, Teasdale GM, Graham DI (1987) Quantitative assessment of early brain damage in a rat model of focal cerebral ischaemia. J Neurol Neurosurg Psychiatry 50:402-410 Park CK, Nehls DG, Graham DI, Teasdale GM, McCulloch J (l988a) Focal cerebral ischaemia in the cat: Treatment with the glutamate antagonist MK-801 after induction of ischaemia. J Cereb Blood Flow Metab 8:757-762 Park CK, Nehls DG, Graham DI, Teasdale GM, McCulloch J (l988b) The glutamate antagonist MK-801 reduces focal ischaemic brain damage in the rat. Ann Neurol 24:543-554 Rader RK, Lanthorn TH (1989) Experimental ischemia induces a persistent depolarization blocked by decreased calcium and NMDA antagonists. Neurosci Lett 99: 125-130 Rothman S, Olney JW (1986) Glutamate and the pathophysiology of hypoxic-ischaemic brain damage. Ann Neurol19: 105-111 Shinohara M, Dollinger B, Brown G, Rapoport S, Sokoloff L (1979) Cerebral glucose utilization: Local changes during and after recovery from spreading cortical depression. Science 203: 188-190 Siegenbeek van Heukelom J, Dijkstra K, Van Ingen H (1976) A versatile and reproducibly operating microelectrode puller. Med Bioi Eng 14:644-652 Siesj6 BK (1981) Cell damage in the brain: A speculative synthesis. J Cereb Blood Flow Metab 1:155-185 Siesj6 BK, Bengtsson F (1989) Calcium fluxes, calcium antagonists, and calcium-related pathology in brain ischemia, hypoglycemia, and spreading depression: A unifying hypothesis. J Cereb Blood Flow Metab 9:127-140 Somjen GG, Aitken PG, Balestrino M, Herreras 0, Kawasaki K (1990) Spreading depression-like depolarization and selective vulnerability of neurones. Stroke 21(suppl 111):179-183 Strong AJ, Venables GS, Gibson G (1983) The cortical ischemic penumbra associated with occlusion of the middle cerebral artery in the cat. I. Topography of changes in blood flow, potassium ion activity, and EEG. J Cereb Blood Flow Metab 3:86-96 Tamura A, Graham DI, McCulloch J, Teasdale GM (l98ia) Focal cerebral ischaemia in the rat: I. Description of technique and early neuropathological consequences following middle cerebral artery occlusion. J Cereb Blood Flow Metab 1:53-60 Tamura A, Graham DI, McCulloch J, Teasdale GM (1981b) Focal cerebral ischaemia in the rat: 2. Regional cerebral blood flow determined by C4C) iodoantipyrine autoradiography following middle cerebral artery occlusion. J Cereb Blood Flow Metab 1:61-81 Van Harreveld A (1959) Compounds in brain extracts causing spreading depression of cerebral cortical activity and contraction of crustacean muscle. J Neurochem 3:300-315 Westerberg E, Kehr J, Ungerstedt U, Wieloch T (1988) The non-competitive NMDA-antagonist MK-801 reduces extracellular amino acid levels during hypoglycemia and protects against hypoglycemic damage in rat caudate nucleus. Neurosci Res Commun 3: 151-158 Wong EHF, Kemp JA, Priestley T, Knight AR, Woodruff GN, Iversen LL (1986) The anticonvulsant MK-801 is a potent N-methyl-D-aspartate antagonist. Proc Nat! Acad Sci USA 83:7104-710 Zhang E, Hansen AJ, Wieloch T, Lauritzen M (1990) Influence of MK-801 on brain extracellular calcium and potassium activities in severe hypoglycemia. J Cereb Blood Flow Metab 10: 136-139 J Cereb Blood Flow Metab, Vol. 12, No.3. 1992