- Email: [email protected]
The competitive NMDA receptor antagonist CGP 40116 protects against status epilepticus-induced neuronal damage
RESEARCH Epilepsy
Research
17
(1994)207.--2 19
The competitive NMDA receptor antagonist CCP 40116 protects against status epilepticus-induced neuronal damage Demon G. Fujikawa”$b*f, Allan H. Danids”, John S. Kim” “Experimenral Neurolog_vLaboratory, Sepcdvrdu VA Medical Cenier. 161II Plummer Street. Sepulveda, CA 91343. USA. hL3epartmenf of Neurology and &win Researcl~Insiitute. UCLA School qf Medicine, Los Angeles, CA 90024, USA (Received
3 August
1993; revision
received 28 September 1993;accepted 20 October 1993)
Abstract
We studied the efftcacy of the competitive NMDA receptor antagonist CGP 40116 in protecting against seizureinduced neuronal necrosis from Iithium-pilocarpine-induced status epilepticus (SE). Rats were given CGP 40116 either before SE (I 2 mg/kg i.p.) or 15 min after the onset of SE (4, 12 and 24 mgjkg); controls received normal saline 15 min after SE began, Diazepam and phenobarbitai were given i.p. after 3 h of SE to stop the seizures. Rats were killed 24 h later, and their brains were processed for light microscopic examination. Neuronal damage occurred in 24 of 25 brain regions examined in saline-injected animals, Protection was maximal in rats given 12 and 24 mg/kg CGP 40116 after SE onset: 19 and 21 of the 24 damaged regions were protected respectively, but the 24 mg/kg group had a mortality rate comparable to saline-injected controls. No necrotic neurons were found in posterior cinguiate and retrosplenial neurons at the two highest CGP 40116 doses, suggesting that the transient cytoplasmic vacuolization induced by NMDA receptor antagonists does not progress to frank necrosis. In rats given CGP 40116 seizure discharges were not eliminated, but their amplitudes were signi~~antly reduced 2 h after SE began. The periodic epileptiform discharge (PED) EEG pattern, probably a sign of widespread neuronal damage, developed in saline-injected controls after 2-2.5 h of SE but not in rats given 12 and 24 mg/kg of CGP 40116. CGP 40116 provided widespread protection against seizureinduced neuronal necrosis, suggesting that an esssential step in its production is NMDA receptor activation by endogenous glutamate. The neuroprotection provided was not simply an antiepileptic effect, since electrographic seizures persisted despite NMDA receptor blockade. CGP 40116 and NMDA receptor antagonists in general could be useful as adjunctive neuroprotectants in patients with refractory SE. Key wortfs: NMDA receptor antagonist; ~...
--
- .~
.-.-
CGP 40116; Status epilepticus: Neuronal necrosis --
1. Introduction Seizures induced in the rat by the muscarinic cholinergic agonist pilocarpine or by the combina* Corresponding 8 1X-895-9554.
author.
Tel.: 818-891-7711,
ext. 7610; Fax:
0920- 12 11iY4/$7.00 .(:. 1994 Elsevier Science B.V. All rights reserved SSRl 0920-121 1(93)E0089-Z
._--
_~
.-._-,_.-___
__..
_... _.“...__...
tion of lithium and pilocarpine, which reduces the dosage of pilocarpine required about 13-fold, produce widespread neuronal necrosis, especially to limbic structures [3,13,33]. Behaviorally, the seizures resemble kindled seizures, and can be classified by the S-stage system used to describe the latter [28]. Because limbic structures are involved,
these seizures are commonly referred to as limbic seizures, to differentiate them from generalized convulsive or nonconvulsive seizures. The anticholinergic drug atropine has no effect once seizures have begun, suggesting spread of seizure activity beyond the cholinergic system [ 13,151, Based upon the excitotoxic action of exogenously administered t-glutamate and its anaiogues, it has been postulated that excessive presynaptic release of L-glutamate produces seizure-induced neuronal necrosis [23]. Activation of the Nmethyl-D-aspartate (NMDA) subtype of glutamate receptor postsynaptically opens the receptor-operated cation channel within the receptor, permitting calcium influx and increasing the intracellular free calcium concentration [16]. It is currently hypothesized that any pathological process (such as status epilepticus (SE)) which increases glutamate release activates an increased number of NMDA receptors for a critical period of time, leading to neuronal necrosis by elevating [Ca2+li and activating potentially destructive CaZ ’ -dependent enzymes f21]. The noncompetitive NMDA receptor antagonists such as ketamine and MK801, which block calcium entry by binding to a site within or at the mouth of the cation channel [19], protect against neuronal damage from kainate-, lithium-pilocarpineand soman-induced SE despite ongoing electrographic seizures [4,9.10,31]. This suggests that the neuroprotection provided is not simply an antiepileptic effect and that seizures can persist despite NMDA receptor blockade. In contrast to the noncompetitive antagonists, there are no data on the effect of systemically administered competitive NMDA receptor antagonists, which bind to the agonist recognition site on the receptor, in protecting against seizure-induced neuronal damage. For this reason we decided to test the effect of a new competitive NMDA receptor antagonist. I>-(E)-2-amino-4acid (CGP methyl-5-phosphono-3-pentenoic 40116; Ciba-Geigy Ltd., Base], Switzerland). Its racemate (CGP 37849) is effective in suppressing electroshock seizures in rats: after i.v. pretreatment, the greatest protection was obtained 2 and 4 h later (ED50 = 4.3 mgjkg), whereas i.p. pretreatment was most effective after 4 h (ED50 = 2.8 mgi kg) [30]. CGP 40116, the most potent competitive
antagonist to date, shows little or no specific binding to other receptors at pharmacological concentrations [8]. In contrast, the noncompetitive NMDA receptor antagonists ketamine and MKX01 also act at other receptors and ion channels [7.12,29.36,37]. By using CGP 401 I6 wc determined how effective NMDA receptor blockade is in ameliorating the neuronal damage and eiectrographic seizure activity caused by lithium-pilocarpine-induced SE. This by inference suggests to what extent activation of the NMDA receptor by endogenous glutamate is responsible for the damage and the EEG seizure discharges.
2. Methods 2. I. Materials
Pilocarpine HCl, atropine methyl bromide (methylatropine) and lithium chloride were purchased from Sigma Chemical Company. The pilocarpine and methylatropine were dissolved in 0.9% sodium chloride and the lithium chloride was dissolved in distilled water prior to injection (200 mg/ml, 100 mg/ml and 127 mg/ml respectively). Pentobarbital sodium (50 mg/ml), ketamine HCI (50 mg/ml), diazepam (5 mg/ml) and phenobarbital sodium (130 mg/ml) were obtained from the Pharmacy Service at Sepulveda VA Medical Center. CGP 40116 was generously provided by Ciba-Geigy Ltd., Base& Switzerland. It was dissolved in 0.9% sodium chloride and titrated with 0.1 M NaOH to a pH of 4-S. Bouin’s solution for brain fixation was prepared in standard fashion
[181. After the i.p. administration of pentobarbital, 40 mg/kg, and ketamine, 25 mg/kg, maie Wistar rats (200-450 g) had three stainless steel screws implanted into their skulls for later recording of electroencephalographic (EEG) activity. Two screws on either side of the midline anterior to the coronal sutures served as active electrodes, and a screw anterior to the lambdoidal suture to the right of midline was used as a ground electrode. Four to six days later, rats were given lithium
D.G. Fujikawu et al.!Epilepsy Research 17 IlYY4) 207-219
chloride, 3 mEy/kg i.p., and the following day they had continuous monitoring of frontal EEG activity during a 20-min baseline period, following pilocarpine injection, during 3-h seizures and the first 2 h of a 24-h recovery period. In addition, rectal temperatures were taken and behavioral observations were made every 15 min. Following baseline measurements, rats were then given methylatropine, 10 mg/kg i.p., to prevent systemic muscarinic side effects, followed by pilocarpine, 30 mg/kg i.p. If no seizures occurred within 30 min, a second 30 mg/kg dose of pilocarpine was given. Rectal temperatures during seizures and the first 2 h of recovery were kept between 37.0 and 39.O”C. Rats were placed on a bed of ice covered with a cloth towel if their temperatures rose to 38.6”C; rectal probes were attached to a temperature controller which automatically switched on a heat lamp if temperatures ‘dropped to 37.5”C, at which point they were taken off the ice. In order to minimize variations in the duration of SE and the degree of neuronal damage in a given brain region from rat to rat, its onset was defined as the time that continuous behavioral and electrographic seizures began, and antiepileptic drugs (diazepam, 10 mg/kg and phenobarbital, 25 mg/kg i.p.) were given 3 h later. In addition, the muscarinic cholinergic antagonist atropine, which crosses the blood-brain barrier, was given i.p. (1 mgikg) with the antiepileptic drugs to lessen the cholinergic component of the seizures. Three hours was chosen because neuronal injury is found in all but one of the 25 brain regions at that time point (manuscript submitted for publication; see also the data on the saline-injected controls in Table 3) and longer seizures were not deemed necessary. CGP 40116, 12 mg/kg i.p., was given at the time of the first pilocarpine injection in one group of rats (n = 6). and 15 min after SE onset in groups given 4 mg/kg (n = 5), 12 mg/kg (n = 5) and 24 mg/ kg (n=5) i.p. A control group was given normal saline i.p. 15 min after SE onset (n = 5). Fifteen minutes after SE onset corresponded to about 20 min after the first seizure and about SO min after pilocarpine injection (Table 2). This time point was chosen because it is prior to the time that neuronal necrosis first appears in this seizure mod-
209
el [1 I] (manuscript submitted for publication), and it was our intention to study the neuroprotective and antiepileptic effects of CGP 40116 during ongoing SE but before neuronal damage occurred. The amplitude of EEG activity was measured prior to SE during the baseline period, and during 3-h SE the maximal amplitudes and frequencies of EEG seizure discharges were measured and rats’ behavior was recorded every 15 min. Foliowing diazepam and phenobarbital administration EEC seizure activity was also measured and behavioral observations were recorded every 15 min for 2 h. EEG seizure discharges persisted during the 3h observation period in all groups Twenty-four hours after 3-h seizures, EEG and behavioral observations were made, then rats were deeply anesthetized with 400 mg/kg pentobarbital i.p. and underwent transcardiac perfusion-fixation with heparinized Ringer’s lactate solution followed by 200 ml of Bouin’s solution, at a perfusion pressure of 120 mm Hg. Brains were left in situ for 1-3 h, following which they were removed and put in Bouin’s solution overnight. The following morning brains were transferred to 70% ethanol; they were kept in 70% ethanol until further processing. Each brain was then cut coronally into 2---3mm thick blocks, dehydrated in graded ethanol solutions, cleared with xylene, embedded in paraffin, sectioned at 6 pm thickness, deparaffnized and hydrated to water, and stained with hematoyxlin and eosin (H&E). Sections at the level of the caudate-putamen, dorsal hippocampus and ventral hippocampus (corresponding to 0.20 mm, 3.30 mm and 5.60 mm from bregma respectively [27]) were examined to determine the extent of neuronal necrosis in 25 brain regions, which were selected because of their vulnerability to seizure-induced neuronal damage. One section at each level was used to rate the degree of damage, and several adjacent sections at each level were checked to insure that the section used was representative. Frontotemporoparietal and cingulate cortices were examined at the level of the caudate-putamen, retrosplenial cortex at the levels of the dorsal and ventral hippocampi, and thalamic nuclei at the dorsal hippocampal level. The 25 brain regions were rated for the extent
of irreversible neuronal damage according to the following scale: grade 0= no damage, grade 0.5 = < 10% acidophilic neurons (slight damage), grade I .O = IO-25% (mild damage), grade 1.5 = 26645%, grade 2.0 = 46-54% (moderate dagrade 2.5 = 55575% and grade mage), 3.0 = > 75% acidophihc neurons (severe damage). Acidophihc neuronal cytoplasm and nuclear hyperchromatism (pyknosis) were used as the criteria of irreversible neuronal damage [ 14,383. Actual cell counts of acidophilic and normal neurons were done at 200 x magnification in the dorsal and ventral dentate hilus (dorsal and ventral CA4) and in dorsal CA1 (data not shown); counts from both hemispheres were added together, and the combined percentage of acidophilic neurons was calculated for each rat. In the 22 remaining brain regions, visual estimates of the percentage of acidophilic neurons present were made and a histological damage score was assigned because quantitative cell counts would have been impractical given the vast numbers of neurons in each of the regions. The histological damage scores of the two hemispheres were averaged for each rat. In the dentate hilus and dorsal CA1 (data not shown), the visual estimates correlated very well with actual cell counts, which were done subsequently (Tables 3, 4 and 5). Grade 2 damage was associated with at least mild edema and grade 3 damage with severe edema and a disrupted, pale neuropil. 2.3. Statistical analysis In order to determine if the histological damage scores were normally distributed, thereby permitting standard parametric methods of data analy(individual scores less group sis, ‘residuals’ means) were plotted for a representative sample of brain regions. Normal (Gaussian) curves were approximated, so one-way analysis of variance (ANOVA) was used to determine if differences in histological scores existed among the five groups of rats in the 25 brain regions examined. If the overall P value for the F-ratio obtained for a given brain region was 60.05, post-hoc comparisons between groups were made with Fisher’s least significant difference (LSD) test, using pooled SEMs and CI=0.05. One-way ANOVA
with post-hoc Fisher’s LSD test when indicated was also used to compare hilar (CA4) neuronal cell counts in dorsal and ventral dentate gyrus and to analyze seizure and SE latency data and EEG seizure discharge amplitude and frequency data.
3. Results 3.1. Survival data Table 1 summarizes the survival data in salineinjected controls and CGP 40116-injected rats. In this particular group of controls, only seven of 17 (41%) developed SE; of the seven with SE, two (29%) died during seizures. All of the CGP 40116-injected rats developed SE, and four of 26 (15%) died during SE or the first 2 h of the recovery period. Deaths were confined to the groups given 24 mg/kg CGP 40116 (3/9, or 33%) and 12 mg/kg (l/6, or 17%) after SE onset. In addition, one of the nine rats became apneic after it was given 24 mg/kg; it was resuscitated and eventually recovered but was not included in the EEG and histological analyses. The reason that so few saline-injected controls developed SE is not clear: 80% of the rats which did not develop SE received the higher (60 mg/kg) dose of pilocarpine. Moreover, 77% of the CGP 40116-injected rats were given the drug after SE had begun so it could not have been responsible for the 100%
Table I Survival data Group
SE
Saline-injected controls CGP 40116 before SE onset
s
IO
2
17
12 mg/kg CGP 401 I6 after SE onset
6
0
0
6
s 5 6”
0 0 0
0 I ?
5 6 9
4 mgikg 12 mg/kg 24 mg/kg
~.__
No SE .~~._
Died _
Total
~~-~
SE = the number of rats with status epilepticus; no SE = the number of rats which did not develop SE despite 60 mgikg of pilocarpine. “One of the rats became apneic after it was given CGP 40116; it was resuscitated and eventually recovered, but was not included in the EEG and histological analyses.
rate of seizure induction
in the CGP 40116 groups.
3.2. EEG and behavioral observations The time intervals between the first pilocarpine injection and the first seizure and between the first injection and the onset of SE did not vary signiticantly among the five groups tested (Table 2). Four of the five (80%) saline-injected SE controls received 30 mg/kg of pilocarpine; one (20%) got 60 mg/kg. Of the CGP 40116-injected rats, four of 21 (19%) received 30 mg/kg; the remaining 17 (81%) got 60 mg/kg. The CGP 40116 was not responsible for the high percentage requiring 60 mg/kg, since 15 of the 21 (71%) were given CGP 40116 after seizures began. The dose of pilocarpine given had no effect on the latency either to the first seizure or to SE.
Table 2 Times from pilocarpine injection onset of status epilepticus
to the first seizure and to the -
Group
Latency to first seizure (min)
Latency to SE (min)
Saline-injected controls (5)
34+10
68+ 8
57+11
74&15
39* 6 39+ 7 3S& 5
48+ 41+ 43+
CGP 40116 given before the onset of SE I? mg:kg (6) CGP 40116 given after the onset of SE 4 mg:kg (5) I2 m&kg (5) 24 mg,‘kg (5)
7 6 5
The numbers in parentheses refer to the number of rats in each group. Latency to iirst seizure (mean + SEM) refers to the time from the first pilocarpine injectmn to the appearance of the first seizure. Latency to SE refers to the time from the frst pilocarpine injection to the appearance of continuous seizures. or status epilepticus (SE). No differences between groups were found wth respect to either measure.
A
B
BASELINE
1000uv
15-MIN SE (saline i.pJ
BASELINE
15-MIN SE (CGP 401 ‘I6 12 mg/kg i.p.)
1000 pv
1 KE
l-HOUR
SE
1000 jlv
?-HOUR
SE
J-HOUR
SE
\
1 set
3-HOUR
SE
1000
pv L-
1 set
--+----~----2-HOUR
RECOVERY
1000
-2-HOUR
uv
RECOVERY
I_-1 WE
Fig, 1. (A) This is the frontal EEG activity during baseline, status epilepticus (SE) and the early recovery period in a rat given normal saline i.p. 15 min after the onset of SE. (B) This EEG is from a rat given CGP 40116 (12 mg/kg i.p.) 15 min after SE onset at the same time periods as in A. EEG seizure discharges persisted for 3 h in both rats, although the 3-h SE shows a periodic epileptiform discharge (PED) pattern in the saline-injected control and a frequent spike discharge pattern in the CGP 401 l6-injected rat.
Noncompetitive NMDA receptor antagonists have been given before convulsant drug administration [4,9,26,3 11, 5 mm after the first EEG spikes [31], 10 [26,35], 15 [lo] or 40 min [26] after the onset of continuous EEG spikes. or 2 h after 90 min of continuous electrical stimulation of the hippocampus [2]. In this study CGP 40116 was given either at the time of the first pilocarpine injection or 15 min after the onset of continuous electrographic and behavioral seizure activity (SE), which was about 20 min after the first seizure and about 60 min after the first pilocarpine injection was given (Table 2). EEG seizure discharges (Fig. 1) and clonic jerks, often progressing to rearing and falling and corresponding behaviorally to kindled seizure stages 3-5 [28], persisted in the saline-injected controls until atropine (1 mg/kg), diazepam (10 mg/ kg) and phenobarbital (25 mg/kg) were injected i.p. after 3 h of continuous seizures. Although EEG seizure discharges were not eliminated in those rats given CGP 40116, discharge amplitudes were significantly reduced 2 h after SE onset in all
of the rats given CGP 40116 (Fig. 2A). The amplitudes of seizure discharges did not decline in saline-injected seizure animals. The mean seizure discharge amplitudes of the group given CGP 40116 prior to seizure onset were consistently lower than those of the other groups from I5 to 180 min of SE, and those of the saline-injected control group were consistently higher than those of the CGP 40116 groups from 90 to 180 min of SE, although these differences were only statistically significant at 120 min of SE (Fig. 2A). There were significant differences between groups in the frequencies of EEG seizure discharges 15 and 165 min after SE began (Fig. 2B). Discharge frequencies were higher after 15min SE in the rats given 4 and 12 mg/kg CGP 40 I 16 than in those given saline at that time point; the EEG recordings were obtained just before CGP 40 I 16 or saline injections. The groups given 12 and 24 mg/kg after 15 min of SE had higher discharge frequencies than the saline-injected group 165 min after SE began. The seizure discharge frequencies of the saline-
a
5 3 2
2000
2 si
1600
3 %
1200
7 6
z E
800
z cc
400
t: w
Oi 0
30
60
90
120
150
DURATION OF STATUS EPILEPTICUS(MIN)
180
0
30
I
I
60
90
i--,-~-I
120
150
180
DURATION OF STATUS EPILEPTICUS (MIN)
Fig. 2. Saline-injected control group (n = 5, q), group given 12 mg/kg CGP 40116 prior to SE onset (n = 6, n ), group given 4 mg/kg CGP 40116 15 min after SE onset (n = 5, l ), group given I2 mg/kg CGP 401 I6 after SE onset (n = 5, A) and group given 24 mg/kg CGP 40116 after SE onset (n = 5, v), Values represent mean f SEM; standard error bars were omitted for clarity. The arrow indicates when CGP 40116 was given in the latter three groups, (A) EEG seizure discharge amplitudes in each of the tive seizure groups during the 3-h SE period. ‘P
injected control group progressively declined, and 120-l 50 min after the onset of SE to the end of the 3-h seizure period a periodic epileptiform discharge (PED) pattern was seen, in which low-frequency epileptiform bursts occurred on a flat background (Figs. 1A and 2B). Groups given CGP 40116 had higher seizure discharge frequencies (statistically significant only at 15 and 165 min of SE) and higher voltage EEG background activity and did not develop PEDs. Behavioral seizures were abolished in rats given 12 and 24 mg/kg CGP 40116, which lay prone on their abdomens with limbs flexed or occasionally
spread laterally. No clonic jerks were seen, but fine body tremors were often present. Diazepam and phenobarbital decreased the amplitude and frequency of PEDs in the saline-injected control group (to 227$67 pV, 0.4fO.l Hz (mean+ SEM)) at 2 h of recovery; in the 2-h recovery EEG shown in Fig. 1A the PEDs were of such low voltage that they were not visible at the amplifier gain shown. In general, by 2 h of recovery seizure discharges in the CGP 40116-treated rats were replaced by a 1 Hz or less burst suppression pattern, with bursts consisting of 20-30 Hz rhythmic activity lasting up to 800 ms. In the 2-h recov-
Fig. 3. These are photomicrographs of the ventral CAI pyramidal cell layer (A and B) and the medial nucleus of the amygdala (C and D) in the same rats as in Fig. I. The rat given saline i.p. (Fig. IA) is shown in A and C. and the one given CGP 401 16 (Fig. I B) is shown in B and D (H and E. x 132; scale bar = 40 /lrn). In contrast to the extensive neuronal necrosis (arrows) and edema (asterisks) in A and C. the neurons and neuropil in B and D are normal in appearance.
Histological
damage
scores
Brain region
Dorsal hippocampus CA1 CA2 CA3 Ventral hippocampus CA1 CA2 CA3 Dorsal dentate gyrus Dentate granule cells Hilar region Ventral dentate gyrus Dentate granule cells Hilar region Amygdala Piriform cortex Entorhinal cortex Thalamus Mediodorsal nucleus Lateroposterior Laterodorsal nucleus Centromedian nucleus Reuniens nucleus Cerebral cortex Frontoparietotemporal Cingulate Retrosplenial Septal nuclei Caudatejputamen Substantia nigra Pars compacta Pars reticulata
Saline-injected
CGP 40116 Before SE
15 min after SE onset
12 mg/kg
4 mg/kg
12w&g
24 mgjkg
0.5kO.2 0.2*0.1 0.7+0.1
0.1 &o.l* o.o+o.o* 0.5*0.1
0.1_+0.1* o.o+o.o* 0.6iO.2
o.o~o.o* 0.0 * 0.0 0.2+0.1*
0.0 * o.o* 0.0 * o.o* 0.2*0.1*
2.2kO.l 2.OkO.2 I .4* 0. I
0.9*0.3* 0.8 +0.4* 0.6f0.3*
I .o f 0.4* 1.6kO.5 0.9+0.3
0.2fo.l* o.o+o.o*~*** 0.3;0.1*
0.2+0.1*.*** o.o+o.o*~*** 0.3+0.1*
0.6+0.1 2.4kO.O
0.0 * o.o* 2.OkO.4
0.2*0.1* 1.g+o.3
o.o*o.o* 2.4+O.l
o.o*o.o* 2.1&O.?
0.5*0.1 2.6LO.l 1.5+0.2 2.6kO.3 2.OkO.4
0.0 * o.o* 2.0 5 0.4 0.6iO.l* 2.1 TO.4 0.9*0.2*
0.1 *o.i* 2.2kO.4 0.5*0.2* 2.5kO.5 1.610.4
o.o*o.o* 2.6kO.i 0.2*0.1* 0.6+0.0*‘**‘*** 0.6 Itr O.O*‘***
o.o~o.o* 2.2kO.l 0.2+0.1* 0.5+0.0*‘**‘*** 0.5 *o.o*.***
1.2*0.1 l.2iO.3 1.3+0.4 I .7 * 0.4 2.1 io.3
0.4+0.2* 0.1 fo.l*‘*** o.o+o.o* o.g*o.5 o.g+o.4*
0.4*0.1* 1.0+0.5 0.6+0.2* 1.0+0.4 1.5io.4
o.o*o.o* o.o~o.o*~*** o.o+o.o* 0.6+0.2 0.2+0.1*~***
o.o+o.o* 0.1+0.1* 0.2+0.1* 0.2+0.1* 0.3+0.1*,***
0.5*0.1 0.5+0.0 0.4f0.1 I .o * 0.2 1.2kO.3
0.0 +o.o*,*** o.o*o.o*,*** 0.0 +O.O*‘*** 0.5:0.2* 0.3&0.2*
0.6kO.l 0.2+0.1* 0.2*0.1 0.7io.3 0.6iO.3
0.3i:o.1**,*** o.o*o.o*~*** 0.1 kO.1’ O.O+O.O*~**~*** 0.3iO.l’
0.2*0.
0.0 * 0.0 2.2kO.5
o.o*o.o 1.0*0.5
0.0 f 0.0 1.9+0.4
o.o*o.o 2.OkO.3
o.o+o.o 1.9kO.5
I *.**,*** o.o+O.O*~*** o.o+o.o*~*** 0.0+o.o*‘*** 0.2+0.1*
SE = status epilepticus. The data represent grades of histological damage (mean f SEM) in the live 3-h seizure groups (grade 0, no damage; grade I, mild; grade 2, moderate; grade 3, severe damage; see Methods for details). One-way ANOVA was followed by posthoc Fisher least significant difference comparisons using pooled SEMs and SDS (see Methods for details). *P
ery EEG in Fig. 1B some of the low-voltage bursts are visible. Behavioral seizures stopped in the saline-injected group, and both saline- and CGP 40 116-injected animals lay in whatever position they were placed. Twenty-four hours later saline-injected control
rats had EEGs showing low-voltage PEDs (4541139 pV, 0.8 +0.2 Hz) on a flat background; the CGP 40116 groups had low-voltage seizure discharges with otherwise unremarkable EEGs (means f SEMs for discharge amplitudes were between 243 + 127 and 354 + 139 pV, and there were
between 5 * 2 and 9 &-2 discharges every 10 s). The saline-injected rats were quiet, sitting with backs hunched and with little movement. The group given CGP 40116 prior to SE behaved normally. The behavior of the other rats given CGP 40116 varied from normal to those with slow movements with head jerks in the 4 and 12 mg/kg groups; three of four in the 24 mg/kg group moved by sliding around on their abdomens and one of the four appeared ataxic while walking. The CGP 40116 behavioral changes would probably have normalized had the survival period been extended beyond 24 h. Rectal temperatures were monitored before, during and following seizures and did not change (data not shown). 3.3. Histotogicaf results CGP 40116 provided protection against SE-induced neuronal damage when given either before or 15 min after the onset of SE (Fig. 3 and Table 3). The number of protected brain regions increased as the CGP 40116 dose increased: at 4 mg/kg nine of 24 damaged regions were protected, at 12 mg/kg before and after SE 18 and 19 regions respectively were protected, and at 24 mg/ kg 21 regions were protected (Table 3). When the groups given 12 mg/kg and 24 mg/kg after SE onset were compared, no differences were found between them. However, as mentioned previously, the 24 mg/kg group had a mortality rate comparable to seizure controls, and apnea occurred after CGP 40116 injection in one of nine (Table 1). Giving 12 mg/kg CGP 40 116 after seizure onset provided comparable protection to giving it prior to seizures. A comparison of the two groups showed that two brain regions had less damage and one region had more damage in the group given CGP 40116 after SE began (Table 3). Four brain regions had greater damage in the group given 4 mg/kg CGP 40116 when it was compared to the one given 12 mg/kg before SE. This increased to eight and nine regions when the 4 mg/ kg group was compared to the groups given 12 and 24 mg/kg respectively after the onset of SE. Posterior cingu~ate and retrosplenial cortices were examined to determine if the NMDA receptor antagonist-induced transient cytoplasmic va-
Table 4 Dorsal hiiar (CA4)
~--
neurons
_----
AN
5414 Saline-injected SE controls CGP 40116 before SE onset SO-i_12 12 mg/kg i.p. CGP 40116 1S-min after SE onset 34k8 4 mg/kg i.p. 4514 12 mg;kg i.p. 4112 24 mg/kg i.p.
NN
O/oAN --
32+3
63&4
43&X
52111
35&h 29_+2
4819 61_t4 5224 38i4 _.---l.-
SE = status epilepticus, AN = number of acidophilic neurons, NN = number of normal-appearing neurons, % AN = percentage of the total number of neurons (AN + NN) which are acidophilic. Values represent mean k SEM. The percentages of acidophilic neurons in each group correspond closely to the histological damage scores in Table 3. No differences between groups were found.
cuolization noted by Olney and colleagues [24,25] progresses to neuronal necrosis. The degree of neuronal necrosis was always less than that which occurred in frontoparietotemporal cortex (layers 2-3 were always involved, and occasionally layers 4-6), and there were often no damaged neurons even when they appeared in frontoparietotemporal cortex (Table 3). At the 12 mg/kg dose of CGP 40 116 given before SE and the 24 mgjkg dose, no necrotic neurons were seen in any of the cerebral cortical regions examined, and at the 12 mg/kg dose given after SE began there were no necrotic neurons in cingulate cortex despite slight cortical damage elsewhere (Table 3). Table 5 Ventral hilar (CA4) neurons _ Saline-injected SE controls 8ii8 CGP 40116 before SE onset 12mg: kg i.p. ho+ 13 CGP 40116 I5 min after SE anset 4 mgikg i.p. 63114 12 mg;kg i.p. 68&J 24 mg:kg i.p. 56+S
30*4
-73*2
52*i2
53jll
43+13 30*2 48,5
60~12 69+2 54+3 -.“-
SE = status epilepticus, AN = number of a~idophiiic neurons, NN = number of normal-appearing neurons. % AN = percentage of the total number of neurons (AN + NN) which are acidophilic. Values represent mean + SEM. As in Table 4, the percentages of a~idophilic neurons in each group corresponded closely to the histological damage scores in Table 3 and there were no differences between groups,
In the hilus of the hippocampal dentate gyrus (CA4), where cell counts were performed, there were no differences in the number and percentage of acidophilic neurons between saline-injected rats and any of the groups given CGP 40116 (Tables 4 and 5). In addition to the dentate hilus. the only other damaged brain region not protected by CGP 401 16 was substantia nigra pars reticulata (Table 3). Within the pars reticulata, well circumscribed, oval-shaped lesions of variable size were found. They were characterized by a pale, disrupted neuropil, necrosis of all neurons and a variable degree of astrocytic necrosis. The adjacent substantia nigra pars compacta was normal in the saline-injected seizure group as well as all of the CGP 40 I I6 groups.
4. Discussion 4.1. The neuroprotective efikcts of’ CGP 40116 and noncompetitive NMDA receptor antagonists To our knowledge, this is the first study to show that a parenterally administered competitive NMDA receptor antagonist protects against SEinduced neuronal necrosis. The protection offered by the competitive NMDA receptor antagonist CGP 40116 against seizure-induced neuronal damage, even when it is given 15 min after the onset of SE, suggests that seizure-induced activation of this receptor participates in producing the damage. Maximal protection was provided by 12 mg/kg CGP 40116 (19 of 24 damaged regions) and 24 mg/kg (21 of 24) given 15 min after seizure onset (Table 3). When these groups were compared to each other, no differences in the degree of protection were found in any brain region. No additional benefit was provided by the higher dose, so the 12 mg/kg dose can be considered the most effective, especially since the group given 24 mg/kg had a mortality rate comparable to seizure controls (Table I). Giving 12 mg/kg CGP 401 I6 before SE onset offered no advantage compared to the same dose given after SE onset. Despite the protection provided by 12 mg/kg CGP 15 min after SE onset, five regions were not significantly improved. Two of these five regions had only slight damage, and one of the two was
compared to a region which itself had only been slightly damaged (Table 3). In the other three rcgions the moderate-to-severe neuronal damage seen in both the dorsal and ventral hilus and in substantia nigra pars reticulata was not amehorated by CGP 40116 (Tables 35). It is possible that in the latter three regions the seizure-induced neuronal damage was not mediated by the NMDA receptor. The well circumscribed, ovalshaped lesions within the substantia nigra pars reticulata are similar to those described in rats 1 week after flurothyl-induced SE [22], except that in the current study macrophage infiltration was not seen, probably because the recovery period was shorter. The npncompetitive NMDA receptor antagonists MK-801 and ketamine protect against SEinduced neuronal damage despite ongoing electrographic seizures, like CGP 40116 [4,9,10,31]. However, ketamine and MK-801 have pharmacological actions other than at the NMDA receptor so their neuroprotective effect [7.12,29,36.37], may not be due strictly to NMDA receptor antagonism. When given 15 min after the onset of SE, ketamine (100 mg/kg i.p.) protected against seizure-induced neuronal damage in 23 of the 24 brain regions damaged by lithium-pilocarpine SE (dorsal CA2, the one unprotected region, was only minimally damaged in saline-injected controls) [lo] (manuscript submitted for publication). In the current study 19 of the 24 damaged regions were protected by 12 mgikg CGP 40116 given 15 min after SE onset in the same seizure model. Neuronal damage may not have been NMDA receptor-mediated in the three regions which showed persistent moderate-to-severe damage (dorsal and ventral hippocampal hilar regions and substantia nigra pars reticulata). Except for these three regions, the neuroprotective effects of CGP 40116 and ketamine were comparable. 4.2. The antiepileptic effixts qf’ CGP 40116 und noncompetitive NMDA receptor antagonists Early reports on parenterally administered competitive NMDA receptor antagonists focussed on their anticonvulsant properties in rodent seizure models [5,6]. Only the behavioral manifestations of seizures. which were prevented
by pre-administration of the antagonist, were studied. It was later found that these antagonists inhibited seizure discharges in the hippocampal slice [34]. However, studies with the noncompetitive antagonists ketamine and MK-801 [4,9,10,31] and this study with CGP 40116 have shown that after parenteral injection of these agents electrographic seizure discharges may persist, even though the behavioral manifestations of seizures are suppressed. Although electrographic seizure discharges persisted in this study, their amplitudes were significantly reduced compared to controls by 2 h after the onset of SE, and the PED pattern which evolved in saline-injected control rats at 2-2.5 h of SE did not occur in the rats given CGP 40116 (Fig. 2A,B). Similar results have been reported for MK-801 [26,32]. The development of PEDs as a late stage in SE has been described in both experimentally induced 117,321 and human SE 1321. Our data support the view that the PED pattern is an EEG marker of widespread neuronal damage 1351. The higher seizure discharge frequencies at 15 min of SE in the groups given 4 and 12 mglkg CGP 40116 compared to those given saline at that time point (Fig. 2B) are not the result of the paradoxical enhancement of electrographic seizure discharges which may occur with ketamine or MK-801 [2,9]. Since the electrographic record was obtained just prior to CGP 40116 injection, the differences could not have been due to CGP 40116 and are probably fortuitous. On the other hand, the higher discharge frequencies noted at 165 min after SE onset in the groups given 12 and 24 mg/kg after SE onset compared to controls may be analogous to the rapid continuous spike activity produced by MK-801 in the lithium-pilocarpine seizure model 1351. Walton and Treiman [RS] contend that this EEG pattern and a lack of progression to PEDs late in SE is a good prognostic sign. Our histological data support this concept. Ketamine and MK-801 also induced increased seizure discharge frequencies in SE produced by continuous hippocampal stimulation, but increased discharge frequency was found in only one of four rats given the competitive NMDA receptor antagonist 3(2-carhoxypiperazine-4-yl)propyl-I-phosphonic acid (CPP), and
only in short bursts with CGP 40116.
[2], in contrast
to our results
Olney and colleagues have reported that both competitive and noncompetitive NMDA receptor antagonists produce transient cytoplasmic vacuolization in rat posterior cingulate and retrosplenial neurons, which they postulate may be responsible for the psychomimetic effects of these agents [24,25]. However, these antagonists do not in general produce lethal damage: at a dose of 5 mgikg MK-80I i.p. (live times that initially reported), the overwhelming majority of cingulate cortical neurons were normal 48 h after MK-801 administration; necrotic neurons were rarely seen [I]. In the current study. no necrotic neurons were seen in cerebral cortex (including posterior cingulate and retrosplenial cortices) in rats given 12 mg’kg CGP 40116 before and those given 24 mg/kg after seizure onset. During SE or other neurological emergencies. the risk of producing transient cingulate neuronal vacuolization and a possible transient psychosis by giving NMDA receptor antagonists would be far outweighed by the potential benefit of preventing widespread neuronal necrosis and permanent neuropsychological seyuelae.
In the lithium-pilocarpine seizure model in adult rats, neuronal necrosis first appeared 40 min after the onset of SE in cerebral cortex and atnygdala and became more widespread after I It of SE [ 11J (lnanuscript submitted for publication). Investigators studying llurothyl-induced SE in the rat also found neuronal damage within I h in several brain regions [l4,22]. In adolescent baboons with bicuculline-induced SE neuronal damage appeared within 1.5 h of SE onset [20]. The implications of these animal studies for the treatment of human SE are that prolonged seizures damage the brain and that persistent or repetitive seizures should be stopped as soon as possible. Heretofore the therapeutic emphasis has been on using drugs which either stop ongoing behavioral and electrographic seizures or prevent their recurrence. Since
NMDA receptor antagonists protect against SEinduced neuronal damage, they could be useful adjunctive agents in the treatment of refractory SE in humans. However, further studies are needed, especially with respect to the margin of safety between the most efficacious and toxic doses of CGP 401 16, Criteria for clinical effkacy must also be developed before human trials are considered.
5. Acknowledgements This research was supported by Ciba-Geigy Ltd., Basel, Switzerland, and by the Research Service of the Department of Veterans Affairs. The assistance of Dr. Jeffrey Gornbein of the UCLA Department of Biomathematics Statistical/Biomathematical Consulting Clinic in the statistical analysis of the data is gratefully acknowledged.
6. References III Allen. H.L. and Iversen, L.L., Phencyclidine,
dizocilpine, and cerebrocortical neurons, Science, 247 (1990) 22 I PI Bertram, E.H. and Lothman. E.W., NMDA receptor antagonists and limbic status epilepticus: a comparison with standard anticonvulsants. Epilrps~ Rrs., 5 (1990) I71 I X4. PI Clifford, D.B.. Olnry, J.W., Maniotis, A., Collins, R.C. and Zorumski. C.F.. The functional anatomy and pathology of lithium-pilocdrpine and high-dose pilocarpine seizures, Neumcimv, 23 (1987) 953-968. [41Clifford D.B., Olney, J.W., Benz, A.M.. Fuller. T.A. and Zorumski. C.F., Ketamine, phencyclidine and MK-801 protect against kainic acid-induced seizure-related brain damage, E~ilqxiu, 3 I ( 1990) 382- 390. M.J., Collins. J.F. and Meldrum, B.S.. AntiISICroucher. convulsant action of excitatory amino acid antagonists. S&ncP. 216 (1982) x99-901. against 161 Czuczwar. S.J. and Meldrum, B.. Protection chemically induced seizures by 2-amino-7_phosphonoheptanoic acid. ,%:ur.J. Phurmuco/.. X3 (1982) 335-338. [7] Ereciliska. M.. Nelson, D. and Silver. I.A., Interactions of benztropine. atropine and ketamine with veratridineactivated sodium channels: effects on membrane depolartLation, K * -efflux and neurotransmitter amino acid release. Br, J. ~~~ur~~ff~~)~., 94 ( 1988) U-881. [X] Fag, G.E.. Olpe, H.-R., Pozza. M.F. et al., CGP 37849 and CGP 39551: novel and potent competitive N-methylI)-aspartate receptor antagonists with oral activity. Br. J. Pkort?tucol. ( 99 (I 990) 79 I 797.
Faricllo, R.G.. Golden, <;.‘I ., Smith. G.G. and Rcycb. P.F.. Potentiation of kaimc acid epileptogenicity and q~aring from neuronal damage by an NMDA rcceptut antagonist, Epipikeps~~Rev.. 3 (1989) 206 213. Fujikawa. D.G., Ketamine given after the onset of limblc status epilepticus protects against seizure-induced neuronal damage. t~~~;~~l~.~jff. 31 ( 1990) 635. Fujikawa, D.G.. Kim. J.S. and Daniels, A.H.. When and where does neuronal necrosis first appear in the lithiumpilocarpine sel7ure model? f$ik,/>.?icr. 33. Suppl. 3 (1492) 36. Gage. P.W. and Robertson, B.. Prolongation of inhibitory postsynaptic currents by pentobarbitone, halothane and ketamine in CA1 pyramidal cells in rat hippocampus, IS?. J. Pfwmwol., 85 (1985) 675 68 I. Honchar. M.P.. Olney, 3.W. and Sherman, W.R.. Systemic eholinergic agents induce seizures and brain damage in lithium-treated rats, Sci~rz~e. 220 (1983) 323~ 325. 1141 Ingvar. M., Morgan, P.F. and Auer, R.N., The nature and timing of excitotoxic neuronal necrosis in the cerebral cortex, hlppocampus and thalamus due to flurothylinduced status epilepticus. A(,rtr !Veuropurhol. I Bed. ). 75 (1988) 362 369. [IS] Jope, R.S., Morrisett, R.A. and Snead. O.C., Characteristlcs of lithium potentiation of pilocarpine-induced status epilepticus in rats, E.qI. Il’curoi., 91 (I 986) 471 480. [16] Lodge. D. and Coliinridge. G.. Lrs qgcgrnrs~r(~~~~~~~~~~~.~~ 3 series on the pharmacology of excitatory amino acids. Trcwt.v Phurmuc~oi. Sci.. I I (1990) 22-24. [I 71 Lothman. E.W., Bertram, E.H.. Beckenstein, J.W. and Perlin, J.B.. Self-sustaining limbic status epilepticus Induced by ‘continuous’ hippocampal stimulation: electrographic and behavioral characteristics. Epikpsy Rcs., 3 (1989) 107 119. Siujnj~r~ ~(,1~7~~[~.s of’ rhc 1181Luna. L.G.. :~ut~~~ll fff ~1.~~~~~~~~~~ Armd Forcr.r r~z,~tjt~t~~ of ~~t~?(~~~/~.~. McGraw-Hill. New York. NY, 1972. 258 pp. J.F. and Nowak, L.M., Mechanisms of 1191 MacDonald, blockade of excitatory amino acid receptor channels. 7’ren& Phurmat ol. Sci.. 1I ( 1990) 167 172. B.S. and Brierley, J.B., Prolonged epileptic [201 Meldrum. seizures in primates: ischemic cell change and its relation to octal phystological events, Arch. ,Ve~rol., 28 (1973) IO -I 7. PII Meldrum, B. and Garthwaite, 3.. Excitatory ammo acid neLirotoxicity and lleurode~encrati~~e disease. Trt*nrl,s P/?unnotvri. Sri., 11 (1990) 379 3x7. PI Ncvander. G.. Ingvar. M.. Auer. R. and Siesjii. B.K.. Status cpilepticus in well-oxygenated rats causes neuronal nccrosib. Alltr. ~Ywrol.. 18 ( 198.5) 281 290. transmitters and epilepsy-related [23] Olney. J.W.. Excitatory brain damage, It,/. Kav. h’crtrohid., 27 (1986) 337 362. P41 Oiney. J.W., Labruyere, J. and Price, M.T., Pathological changes induced in cerebrocortical neurons by phencyclidint and related drugs, .S’&nc~. 344 (1989) 1360 1362. J.. Wang, G., Wozniak, D.F.. P51 Olney. J.W.. Labruyere, Price. M.T. and Sesma, M.A.. NMDA antagonist ncurotoxicity: mechanism and prevention, Scirncc,. 254
(1991) 1515~1518. [26] Ormandy, G.C.. Jope, R.S. and Snead, O.C., Anticonvulsant actions of MK-801 on the lithium-pilocarpine model of status epilepticus in rats. E-up. Neural., 106 (1989) 1722180. [27] Paxinos, G. and Watson, C., The Rar Brain in Sterrofasic C’oordininates. Academic Press, Sydney, 1986. [28] Racine, R.J., Modification of seizure activity by electrical stimulation. II. Motor seizure, Elecrroenccph. Clirl. Nruroph~,siol.. 32 ( 1972) 28 l-294. [29] Rothman, S.. Noncompetitive N-metyl-D-aspartate antagonists affect multiple ionic currents, J. Pharmacol. E-x-p. Ther.. 246 (1988) 137-142. [30] Schmutz, M.. Portet, C., Jeker, A.. Klebs. K.. Vassout. A.. Allgeier, H.. Heckendorn, R., Fagg. G.E., Olpe. H.-R. and van Rienen. H., The competitive NMDA receptor antagonists CGP 37849 and CGP 39551 are potent, orally active anticonvulsants in rodents, Nuun~n-S~hmie[leher~‘.~ Arch. Phrrrmcrcol., 342 (1990) 61-66. [31] Sparenborg. S., Brennecke, L.H., Jaax, N.K. and Braitman, D.J., Dizocilpine (MK-801) arrests ,stata.~ epi/ep/icu,s and prevents brain damage induced by soman. Neuropharmacology. [32] Treiman.
3 I ( 1992) 357-368.
D.M., Walton. N.Y. and Kendrick, progressive sequence of electrographic changes generalized convulsive status epilepticus. Epikps~ (I 990) 49 60.
[33] Turski, W.A., Cavalheiro, E.A.. Schwarz, M., Czuczwar, S.J.. Kleinrok, Z. and Turski. L.. Limbic seizures produced by pilocarpine in rats: behavioural, electroencephalographic and neuropathological study. B&Jv. Bruin RLJS.. 9 (1983) 315-335. [34] Walther. H.. Lambert. J.D.C., Jones, R.S.G.. Heinemann. U. and Hamon, B., Epilepttform activity in combined slices of the hippocampus, subiculum and entorhinal cortex during perfusion with low magnesium medium. Ncuro.~ci. Le/r.. 69 (I 986) l56m I6 I. [35] Walton, N.Y. and Treiman. D.M.. Motor and electroencephalographic response of refractory experimental status epilepticus in rats to treatment with MK-801, diazepam. or MK-801 plus diazepam. &air? Rex.. 553 (1991) 97 104. [36] Wamil. A.W. and McLean, M.J., Use-. concentrationand voltage-dependent limitation by MK-801 of action potential tiring frequency in mouse central neurons in cell culture. J. Phurmacol. Esp. Ther., 260 (1992) 376 383. [37] Wood. J.D. and Hertz. L.. Ketamine-induced changes in the GABA system of mouse brain, Neuropharmrrcolo~~, 19
(1980)805-808. [38] Wyllie.
C.. A during Rrs..
5
apoptosis Lockshm Chapman
A.H., Cell death: a new classification from necrosis. In: I.D. Bowen (Eds.). Cell Death WI Biology md and Hall, London. 1981, pp. 9~34.
separating and R.A. Parholo~~,
Research
17
(1994)207.--2 19
The competitive NMDA receptor antagonist CCP 40116 protects against status epilepticus-induced neuronal damage Demon G. Fujikawa”$b*f, Allan H. Danids”, John S. Kim” “Experimenral Neurolog_vLaboratory, Sepcdvrdu VA Medical Cenier. 161II Plummer Street. Sepulveda, CA 91343. USA. hL3epartmenf of Neurology and &win Researcl~Insiitute. UCLA School qf Medicine, Los Angeles, CA 90024, USA (Received
3 August
1993; revision
received 28 September 1993;accepted 20 October 1993)
Abstract
We studied the efftcacy of the competitive NMDA receptor antagonist CGP 40116 in protecting against seizureinduced neuronal necrosis from Iithium-pilocarpine-induced status epilepticus (SE). Rats were given CGP 40116 either before SE (I 2 mg/kg i.p.) or 15 min after the onset of SE (4, 12 and 24 mgjkg); controls received normal saline 15 min after SE began, Diazepam and phenobarbitai were given i.p. after 3 h of SE to stop the seizures. Rats were killed 24 h later, and their brains were processed for light microscopic examination. Neuronal damage occurred in 24 of 25 brain regions examined in saline-injected animals, Protection was maximal in rats given 12 and 24 mg/kg CGP 40116 after SE onset: 19 and 21 of the 24 damaged regions were protected respectively, but the 24 mg/kg group had a mortality rate comparable to saline-injected controls. No necrotic neurons were found in posterior cinguiate and retrosplenial neurons at the two highest CGP 40116 doses, suggesting that the transient cytoplasmic vacuolization induced by NMDA receptor antagonists does not progress to frank necrosis. In rats given CGP 40116 seizure discharges were not eliminated, but their amplitudes were signi~~antly reduced 2 h after SE began. The periodic epileptiform discharge (PED) EEG pattern, probably a sign of widespread neuronal damage, developed in saline-injected controls after 2-2.5 h of SE but not in rats given 12 and 24 mg/kg of CGP 40116. CGP 40116 provided widespread protection against seizureinduced neuronal necrosis, suggesting that an esssential step in its production is NMDA receptor activation by endogenous glutamate. The neuroprotection provided was not simply an antiepileptic effect, since electrographic seizures persisted despite NMDA receptor blockade. CGP 40116 and NMDA receptor antagonists in general could be useful as adjunctive neuroprotectants in patients with refractory SE. Key wortfs: NMDA receptor antagonist; ~...
--
- .~
.-.-
CGP 40116; Status epilepticus: Neuronal necrosis --
1. Introduction Seizures induced in the rat by the muscarinic cholinergic agonist pilocarpine or by the combina* Corresponding 8 1X-895-9554.
author.
Tel.: 818-891-7711,
ext. 7610; Fax:
0920- 12 11iY4/$7.00 .(:. 1994 Elsevier Science B.V. All rights reserved SSRl 0920-121 1(93)E0089-Z
._--
_~
.-._-,_.-___
__..
_... _.“...__...
tion of lithium and pilocarpine, which reduces the dosage of pilocarpine required about 13-fold, produce widespread neuronal necrosis, especially to limbic structures [3,13,33]. Behaviorally, the seizures resemble kindled seizures, and can be classified by the S-stage system used to describe the latter [28]. Because limbic structures are involved,
these seizures are commonly referred to as limbic seizures, to differentiate them from generalized convulsive or nonconvulsive seizures. The anticholinergic drug atropine has no effect once seizures have begun, suggesting spread of seizure activity beyond the cholinergic system [ 13,151, Based upon the excitotoxic action of exogenously administered t-glutamate and its anaiogues, it has been postulated that excessive presynaptic release of L-glutamate produces seizure-induced neuronal necrosis [23]. Activation of the Nmethyl-D-aspartate (NMDA) subtype of glutamate receptor postsynaptically opens the receptor-operated cation channel within the receptor, permitting calcium influx and increasing the intracellular free calcium concentration [16]. It is currently hypothesized that any pathological process (such as status epilepticus (SE)) which increases glutamate release activates an increased number of NMDA receptors for a critical period of time, leading to neuronal necrosis by elevating [Ca2+li and activating potentially destructive CaZ ’ -dependent enzymes f21]. The noncompetitive NMDA receptor antagonists such as ketamine and MK801, which block calcium entry by binding to a site within or at the mouth of the cation channel [19], protect against neuronal damage from kainate-, lithium-pilocarpineand soman-induced SE despite ongoing electrographic seizures [4,9.10,31]. This suggests that the neuroprotection provided is not simply an antiepileptic effect and that seizures can persist despite NMDA receptor blockade. In contrast to the noncompetitive antagonists, there are no data on the effect of systemically administered competitive NMDA receptor antagonists, which bind to the agonist recognition site on the receptor, in protecting against seizure-induced neuronal damage. For this reason we decided to test the effect of a new competitive NMDA receptor antagonist. I>-(E)-2-amino-4acid (CGP methyl-5-phosphono-3-pentenoic 40116; Ciba-Geigy Ltd., Base], Switzerland). Its racemate (CGP 37849) is effective in suppressing electroshock seizures in rats: after i.v. pretreatment, the greatest protection was obtained 2 and 4 h later (ED50 = 4.3 mgjkg), whereas i.p. pretreatment was most effective after 4 h (ED50 = 2.8 mgi kg) [30]. CGP 40116, the most potent competitive
antagonist to date, shows little or no specific binding to other receptors at pharmacological concentrations [8]. In contrast, the noncompetitive NMDA receptor antagonists ketamine and MKX01 also act at other receptors and ion channels [7.12,29.36,37]. By using CGP 401 I6 wc determined how effective NMDA receptor blockade is in ameliorating the neuronal damage and eiectrographic seizure activity caused by lithium-pilocarpine-induced SE. This by inference suggests to what extent activation of the NMDA receptor by endogenous glutamate is responsible for the damage and the EEG seizure discharges.
2. Methods 2. I. Materials
Pilocarpine HCl, atropine methyl bromide (methylatropine) and lithium chloride were purchased from Sigma Chemical Company. The pilocarpine and methylatropine were dissolved in 0.9% sodium chloride and the lithium chloride was dissolved in distilled water prior to injection (200 mg/ml, 100 mg/ml and 127 mg/ml respectively). Pentobarbital sodium (50 mg/ml), ketamine HCI (50 mg/ml), diazepam (5 mg/ml) and phenobarbital sodium (130 mg/ml) were obtained from the Pharmacy Service at Sepulveda VA Medical Center. CGP 40116 was generously provided by Ciba-Geigy Ltd., Base& Switzerland. It was dissolved in 0.9% sodium chloride and titrated with 0.1 M NaOH to a pH of 4-S. Bouin’s solution for brain fixation was prepared in standard fashion
[181. After the i.p. administration of pentobarbital, 40 mg/kg, and ketamine, 25 mg/kg, maie Wistar rats (200-450 g) had three stainless steel screws implanted into their skulls for later recording of electroencephalographic (EEG) activity. Two screws on either side of the midline anterior to the coronal sutures served as active electrodes, and a screw anterior to the lambdoidal suture to the right of midline was used as a ground electrode. Four to six days later, rats were given lithium
D.G. Fujikawu et al.!Epilepsy Research 17 IlYY4) 207-219
chloride, 3 mEy/kg i.p., and the following day they had continuous monitoring of frontal EEG activity during a 20-min baseline period, following pilocarpine injection, during 3-h seizures and the first 2 h of a 24-h recovery period. In addition, rectal temperatures were taken and behavioral observations were made every 15 min. Following baseline measurements, rats were then given methylatropine, 10 mg/kg i.p., to prevent systemic muscarinic side effects, followed by pilocarpine, 30 mg/kg i.p. If no seizures occurred within 30 min, a second 30 mg/kg dose of pilocarpine was given. Rectal temperatures during seizures and the first 2 h of recovery were kept between 37.0 and 39.O”C. Rats were placed on a bed of ice covered with a cloth towel if their temperatures rose to 38.6”C; rectal probes were attached to a temperature controller which automatically switched on a heat lamp if temperatures ‘dropped to 37.5”C, at which point they were taken off the ice. In order to minimize variations in the duration of SE and the degree of neuronal damage in a given brain region from rat to rat, its onset was defined as the time that continuous behavioral and electrographic seizures began, and antiepileptic drugs (diazepam, 10 mg/kg and phenobarbital, 25 mg/kg i.p.) were given 3 h later. In addition, the muscarinic cholinergic antagonist atropine, which crosses the blood-brain barrier, was given i.p. (1 mgikg) with the antiepileptic drugs to lessen the cholinergic component of the seizures. Three hours was chosen because neuronal injury is found in all but one of the 25 brain regions at that time point (manuscript submitted for publication; see also the data on the saline-injected controls in Table 3) and longer seizures were not deemed necessary. CGP 40116, 12 mg/kg i.p., was given at the time of the first pilocarpine injection in one group of rats (n = 6). and 15 min after SE onset in groups given 4 mg/kg (n = 5), 12 mg/kg (n = 5) and 24 mg/ kg (n=5) i.p. A control group was given normal saline i.p. 15 min after SE onset (n = 5). Fifteen minutes after SE onset corresponded to about 20 min after the first seizure and about SO min after pilocarpine injection (Table 2). This time point was chosen because it is prior to the time that neuronal necrosis first appears in this seizure mod-
209
el [1 I] (manuscript submitted for publication), and it was our intention to study the neuroprotective and antiepileptic effects of CGP 40116 during ongoing SE but before neuronal damage occurred. The amplitude of EEG activity was measured prior to SE during the baseline period, and during 3-h SE the maximal amplitudes and frequencies of EEG seizure discharges were measured and rats’ behavior was recorded every 15 min. Foliowing diazepam and phenobarbital administration EEC seizure activity was also measured and behavioral observations were recorded every 15 min for 2 h. EEG seizure discharges persisted during the 3h observation period in all groups Twenty-four hours after 3-h seizures, EEG and behavioral observations were made, then rats were deeply anesthetized with 400 mg/kg pentobarbital i.p. and underwent transcardiac perfusion-fixation with heparinized Ringer’s lactate solution followed by 200 ml of Bouin’s solution, at a perfusion pressure of 120 mm Hg. Brains were left in situ for 1-3 h, following which they were removed and put in Bouin’s solution overnight. The following morning brains were transferred to 70% ethanol; they were kept in 70% ethanol until further processing. Each brain was then cut coronally into 2---3mm thick blocks, dehydrated in graded ethanol solutions, cleared with xylene, embedded in paraffin, sectioned at 6 pm thickness, deparaffnized and hydrated to water, and stained with hematoyxlin and eosin (H&E). Sections at the level of the caudate-putamen, dorsal hippocampus and ventral hippocampus (corresponding to 0.20 mm, 3.30 mm and 5.60 mm from bregma respectively [27]) were examined to determine the extent of neuronal necrosis in 25 brain regions, which were selected because of their vulnerability to seizure-induced neuronal damage. One section at each level was used to rate the degree of damage, and several adjacent sections at each level were checked to insure that the section used was representative. Frontotemporoparietal and cingulate cortices were examined at the level of the caudate-putamen, retrosplenial cortex at the levels of the dorsal and ventral hippocampi, and thalamic nuclei at the dorsal hippocampal level. The 25 brain regions were rated for the extent
of irreversible neuronal damage according to the following scale: grade 0= no damage, grade 0.5 = < 10% acidophilic neurons (slight damage), grade I .O = IO-25% (mild damage), grade 1.5 = 26645%, grade 2.0 = 46-54% (moderate dagrade 2.5 = 55575% and grade mage), 3.0 = > 75% acidophihc neurons (severe damage). Acidophihc neuronal cytoplasm and nuclear hyperchromatism (pyknosis) were used as the criteria of irreversible neuronal damage [ 14,383. Actual cell counts of acidophilic and normal neurons were done at 200 x magnification in the dorsal and ventral dentate hilus (dorsal and ventral CA4) and in dorsal CA1 (data not shown); counts from both hemispheres were added together, and the combined percentage of acidophilic neurons was calculated for each rat. In the 22 remaining brain regions, visual estimates of the percentage of acidophilic neurons present were made and a histological damage score was assigned because quantitative cell counts would have been impractical given the vast numbers of neurons in each of the regions. The histological damage scores of the two hemispheres were averaged for each rat. In the dentate hilus and dorsal CA1 (data not shown), the visual estimates correlated very well with actual cell counts, which were done subsequently (Tables 3, 4 and 5). Grade 2 damage was associated with at least mild edema and grade 3 damage with severe edema and a disrupted, pale neuropil. 2.3. Statistical analysis In order to determine if the histological damage scores were normally distributed, thereby permitting standard parametric methods of data analy(individual scores less group sis, ‘residuals’ means) were plotted for a representative sample of brain regions. Normal (Gaussian) curves were approximated, so one-way analysis of variance (ANOVA) was used to determine if differences in histological scores existed among the five groups of rats in the 25 brain regions examined. If the overall P value for the F-ratio obtained for a given brain region was 60.05, post-hoc comparisons between groups were made with Fisher’s least significant difference (LSD) test, using pooled SEMs and CI=0.05. One-way ANOVA
with post-hoc Fisher’s LSD test when indicated was also used to compare hilar (CA4) neuronal cell counts in dorsal and ventral dentate gyrus and to analyze seizure and SE latency data and EEG seizure discharge amplitude and frequency data.
3. Results 3.1. Survival data Table 1 summarizes the survival data in salineinjected controls and CGP 40116-injected rats. In this particular group of controls, only seven of 17 (41%) developed SE; of the seven with SE, two (29%) died during seizures. All of the CGP 40116-injected rats developed SE, and four of 26 (15%) died during SE or the first 2 h of the recovery period. Deaths were confined to the groups given 24 mg/kg CGP 40116 (3/9, or 33%) and 12 mg/kg (l/6, or 17%) after SE onset. In addition, one of the nine rats became apneic after it was given 24 mg/kg; it was resuscitated and eventually recovered but was not included in the EEG and histological analyses. The reason that so few saline-injected controls developed SE is not clear: 80% of the rats which did not develop SE received the higher (60 mg/kg) dose of pilocarpine. Moreover, 77% of the CGP 40116-injected rats were given the drug after SE had begun so it could not have been responsible for the 100%
Table I Survival data Group
SE
Saline-injected controls CGP 40116 before SE onset
s
IO
2
17
12 mg/kg CGP 401 I6 after SE onset
6
0
0
6
s 5 6”
0 0 0
0 I ?
5 6 9
4 mgikg 12 mg/kg 24 mg/kg
~.__
No SE .~~._
Died _
Total
~~-~
SE = the number of rats with status epilepticus; no SE = the number of rats which did not develop SE despite 60 mgikg of pilocarpine. “One of the rats became apneic after it was given CGP 40116; it was resuscitated and eventually recovered, but was not included in the EEG and histological analyses.
rate of seizure induction
in the CGP 40116 groups.
3.2. EEG and behavioral observations The time intervals between the first pilocarpine injection and the first seizure and between the first injection and the onset of SE did not vary signiticantly among the five groups tested (Table 2). Four of the five (80%) saline-injected SE controls received 30 mg/kg of pilocarpine; one (20%) got 60 mg/kg. Of the CGP 40116-injected rats, four of 21 (19%) received 30 mg/kg; the remaining 17 (81%) got 60 mg/kg. The CGP 40116 was not responsible for the high percentage requiring 60 mg/kg, since 15 of the 21 (71%) were given CGP 40116 after seizures began. The dose of pilocarpine given had no effect on the latency either to the first seizure or to SE.
Table 2 Times from pilocarpine injection onset of status epilepticus
to the first seizure and to the -
Group
Latency to first seizure (min)
Latency to SE (min)
Saline-injected controls (5)
34+10
68+ 8
57+11
74&15
39* 6 39+ 7 3S& 5
48+ 41+ 43+
CGP 40116 given before the onset of SE I? mg:kg (6) CGP 40116 given after the onset of SE 4 mg:kg (5) I2 m&kg (5) 24 mg,‘kg (5)
7 6 5
The numbers in parentheses refer to the number of rats in each group. Latency to iirst seizure (mean + SEM) refers to the time from the first pilocarpine injectmn to the appearance of the first seizure. Latency to SE refers to the time from the frst pilocarpine injection to the appearance of continuous seizures. or status epilepticus (SE). No differences between groups were found wth respect to either measure.
A
B
BASELINE
1000uv
15-MIN SE (saline i.pJ
BASELINE
15-MIN SE (CGP 401 ‘I6 12 mg/kg i.p.)
1000 pv
1 KE
l-HOUR
SE
1000 jlv
?-HOUR
SE
J-HOUR
SE
\
1 set
3-HOUR
SE
1000
pv L-
1 set
--+----~----2-HOUR
RECOVERY
1000
-2-HOUR
uv
RECOVERY
I_-1 WE
Fig, 1. (A) This is the frontal EEG activity during baseline, status epilepticus (SE) and the early recovery period in a rat given normal saline i.p. 15 min after the onset of SE. (B) This EEG is from a rat given CGP 40116 (12 mg/kg i.p.) 15 min after SE onset at the same time periods as in A. EEG seizure discharges persisted for 3 h in both rats, although the 3-h SE shows a periodic epileptiform discharge (PED) pattern in the saline-injected control and a frequent spike discharge pattern in the CGP 401 l6-injected rat.
Noncompetitive NMDA receptor antagonists have been given before convulsant drug administration [4,9,26,3 11, 5 mm after the first EEG spikes [31], 10 [26,35], 15 [lo] or 40 min [26] after the onset of continuous EEG spikes. or 2 h after 90 min of continuous electrical stimulation of the hippocampus [2]. In this study CGP 40116 was given either at the time of the first pilocarpine injection or 15 min after the onset of continuous electrographic and behavioral seizure activity (SE), which was about 20 min after the first seizure and about 60 min after the first pilocarpine injection was given (Table 2). EEG seizure discharges (Fig. 1) and clonic jerks, often progressing to rearing and falling and corresponding behaviorally to kindled seizure stages 3-5 [28], persisted in the saline-injected controls until atropine (1 mg/kg), diazepam (10 mg/ kg) and phenobarbital (25 mg/kg) were injected i.p. after 3 h of continuous seizures. Although EEG seizure discharges were not eliminated in those rats given CGP 40116, discharge amplitudes were significantly reduced 2 h after SE onset in all
of the rats given CGP 40116 (Fig. 2A). The amplitudes of seizure discharges did not decline in saline-injected seizure animals. The mean seizure discharge amplitudes of the group given CGP 40116 prior to seizure onset were consistently lower than those of the other groups from I5 to 180 min of SE, and those of the saline-injected control group were consistently higher than those of the CGP 40116 groups from 90 to 180 min of SE, although these differences were only statistically significant at 120 min of SE (Fig. 2A). There were significant differences between groups in the frequencies of EEG seizure discharges 15 and 165 min after SE began (Fig. 2B). Discharge frequencies were higher after 15min SE in the rats given 4 and 12 mg/kg CGP 40 I 16 than in those given saline at that time point; the EEG recordings were obtained just before CGP 40 I 16 or saline injections. The groups given 12 and 24 mg/kg after 15 min of SE had higher discharge frequencies than the saline-injected group 165 min after SE began. The seizure discharge frequencies of the saline-
a
5 3 2
2000
2 si
1600
3 %
1200
7 6
z E
800
z cc
400
t: w
Oi 0
30
60
90
120
150
DURATION OF STATUS EPILEPTICUS(MIN)
180
0
30
I
I
60
90
i--,-~-I
120
150
180
DURATION OF STATUS EPILEPTICUS (MIN)
Fig. 2. Saline-injected control group (n = 5, q), group given 12 mg/kg CGP 40116 prior to SE onset (n = 6, n ), group given 4 mg/kg CGP 40116 15 min after SE onset (n = 5, l ), group given I2 mg/kg CGP 401 I6 after SE onset (n = 5, A) and group given 24 mg/kg CGP 40116 after SE onset (n = 5, v), Values represent mean f SEM; standard error bars were omitted for clarity. The arrow indicates when CGP 40116 was given in the latter three groups, (A) EEG seizure discharge amplitudes in each of the tive seizure groups during the 3-h SE period. ‘P
injected control group progressively declined, and 120-l 50 min after the onset of SE to the end of the 3-h seizure period a periodic epileptiform discharge (PED) pattern was seen, in which low-frequency epileptiform bursts occurred on a flat background (Figs. 1A and 2B). Groups given CGP 40116 had higher seizure discharge frequencies (statistically significant only at 15 and 165 min of SE) and higher voltage EEG background activity and did not develop PEDs. Behavioral seizures were abolished in rats given 12 and 24 mg/kg CGP 40116, which lay prone on their abdomens with limbs flexed or occasionally
spread laterally. No clonic jerks were seen, but fine body tremors were often present. Diazepam and phenobarbital decreased the amplitude and frequency of PEDs in the saline-injected control group (to 227$67 pV, 0.4fO.l Hz (mean+ SEM)) at 2 h of recovery; in the 2-h recovery EEG shown in Fig. 1A the PEDs were of such low voltage that they were not visible at the amplifier gain shown. In general, by 2 h of recovery seizure discharges in the CGP 40116-treated rats were replaced by a 1 Hz or less burst suppression pattern, with bursts consisting of 20-30 Hz rhythmic activity lasting up to 800 ms. In the 2-h recov-
Fig. 3. These are photomicrographs of the ventral CAI pyramidal cell layer (A and B) and the medial nucleus of the amygdala (C and D) in the same rats as in Fig. I. The rat given saline i.p. (Fig. IA) is shown in A and C. and the one given CGP 401 16 (Fig. I B) is shown in B and D (H and E. x 132; scale bar = 40 /lrn). In contrast to the extensive neuronal necrosis (arrows) and edema (asterisks) in A and C. the neurons and neuropil in B and D are normal in appearance.
Histological
damage
scores
Brain region
Dorsal hippocampus CA1 CA2 CA3 Ventral hippocampus CA1 CA2 CA3 Dorsal dentate gyrus Dentate granule cells Hilar region Ventral dentate gyrus Dentate granule cells Hilar region Amygdala Piriform cortex Entorhinal cortex Thalamus Mediodorsal nucleus Lateroposterior Laterodorsal nucleus Centromedian nucleus Reuniens nucleus Cerebral cortex Frontoparietotemporal Cingulate Retrosplenial Septal nuclei Caudatejputamen Substantia nigra Pars compacta Pars reticulata
Saline-injected
CGP 40116 Before SE
15 min after SE onset
12 mg/kg
4 mg/kg
12w&g
24 mgjkg
0.5kO.2 0.2*0.1 0.7+0.1
0.1 &o.l* o.o+o.o* 0.5*0.1
0.1_+0.1* o.o+o.o* 0.6iO.2
o.o~o.o* 0.0 * 0.0 0.2+0.1*
0.0 * o.o* 0.0 * o.o* 0.2*0.1*
2.2kO.l 2.OkO.2 I .4* 0. I
0.9*0.3* 0.8 +0.4* 0.6f0.3*
I .o f 0.4* 1.6kO.5 0.9+0.3
0.2fo.l* o.o+o.o*~*** 0.3;0.1*
0.2+0.1*.*** o.o+o.o*~*** 0.3+0.1*
0.6+0.1 2.4kO.O
0.0 * o.o* 2.OkO.4
0.2*0.1* 1.g+o.3
o.o*o.o* 2.4+O.l
o.o*o.o* 2.1&O.?
0.5*0.1 2.6LO.l 1.5+0.2 2.6kO.3 2.OkO.4
0.0 * o.o* 2.0 5 0.4 0.6iO.l* 2.1 TO.4 0.9*0.2*
0.1 *o.i* 2.2kO.4 0.5*0.2* 2.5kO.5 1.610.4
o.o*o.o* 2.6kO.i 0.2*0.1* 0.6+0.0*‘**‘*** 0.6 Itr O.O*‘***
o.o~o.o* 2.2kO.l 0.2+0.1* 0.5+0.0*‘**‘*** 0.5 *o.o*.***
1.2*0.1 l.2iO.3 1.3+0.4 I .7 * 0.4 2.1 io.3
0.4+0.2* 0.1 fo.l*‘*** o.o+o.o* o.g*o.5 o.g+o.4*
0.4*0.1* 1.0+0.5 0.6+0.2* 1.0+0.4 1.5io.4
o.o*o.o* o.o~o.o*~*** o.o+o.o* 0.6+0.2 0.2+0.1*~***
o.o+o.o* 0.1+0.1* 0.2+0.1* 0.2+0.1* 0.3+0.1*,***
0.5*0.1 0.5+0.0 0.4f0.1 I .o * 0.2 1.2kO.3
0.0 +o.o*,*** o.o*o.o*,*** 0.0 +O.O*‘*** 0.5:0.2* 0.3&0.2*
0.6kO.l 0.2+0.1* 0.2*0.1 0.7io.3 0.6iO.3
0.3i:o.1**,*** o.o*o.o*~*** 0.1 kO.1’ O.O+O.O*~**~*** 0.3iO.l’
0.2*0.
0.0 * 0.0 2.2kO.5
o.o*o.o 1.0*0.5
0.0 f 0.0 1.9+0.4
o.o*o.o 2.OkO.3
o.o+o.o 1.9kO.5
I *.**,*** o.o+O.O*~*** o.o+o.o*~*** 0.0+o.o*‘*** 0.2+0.1*
SE = status epilepticus. The data represent grades of histological damage (mean f SEM) in the live 3-h seizure groups (grade 0, no damage; grade I, mild; grade 2, moderate; grade 3, severe damage; see Methods for details). One-way ANOVA was followed by posthoc Fisher least significant difference comparisons using pooled SEMs and SDS (see Methods for details). *P
ery EEG in Fig. 1B some of the low-voltage bursts are visible. Behavioral seizures stopped in the saline-injected group, and both saline- and CGP 40 116-injected animals lay in whatever position they were placed. Twenty-four hours later saline-injected control
rats had EEGs showing low-voltage PEDs (4541139 pV, 0.8 +0.2 Hz) on a flat background; the CGP 40116 groups had low-voltage seizure discharges with otherwise unremarkable EEGs (means f SEMs for discharge amplitudes were between 243 + 127 and 354 + 139 pV, and there were
between 5 * 2 and 9 &-2 discharges every 10 s). The saline-injected rats were quiet, sitting with backs hunched and with little movement. The group given CGP 40116 prior to SE behaved normally. The behavior of the other rats given CGP 40116 varied from normal to those with slow movements with head jerks in the 4 and 12 mg/kg groups; three of four in the 24 mg/kg group moved by sliding around on their abdomens and one of the four appeared ataxic while walking. The CGP 40116 behavioral changes would probably have normalized had the survival period been extended beyond 24 h. Rectal temperatures were monitored before, during and following seizures and did not change (data not shown). 3.3. Histotogicaf results CGP 40116 provided protection against SE-induced neuronal damage when given either before or 15 min after the onset of SE (Fig. 3 and Table 3). The number of protected brain regions increased as the CGP 40116 dose increased: at 4 mg/kg nine of 24 damaged regions were protected, at 12 mg/kg before and after SE 18 and 19 regions respectively were protected, and at 24 mg/ kg 21 regions were protected (Table 3). When the groups given 12 mg/kg and 24 mg/kg after SE onset were compared, no differences were found between them. However, as mentioned previously, the 24 mg/kg group had a mortality rate comparable to seizure controls, and apnea occurred after CGP 40116 injection in one of nine (Table 1). Giving 12 mg/kg CGP 40 116 after seizure onset provided comparable protection to giving it prior to seizures. A comparison of the two groups showed that two brain regions had less damage and one region had more damage in the group given CGP 40116 after SE began (Table 3). Four brain regions had greater damage in the group given 4 mg/kg CGP 40116 when it was compared to the one given 12 mg/kg before SE. This increased to eight and nine regions when the 4 mg/ kg group was compared to the groups given 12 and 24 mg/kg respectively after the onset of SE. Posterior cingu~ate and retrosplenial cortices were examined to determine if the NMDA receptor antagonist-induced transient cytoplasmic va-
Table 4 Dorsal hiiar (CA4)
~--
neurons
_----
AN
5414 Saline-injected SE controls CGP 40116 before SE onset SO-i_12 12 mg/kg i.p. CGP 40116 1S-min after SE onset 34k8 4 mg/kg i.p. 4514 12 mg;kg i.p. 4112 24 mg/kg i.p.
NN
O/oAN --
32+3
63&4
43&X
52111
35&h 29_+2
4819 61_t4 5224 38i4 _.---l.-
SE = status epilepticus, AN = number of acidophilic neurons, NN = number of normal-appearing neurons, % AN = percentage of the total number of neurons (AN + NN) which are acidophilic. Values represent mean k SEM. The percentages of acidophilic neurons in each group correspond closely to the histological damage scores in Table 3. No differences between groups were found.
cuolization noted by Olney and colleagues [24,25] progresses to neuronal necrosis. The degree of neuronal necrosis was always less than that which occurred in frontoparietotemporal cortex (layers 2-3 were always involved, and occasionally layers 4-6), and there were often no damaged neurons even when they appeared in frontoparietotemporal cortex (Table 3). At the 12 mg/kg dose of CGP 40 116 given before SE and the 24 mgjkg dose, no necrotic neurons were seen in any of the cerebral cortical regions examined, and at the 12 mg/kg dose given after SE began there were no necrotic neurons in cingulate cortex despite slight cortical damage elsewhere (Table 3). Table 5 Ventral hilar (CA4) neurons _ Saline-injected SE controls 8ii8 CGP 40116 before SE onset 12mg: kg i.p. ho+ 13 CGP 40116 I5 min after SE anset 4 mgikg i.p. 63114 12 mg;kg i.p. 68&J 24 mg:kg i.p. 56+S
30*4
-73*2
52*i2
53jll
43+13 30*2 48,5
60~12 69+2 54+3 -.“-
SE = status epilepticus, AN = number of a~idophiiic neurons, NN = number of normal-appearing neurons. % AN = percentage of the total number of neurons (AN + NN) which are acidophilic. Values represent mean + SEM. As in Table 4, the percentages of a~idophilic neurons in each group corresponded closely to the histological damage scores in Table 3 and there were no differences between groups,
In the hilus of the hippocampal dentate gyrus (CA4), where cell counts were performed, there were no differences in the number and percentage of acidophilic neurons between saline-injected rats and any of the groups given CGP 40116 (Tables 4 and 5). In addition to the dentate hilus. the only other damaged brain region not protected by CGP 401 16 was substantia nigra pars reticulata (Table 3). Within the pars reticulata, well circumscribed, oval-shaped lesions of variable size were found. They were characterized by a pale, disrupted neuropil, necrosis of all neurons and a variable degree of astrocytic necrosis. The adjacent substantia nigra pars compacta was normal in the saline-injected seizure group as well as all of the CGP 40 I I6 groups.
4. Discussion 4.1. The neuroprotective efikcts of’ CGP 40116 and noncompetitive NMDA receptor antagonists To our knowledge, this is the first study to show that a parenterally administered competitive NMDA receptor antagonist protects against SEinduced neuronal necrosis. The protection offered by the competitive NMDA receptor antagonist CGP 40116 against seizure-induced neuronal damage, even when it is given 15 min after the onset of SE, suggests that seizure-induced activation of this receptor participates in producing the damage. Maximal protection was provided by 12 mg/kg CGP 40116 (19 of 24 damaged regions) and 24 mg/kg (21 of 24) given 15 min after seizure onset (Table 3). When these groups were compared to each other, no differences in the degree of protection were found in any brain region. No additional benefit was provided by the higher dose, so the 12 mg/kg dose can be considered the most effective, especially since the group given 24 mg/kg had a mortality rate comparable to seizure controls (Table I). Giving 12 mg/kg CGP 401 I6 before SE onset offered no advantage compared to the same dose given after SE onset. Despite the protection provided by 12 mg/kg CGP 15 min after SE onset, five regions were not significantly improved. Two of these five regions had only slight damage, and one of the two was
compared to a region which itself had only been slightly damaged (Table 3). In the other three rcgions the moderate-to-severe neuronal damage seen in both the dorsal and ventral hilus and in substantia nigra pars reticulata was not amehorated by CGP 40116 (Tables 35). It is possible that in the latter three regions the seizure-induced neuronal damage was not mediated by the NMDA receptor. The well circumscribed, ovalshaped lesions within the substantia nigra pars reticulata are similar to those described in rats 1 week after flurothyl-induced SE [22], except that in the current study macrophage infiltration was not seen, probably because the recovery period was shorter. The npncompetitive NMDA receptor antagonists MK-801 and ketamine protect against SEinduced neuronal damage despite ongoing electrographic seizures, like CGP 40116 [4,9,10,31]. However, ketamine and MK-801 have pharmacological actions other than at the NMDA receptor so their neuroprotective effect [7.12,29,36.37], may not be due strictly to NMDA receptor antagonism. When given 15 min after the onset of SE, ketamine (100 mg/kg i.p.) protected against seizure-induced neuronal damage in 23 of the 24 brain regions damaged by lithium-pilocarpine SE (dorsal CA2, the one unprotected region, was only minimally damaged in saline-injected controls) [lo] (manuscript submitted for publication). In the current study 19 of the 24 damaged regions were protected by 12 mgikg CGP 40116 given 15 min after SE onset in the same seizure model. Neuronal damage may not have been NMDA receptor-mediated in the three regions which showed persistent moderate-to-severe damage (dorsal and ventral hippocampal hilar regions and substantia nigra pars reticulata). Except for these three regions, the neuroprotective effects of CGP 40116 and ketamine were comparable. 4.2. The antiepileptic effixts qf’ CGP 40116 und noncompetitive NMDA receptor antagonists Early reports on parenterally administered competitive NMDA receptor antagonists focussed on their anticonvulsant properties in rodent seizure models [5,6]. Only the behavioral manifestations of seizures. which were prevented
by pre-administration of the antagonist, were studied. It was later found that these antagonists inhibited seizure discharges in the hippocampal slice [34]. However, studies with the noncompetitive antagonists ketamine and MK-801 [4,9,10,31] and this study with CGP 40116 have shown that after parenteral injection of these agents electrographic seizure discharges may persist, even though the behavioral manifestations of seizures are suppressed. Although electrographic seizure discharges persisted in this study, their amplitudes were significantly reduced compared to controls by 2 h after the onset of SE, and the PED pattern which evolved in saline-injected control rats at 2-2.5 h of SE did not occur in the rats given CGP 40116 (Fig. 2A,B). Similar results have been reported for MK-801 [26,32]. The development of PEDs as a late stage in SE has been described in both experimentally induced 117,321 and human SE 1321. Our data support the view that the PED pattern is an EEG marker of widespread neuronal damage 1351. The higher seizure discharge frequencies at 15 min of SE in the groups given 4 and 12 mglkg CGP 40116 compared to those given saline at that time point (Fig. 2B) are not the result of the paradoxical enhancement of electrographic seizure discharges which may occur with ketamine or MK-801 [2,9]. Since the electrographic record was obtained just prior to CGP 40116 injection, the differences could not have been due to CGP 40116 and are probably fortuitous. On the other hand, the higher discharge frequencies noted at 165 min after SE onset in the groups given 12 and 24 mg/kg after SE onset compared to controls may be analogous to the rapid continuous spike activity produced by MK-801 in the lithium-pilocarpine seizure model 1351. Walton and Treiman [RS] contend that this EEG pattern and a lack of progression to PEDs late in SE is a good prognostic sign. Our histological data support this concept. Ketamine and MK-801 also induced increased seizure discharge frequencies in SE produced by continuous hippocampal stimulation, but increased discharge frequency was found in only one of four rats given the competitive NMDA receptor antagonist 3(2-carhoxypiperazine-4-yl)propyl-I-phosphonic acid (CPP), and
only in short bursts with CGP 40116.
[2], in contrast
to our results
Olney and colleagues have reported that both competitive and noncompetitive NMDA receptor antagonists produce transient cytoplasmic vacuolization in rat posterior cingulate and retrosplenial neurons, which they postulate may be responsible for the psychomimetic effects of these agents [24,25]. However, these antagonists do not in general produce lethal damage: at a dose of 5 mgikg MK-80I i.p. (live times that initially reported), the overwhelming majority of cingulate cortical neurons were normal 48 h after MK-801 administration; necrotic neurons were rarely seen [I]. In the current study. no necrotic neurons were seen in cerebral cortex (including posterior cingulate and retrosplenial cortices) in rats given 12 mg’kg CGP 40116 before and those given 24 mg/kg after seizure onset. During SE or other neurological emergencies. the risk of producing transient cingulate neuronal vacuolization and a possible transient psychosis by giving NMDA receptor antagonists would be far outweighed by the potential benefit of preventing widespread neuronal necrosis and permanent neuropsychological seyuelae.
In the lithium-pilocarpine seizure model in adult rats, neuronal necrosis first appeared 40 min after the onset of SE in cerebral cortex and atnygdala and became more widespread after I It of SE [ 11J (lnanuscript submitted for publication). Investigators studying llurothyl-induced SE in the rat also found neuronal damage within I h in several brain regions [l4,22]. In adolescent baboons with bicuculline-induced SE neuronal damage appeared within 1.5 h of SE onset [20]. The implications of these animal studies for the treatment of human SE are that prolonged seizures damage the brain and that persistent or repetitive seizures should be stopped as soon as possible. Heretofore the therapeutic emphasis has been on using drugs which either stop ongoing behavioral and electrographic seizures or prevent their recurrence. Since
NMDA receptor antagonists protect against SEinduced neuronal damage, they could be useful adjunctive agents in the treatment of refractory SE in humans. However, further studies are needed, especially with respect to the margin of safety between the most efficacious and toxic doses of CGP 401 16, Criteria for clinical effkacy must also be developed before human trials are considered.
5. Acknowledgements This research was supported by Ciba-Geigy Ltd., Basel, Switzerland, and by the Research Service of the Department of Veterans Affairs. The assistance of Dr. Jeffrey Gornbein of the UCLA Department of Biomathematics Statistical/Biomathematical Consulting Clinic in the statistical analysis of the data is gratefully acknowledged.
6. References III Allen. H.L. and Iversen, L.L., Phencyclidine,
dizocilpine, and cerebrocortical neurons, Science, 247 (1990) 22 I PI Bertram, E.H. and Lothman. E.W., NMDA receptor antagonists and limbic status epilepticus: a comparison with standard anticonvulsants. Epilrps~ Rrs., 5 (1990) I71 I X4. PI Clifford, D.B.. Olnry, J.W., Maniotis, A., Collins, R.C. and Zorumski. C.F.. The functional anatomy and pathology of lithium-pilocdrpine and high-dose pilocarpine seizures, Neumcimv, 23 (1987) 953-968. [41Clifford D.B., Olney, J.W., Benz, A.M.. Fuller. T.A. and Zorumski. C.F., Ketamine, phencyclidine and MK-801 protect against kainic acid-induced seizure-related brain damage, E~ilqxiu, 3 I ( 1990) 382- 390. M.J., Collins. J.F. and Meldrum, B.S.. AntiISICroucher. convulsant action of excitatory amino acid antagonists. S&ncP. 216 (1982) x99-901. against 161 Czuczwar. S.J. and Meldrum, B.. Protection chemically induced seizures by 2-amino-7_phosphonoheptanoic acid. ,%:ur.J. Phurmuco/.. X3 (1982) 335-338. [7] Ereciliska. M.. Nelson, D. and Silver. I.A., Interactions of benztropine. atropine and ketamine with veratridineactivated sodium channels: effects on membrane depolartLation, K * -efflux and neurotransmitter amino acid release. Br, J. ~~~ur~~ff~~)~., 94 ( 1988) U-881. [X] Fag, G.E.. Olpe, H.-R., Pozza. M.F. et al., CGP 37849 and CGP 39551: novel and potent competitive N-methylI)-aspartate receptor antagonists with oral activity. Br. J. Pkort?tucol. ( 99 (I 990) 79 I 797.
Faricllo, R.G.. Golden, <;.‘I ., Smith. G.G. and Rcycb. P.F.. Potentiation of kaimc acid epileptogenicity and q~aring from neuronal damage by an NMDA rcceptut antagonist, Epipikeps~~Rev.. 3 (1989) 206 213. Fujikawa. D.G., Ketamine given after the onset of limblc status epilepticus protects against seizure-induced neuronal damage. t~~~;~~l~.~jff. 31 ( 1990) 635. Fujikawa, D.G.. Kim. J.S. and Daniels, A.H.. When and where does neuronal necrosis first appear in the lithiumpilocarpine sel7ure model? f$ik,/>.?icr. 33. Suppl. 3 (1492) 36. Gage. P.W. and Robertson, B.. Prolongation of inhibitory postsynaptic currents by pentobarbitone, halothane and ketamine in CA1 pyramidal cells in rat hippocampus, IS?. J. Pfwmwol., 85 (1985) 675 68 I. Honchar. M.P.. Olney, 3.W. and Sherman, W.R.. Systemic eholinergic agents induce seizures and brain damage in lithium-treated rats, Sci~rz~e. 220 (1983) 323~ 325. 1141 Ingvar. M., Morgan, P.F. and Auer, R.N., The nature and timing of excitotoxic neuronal necrosis in the cerebral cortex, hlppocampus and thalamus due to flurothylinduced status epilepticus. A(,rtr !Veuropurhol. I Bed. ). 75 (1988) 362 369. [IS] Jope, R.S., Morrisett, R.A. and Snead. O.C., Characteristlcs of lithium potentiation of pilocarpine-induced status epilepticus in rats, E.qI. Il’curoi., 91 (I 986) 471 480. [16] Lodge. D. and Coliinridge. G.. Lrs qgcgrnrs~r(~~~~~~~~~~~.~~ 3 series on the pharmacology of excitatory amino acids. Trcwt.v Phurmuc~oi. Sci.. I I (1990) 22-24. [I 71 Lothman. E.W., Bertram, E.H.. Beckenstein, J.W. and Perlin, J.B.. Self-sustaining limbic status epilepticus Induced by ‘continuous’ hippocampal stimulation: electrographic and behavioral characteristics. Epikpsy Rcs., 3 (1989) 107 119. Siujnj~r~ ~(,1~7~~[~.s of’ rhc 1181Luna. L.G.. :~ut~~~ll fff ~1.~~~~~~~~~~ Armd Forcr.r r~z,~tjt~t~~ of ~~t~?(~~~/~.~. McGraw-Hill. New York. NY, 1972. 258 pp. J.F. and Nowak, L.M., Mechanisms of 1191 MacDonald, blockade of excitatory amino acid receptor channels. 7’ren& Phurmat ol. Sci.. 1I ( 1990) 167 172. B.S. and Brierley, J.B., Prolonged epileptic [201 Meldrum. seizures in primates: ischemic cell change and its relation to octal phystological events, Arch. ,Ve~rol., 28 (1973) IO -I 7. PII Meldrum, B. and Garthwaite, 3.. Excitatory ammo acid neLirotoxicity and lleurode~encrati~~e disease. Trt*nrl,s P/?unnotvri. Sri., 11 (1990) 379 3x7. PI Ncvander. G.. Ingvar. M.. Auer. R. and Siesjii. B.K.. Status cpilepticus in well-oxygenated rats causes neuronal nccrosib. Alltr. ~Ywrol.. 18 ( 198.5) 281 290. transmitters and epilepsy-related [23] Olney. J.W.. Excitatory brain damage, It,/. Kav. h’crtrohid., 27 (1986) 337 362. P41 Oiney. J.W., Labruyere, J. and Price, M.T., Pathological changes induced in cerebrocortical neurons by phencyclidint and related drugs, .S’&nc~. 344 (1989) 1360 1362. J.. Wang, G., Wozniak, D.F.. P51 Olney. J.W.. Labruyere, Price. M.T. and Sesma, M.A.. NMDA antagonist ncurotoxicity: mechanism and prevention, Scirncc,. 254
(1991) 1515~1518. [26] Ormandy, G.C.. Jope, R.S. and Snead, O.C., Anticonvulsant actions of MK-801 on the lithium-pilocarpine model of status epilepticus in rats. E-up. Neural., 106 (1989) 1722180. [27] Paxinos, G. and Watson, C., The Rar Brain in Sterrofasic C’oordininates. Academic Press, Sydney, 1986. [28] Racine, R.J., Modification of seizure activity by electrical stimulation. II. Motor seizure, Elecrroenccph. Clirl. Nruroph~,siol.. 32 ( 1972) 28 l-294. [29] Rothman, S.. Noncompetitive N-metyl-D-aspartate antagonists affect multiple ionic currents, J. Pharmacol. E-x-p. Ther.. 246 (1988) 137-142. [30] Schmutz, M.. Portet, C., Jeker, A.. Klebs. K.. Vassout. A.. Allgeier, H.. Heckendorn, R., Fagg. G.E., Olpe. H.-R. and van Rienen. H., The competitive NMDA receptor antagonists CGP 37849 and CGP 39551 are potent, orally active anticonvulsants in rodents, Nuun~n-S~hmie[leher~‘.~ Arch. Phrrrmcrcol., 342 (1990) 61-66. [31] Sparenborg. S., Brennecke, L.H., Jaax, N.K. and Braitman, D.J., Dizocilpine (MK-801) arrests ,stata.~ epi/ep/icu,s and prevents brain damage induced by soman. Neuropharmacology. [32] Treiman.
3 I ( 1992) 357-368.
D.M., Walton. N.Y. and Kendrick, progressive sequence of electrographic changes generalized convulsive status epilepticus. Epikps~ (I 990) 49 60.
[33] Turski, W.A., Cavalheiro, E.A.. Schwarz, M., Czuczwar, S.J.. Kleinrok, Z. and Turski. L.. Limbic seizures produced by pilocarpine in rats: behavioural, electroencephalographic and neuropathological study. B&Jv. Bruin RLJS.. 9 (1983) 315-335. [34] Walther. H.. Lambert. J.D.C., Jones, R.S.G.. Heinemann. U. and Hamon, B., Epilepttform activity in combined slices of the hippocampus, subiculum and entorhinal cortex during perfusion with low magnesium medium. Ncuro.~ci. Le/r.. 69 (I 986) l56m I6 I. [35] Walton, N.Y. and Treiman. D.M.. Motor and electroencephalographic response of refractory experimental status epilepticus in rats to treatment with MK-801, diazepam. or MK-801 plus diazepam. &air? Rex.. 553 (1991) 97 104. [36] Wamil. A.W. and McLean, M.J., Use-. concentrationand voltage-dependent limitation by MK-801 of action potential tiring frequency in mouse central neurons in cell culture. J. Phurmacol. Esp. Ther., 260 (1992) 376 383. [37] Wood. J.D. and Hertz. L.. Ketamine-induced changes in the GABA system of mouse brain, Neuropharmrrcolo~~, 19
(1980)805-808. [38] Wyllie.
C.. A during Rrs..
5
apoptosis Lockshm Chapman
A.H., Cell death: a new classification from necrosis. In: I.D. Bowen (Eds.). Cell Death WI Biology md and Hall, London. 1981, pp. 9~34.
separating and R.A. Parholo~~,