- Email: [email protected]
Myoclonus inducing and seizure modifying effect of cysteamine on cortical and amygdaloid kindled rats
Brain Research, 473 (1988) 99-106 Elsevier
99
BRE 14035
Myoclonus inducing and seizure modifying effect of cysteamine on cortical and amygdaloid kindled rats Akira Takazawa* and David K. Bilkey Department of Psychology and The Neuroscience Centre, Universityof Otago, Dunedin (New Zealand) (Accepted 24 May 1988)
Key words: Cysteamine; Seizure; Myoclonus; Somatostatin; Amygdaloid kindling; Cortical kindling
The effect of the somatostatin depleting substance, cysteamine (100 mg/kg, i.p.), on cortical and amygdaloid kindled seizures was investigated. Cysteamine was tested after the establishment of amygdaloid kindling (AM group) and at two different developmental stages of cortical kindling, namely 'focal-cortical' (FC group) and 'cortico-generalized' seizures (CG group). In control, non-kindled, sham operated animals, cysteamine did not induce any spike activity or myoclonus. However, in all kindled groups clustered spike bursting appeared in the cortex within 5-15 min of the injection. The kindled rats exhibited myoclonic jerks at 10 to 30 min after cysteamine injection, which coincided with the cortical spikes, and continued for about 40 min. In contrast, relatively small amounts of spiking were observed in the amygdala and this did not correlate with the myoclonus. At 4 h after cysteamine injection, the motor seizure and afterdischarge durations of the kindled seizure were prolonged in all kindled groups compared with preinjection levels. However, 24 h later the motor seizure duration and the afterdischarge duration were markedly reduced from the preinjection level in the AM and the CG groups and the tonic seizure component was suppressed in the FC group. This inhibitory effect on seizure activity lasted several days and gradually disappeared. These modifying effects of cysteamine were more marked in cortical kindled, than in amygdaloid kindled animals. The results suggest that the cortex is more sensitive to the effect of cysteamine than the amygdala, and that the effect of cysteamine on kindled seizures involves two phases. The first of these phases is excitatory and the second is inhibitory, presumably reflecting an early release-inducing and a later synthesis-inhibiting effect of cysteamine on brain somatostatin.
INTRODUCTION A n elevation in somatostatin concentration has been reported in the brain of amygdaloid kindled rats 13. This change in somatostatin content continued for at least 2 months after the cessation of kindling stimulation 13, suggesting a permanent change. In contrast, the affinity and number of somatostatin receptors are left unchanged in the cortex and the limbic areas of amygdaloid kindled rats in which somatostatin levels are remarkably elevated 1°. Further, Higuchi et al. have reported that cysteamine (2-mercaptoethylamine), which selectively and reversibly depletes somatostatin from the peripheral organs and the brain 2,7,23-26, retards the development of amygdaloid kindling 12 and temporally suppresses sei-
zure expression 24 h after administration in established amygdaloid kindled animals 9. It has also been reported that intracerebroventricular injection of somatostatin antibody suppresses amygdaloid kindled seizures 9 and that carbamazepine, which is known as the most effective anticonvulsant drug for limbic kindled seizures, induces a decrease of somatostatin content in the kindled it, but not in the naive 27, rat. These reports suggest that in the rat brain, endogenous somatostatin has an important role as a modulator of amygdaloid kindling. Recently, Cottrell and Robertson reported a further effect of cysteamine on the kindled animal 4. They demonstrated that cysteamine (200 mg/kg, i.p.) induces myoclonic jerks during the first hour after administration in amygdaloid and hippocampal
* Present address: Department of Psychiatry, Saitama Medical School, 38 Moro-hongo, Moroyama-machi, Iruma-gun, Saitama, Japan. Correspondence: D.K. Bilkey, Department of Psychology, University of Otago, P.O. Box 56, Dunedin, New Zealand. 0006-8993/88/$03.50 ~ 1988 Elsevier Science Publishers B.V. (Biomedical Division)
100
kindled rats but not in naive rats. This report, taken in conjunction with above mentioned findings, suggests that cysteamine has a proconvulsant, in addition to its anticonvulsant, effect in the limbic kindled animal and that these effects are time dependent. Cortical kindling differs from amygdaloid kindling in several respects. Somatostatin levels are elevated only in the neocortex of cortical kindled rats 12, whereas they are elevated both in the neocortex and in the limbic areas of the amygdaloid kindled rat llA3. It has also been reported that the development pattern of cortical kindling differs from amygdaloid kindling and that established cortical kindled seizures have two different components, early, brief clonictonic convulsions which are then followed by limbictype rearing and falling seizures 3. These reports suggest that the underlying mechanism in cortical kindling, at least partly, differs from amygdaloid kindling. In this report, we investigated the effects of cysteamine on the myoclonus induction and the kindled seizure expression and compared these effects for amygdaloid and cortical kindled animals. In addition, we recorded the E E G activity of the amygdala and the cortex in both amygdaloid kindled and cortical kindled animals during the occurrence of the cysteamine induced behavioral change. MATERIALS AND METHODS Male Sprague-Dawley rats, weighing 300-400 g ' at the time of surgery, were used. All rats were individually housed with a 12 h light-dark cycle. Food and water were available ad libitum. Under pentobarbital anesthesia (60 mg/kg, i.p.), each rat was implanted with tripolar electrodes in both left anterior cortex (2.5 mm anterior and 2.5 mm lateral to bregma, and 2.5 mm below the skull surface), and left amygdala (2.5 mm posterior and 4.8 mm lateral to bregma, and 8.5 mm below the skull surface). All coordinates were determined with the incisor bar 3 mm below the interaural line. The tripolar electrode consisted of 3 twisted Diamel-insulated Nichrome wires (0.18 mm diameter) with the uninsulated tips approximately 0.1 mm apart. Two screw electrodes were placed in the frontal and the occipital skull for grounding the animal and the reference electrode, respectively. All electrodes were attached by insu-
lated wires to small, gold-plated pins, which were inserted into a plastic socket and fixed on the skull by dental acrylic cement. At least 10 days recovery were allowed before the experiments began. Rats were divided into 3 groups, the cortical kindled, the amygdaloid kindled and the unstimulated control groups. Amygdaloid or cortical stimulation was delivered through two poles of the tripolar electrode while the E E G was recorded through the remaining pole and one of the screw electrodes in the skull. A kindling stimulus of biphasic square wave pulses at 100 Hz for 2 s was delivered once daily. The stimulus intensity was 400 p A for cortical kindling and 200 p A for amygdaloid kindling. Movement artifact was reduced by using a field-effect transistor in the cable end, and the signal was amplified, filtered (10-35 Hz half amplitude) and recorded throughout each kindling trial by a Grass model 79D electroencephalograph. The stage of behavior seizures 2°, the motor seizure duration, and the afterdischarge (AD) duration were observed and measured on all occasions. The test dose of 100 mg/kg of cysteamine (Sigma Inc.) was chosen, since it has been shown that this dose can deplete somatostatin content in naive rats to the same extent as the 200 mg/kg dose 7 that was used in previous reports 4-6,9. Unlike the 200 mg/kg dose, however, this dose does not affect the catecholamine content of the cortex 7. The effects of cysteamine were tested on amygdaloid kindling after animals showed 5 consecutive rearing and falling (stage 5) seizures (AM group). Cortical kindling has two different developmental stages, 'focal-cortical' seizures (FC), which are thought to represent a partially kindled stage and involve pure clonic-tonic motor convulsions, and later fully kindled 'cortico-generalized' seizures (CG), which include clonic-tonic convulsions followed by limbic-type, rearing and falling seizures 3. Therefore, cortical kindled animals were tested with cysteamine when they showed 5 consecutive pure clonic-tonic seizures (FC group) or 3 consecutive 'cortico-generalized' seizures (CG group) before testing. Myoclonic behavior and E E G activity were monitored for 90 min after cysteamine administration, and 4 h after the injection a kindling stimulation was delivered. From then on, the animals were tested daily with kindling stimulation until full recovery was seen.
101 All data were analyzed by A N O V A , two-tailed Student's t-test or p a i r e d t-test. Significant level was set at P < 0.05. W h e n all experiments were c o m p l e t e d , the rats were deeply anesthetized and their brains were perfused with 0.9% saline and 10% formalin, sliced into 40/~m sections and stained with thionin. The electrode positions were e x a m i n e d under a light microscope.
probably a form of 'partial generalized' seizure, since it cannot be elicited by cortical stimulation at threshold values until a good deal of kindling has taken place 3. The C G group of cortical kindled animals showed fully developed 'cortico-generalized' kindled seizures after 29.0 + 1.2 A D s (n = 6, range 24-32). There was no difference in the duration of A D between the A M and the C G group, while the A D duration of the F C group was significantly shorter than both the A M and the C G group after kindling establishment ( P < 0.01, see Table I).
RESULTS
Behavioral and electroencephalographic effects of cysteamine
The rate and p a t t e r n of amygdaloid and cortical kindling d e v e l o p m e n t was similar to that described previously 3'13'2°. A m y g d a l a stimulated rats devel-
Control, non-stimulated and sham o p e r a t e d animals (n = 4) which received 100 mg/kg of cysteamine showed no r e m a r k a b l e behavioral change t h r o u g h out the 90 min observation period. The E E G of control rats did not show spike activity in either the amygdala or the cortex during the preinjection period (Fig. 1 Aa). A p p r o x i m a t e l y 10 min after the
o p e d generalized stage 5 seizures after a m e a n of 7.5 + 0.8 A D s ( + S . E . M . ; n = 6, range 6 - 1 1 ) . The F C group of cortical kindled rats showed clonic convulsions at first stimulation and d e v e l o p e d c l o n i c - t o n i c convulsions after a mean of 6.2 + 1.2 A D s (n = 5, range 3 - 9 ) . The tonic phase of cortical kindling is
A A M
a ~-t
C :
.
--:~
:z::
-
; ; z
-
z
~ .
:
z
:
it
~1~I
CX
B
b
d
a
C
b
d T
~?
I l
I
_1
__1
Fig. 1. Examples of early resting EEG change in a control and a kindled rat after cysteamine injection (100 mg/kg, i.p.). A: in a control, non-kindled, sham operated rat there is no spike activity in either the amygdala (AM, upper trace of each recording) or the cortex (CX, lower trace of each recording) at the preinjection period (As). After cysteamine injection, moderate amplitude spike-like activity appeared in both regions (A b and Ac), in open channel recording (Ad) the spike-like activity was observed to be 0 activity. This was not observed during the preinjection period. B: an example from a cortical kindled rat (the cortico-generalizedgroup, see text for details) showing a very small amount of single spike activity during the preinjection period (B~). After cysteamine injection, clustered spike activity appeared and was seen predominantly in the cortex (Bb_d). Later on, myoclonic jerks (indicated by arrow heads) were exhibited, coincident with the cortical spikes (Bb_d). On the open channel recording (Bu), the cortical spike activity frequently rode on the 0 activity. Vertical bars indicate 5 mV in a, b and c, and 1 mV in d. Horizontal bars indicate 5 s in a and b, and 1 s in c and d.
102 cysteamine injection, moderate amplitude spike-like activity appeared in both regions although animals did not show any behavioral change (Fig. 1 Ab_a). The high speed, open channel recording (0.1-35 Hz half amplitude) showed that the spike-like activity was in fact 0 activity, which was seldom seen prior to the preinjection period (Fig. 1 Aa). The spontaneous behavior of kindled animals from any group could not be differentiated from control animals during the preinjection period. Within 10-30 min (mean latency 20.3 min) after the cysteamine injection, however, kindled rats exhibited myoclonic jerks. The myoclonic behavior was very similar in all 3 kindled groups. The appearance of myoclonic jerks lasted for a mean of 37 min (range 13-73 min). The E E G recording showed a very small amount of single spike activity during the preinjection period (Fig. 1 Ba), whereas clustered spike activity appeared within 5-15 min (mean 8.7 min) after the cysteamine injection (Fig. 1 Bb,c). This was predominantly in the cortex (Fig. 1 Bb), and at 5-20 min after the appearance of the spike activity, myoclonic jerks appeared, coincident with the cortical spikes (Fig. 1 Bb-d, indicated by arrow heads). On open channel recording, the cortical spike activity was observed to be true spiking that frequently rode on an underlying 0 activity (Fig. 1 Bd). This finding was c o m m o n to all kindled groups.
jerks during the observation period, the trend for a greater number of myoclonic jerks to occur in the FC group was not statistically significant (F2,1~ = 2.95, P = 0.08). The number of myoclonic jerks was not related to the amount of spike activity in any of the kindled groups.
Modifying effects of cysteamine on kindled seizures The seizure stage of the A M group was not affected 4 h after the cysteamine injection, but 24 h later 2 of 6 rats showed stage 5 seizures, 3 rats exhibited stage 4, and the remaining rat showed short A D activity but no seizure behavior. Within 5 days, all rats in the A M group recovered to show full stage
0
Total Amount of Spike Activity ( 00rain ) 100 200 300 I
i
FC
CG
÷
q
AM
~r J
Effect of cysteamine on the occurrence of myoclonic jerks and the EEG spike activity across kindled groups In the preinjection state, in all kindled groups, a very small number of spontaneous spikes appeared equally in both cortical and amygdaloid regions. After cysteamine injection, the spontaneous spiking increased at both recording sites. The number of amygdaloid spikes induced by cysteamine significantly differed between groups (F2.13 = 3.76, P < 0.05), i.e. a greater number of amygdaloid spikes were observed in the CG and the A M groups compared to the FC group (P < 0.05). The number of cortical spikes was not significantly different between groups (F2,13 = 0.90, P > 0.1). The increase in spiking was significantly greater in the cortex than in the amygdala for the A M group (P < 0.05, Fig. 2). The same tendency was also seen in the FC and the CG group of cortical kindled animals (FC group P = 0.08, CG group P = 0.08). Comparing the total number of myoclonic
400
I
Total
Numberof
50 !
Myoclonic
Jerks
100
(90min)
150 I
FC CG iiiiiiiiiiiiiiiiiiiiiii] AM
i j
Fig. 2. Total number of EEG spikes and myoclonic jerks induced by cysteamine injection (100 mg/kg) in each kindled group. In the graph for total number of EEG spikes, white and black columns indicate the spike number in the amygdala and the cortex, respectively. The number of cortical spikes was significantly greater than amygdaloid spikes in the AM group ('~¢, P < 0.05). The same tendency was seen in the FC and the CG groups but was not statistically significant. A greater number of amygdaloid spikes was observed in the CG and the AM groups compared with the FC group (+, P < 0.05). A greater number of myoclonic jerks was seen in the FC group, but this difference was not statistically significant. FC, the focal-cortical group of the cortical kindling; CG, the cortico-generalized group of the cortical kindling; AM, the amygdaloid kindled group.
103
4 ~0
~3 ~2 N
0
•
control t (~
1
2
3
4
5
6
days
Fig. 3. The mean ( - - ) and individual (0) seizure stage of amygdaloid kindled group are presented for control and days
injections. A l l kindled animals showed stable m o t o r seizure durations during the control period. The motor seizure duration was significantly different between groups during the control period (Fig. 4, F2,t4 = 46.80, P < 0.0001; A M vs C G P < 0.01, A M vs F C P < 0.0001, and C G vs F C P < 0.0001). F o u r hours after the injection, the m o t o r seizure duration did not significantly differ between groups (F2.14 = 2.53, P > 0.1). The duration was p r o l o n g e d in comparison with the control period, but this did not reach statistical significance except for the C G group ( P < 0.01, Fig. 4). In the F C group, 3 out of 5
0-5 after injection of 100 mg/kg of cysteamine. At day 0 cysteamine was injected 4 h prior to kindling stimulation (indicated by arrow).
300.
E
5 seizures (Fig. 3). T h e time course was very similar to the previous r e p o r t 9. The cortical kindled seizures included pure m o t o r convulsions 3, while the amygdaloid kindled seizures involved both focal limbic seizures and secondary generalized convulsions TM. To c o m p a r e the seizure severity of the m o t o r convulsions across groups, we m e a s u r e d the m o t o r seizure duration, i.e. the time of clonic and/or tonic c o m p o n e n t s of kindled s e i z u r e s . The control values were calculated after generating 3 - 5 consecutive kindled seizures before cysteamine sec 40'
._=
}a
,00
-10(
~,
300.
8
.=_
d .~ 2 0 O
.,e -100 ~rgr~r 0 control
tO-
1
2
;
4
;
6
days
Fig. 4. The effect of cystcamine on the motor seizure duration of the focal-cortical group of cortical kindled animals (11), the
cortico-generalized group of cortical kindled animals (&), and the amygdaloid kindled group (0). Cysteamine was injected 4 h before testing at day 0. Animals were subsequently tested every 24 h for 6 days. The vertical bar of each plot indicates standard error of mean. "A', P < 0.05; "i~r~, P < 0.01; ~--~--~, P < 0.001; compared with the control period of each group.
Fig. 5. The effect of cysteamine on the percent change of afterdischarge (AD) in the amygdala and the cortex. The symbols indicate the focal-cortical group of cortical kindled animals (11), the cortico-generalized group of cortical kindled animals (&), and the amygdaloid kindled group (0). Cysteamine was injected 4 h before testing at day 0. Animals were subsequently tested every 24 h for 6 days. The vertical bar of each plot indicates standard error of mean. ¢z, P < 0.05; ¢i'~, P < 0.01; ~-'~-'~, P < 0.001; compared with the control period of each group.
104 TABLE I
Afterdischarge duration (s) AM, afterdischarge duration in th e amygdala; CX, afterdischarge duration in the cortex; FC, focal-cortical group of cortical kindled animals; CG, cortico-generalized group of cortical kindled animals; A M , amygdaloid kindled group.
Control
Cystearnine 4h
AM
CX
FC 6.3+'1.2 a'b'** 6.7+-1.2 a'h'** CG 63.5+10.2 63.8+10.2 AM 68.9+11.4 71.1-+11.6
Day1
Day5
AM
CX
AM
23.1+-7.0~'***:~'** 109.3+6.0 94.2+13.4
25.8+7.5 a'***~b'** 109.6--+6.1 90.5+13.8
7.9+1.7 b'* 9.0+1.1 ~'* 3.8+1.2 :* 4.3+1.1 c* 34.5-+9.4 37.1-+11.9
CX
AM
CX
9.5+0.93"h'* 85.5-+21.9 74.7-+15.7
11.0+0.9 ~'~'* 86.7-+21.8 68.5+15.5
FCvsCG; b FCvsAM; c CGvsAM. *P<0.05; **P<0.01: ***P<0.001.
cases showed clonic-tonic convulsions followed by rearing (stage 4), which was never seen in the preinjection trial. The CG and AM groups, however, showed a prolongation of the motor seizure duration without a change in seizure pattern. In contrast, 24 h later the motor seizure duration of the AM and the CG groups was shortened compared to the preinjection period (AM P = 0.08, and CG P < 0.001, Fig. 4). Further, the motor seizure duration of the CG group was the same as the FC group, probably due to the disappearance of the limbic-type seizure component in all rats of the CG group. In the AM group, however, 5 out of 6 animals showed stage 4 or 5 seizures. There was no major change in the motor seizure duration of the FC group at this time (Fig. 4), but the tonic component of convulsions were almost completely inhibited. These inhibitory effects on the motor seizure disappeared gradually over several days. Four hours after the cysteamine injection, no myoclonic jerking was observed and about half of the animals in each group showed cortical spiking. Twentyfour hours later, both the spontaneous behavior and the EEG activity returned to preinjection level. There seemed to be no strict relationship between the number of myoclonic jerks or the E E G spike activity and the strength of the biphasic modifying effect of cysteamine on the kindled seizures.
Effect of cysteamine on the A D duration The effect of cysteamine on the AD duration of each group was very similar to the effects on the motor seizure duration shown in Fig. 4 (Fig. 5, Table I), except that the change in the AD duration of the AM
group was relatively marked compared to the effect on the motor seizure duration. The AD duration of both regions of the AM group was significantly prolonged at 4 h after cysteamine (P < 0.05), and at 24 h later was significantly shortened (in the amygdala P < 0.05 and in the cortex P < 0.01, Fig. 4). The later suppressing effect of cysteamine on the AD duration of the AM and the CG group lasted for several days. The change in the AD duration thus corresponded well to the change in the motor seizure duration of each group. The potentiating effect of cysteamine on the amygdaloid kindled seizure was also seen 1 h after administration. There was no significant difference in the extent of the effect 1 h after cysteamine administration and the effect at 4 h, both on the AD duration of the amygdala (Fl. 9 = 2.15, P > 0.1) and on the motor seizure duration (El, 9 = 0 . 1 8 , P > 0 . 1 ) .
Histology All amygdala electrodes were in the amygdala body, mainly in the basolateral nucleus of the amygdala, and all cortical electrodes were in the deep layers of the frontal and sensory-motor cortex. DISCUSSION
It has been reported that a'200 mg/kg dose of cysteamine inhibits seizure expression in amygdaloid or hippocampal kindled rats 4'5'9, and in animals chemically kindled by repeated pentylenetetrazol administration t. The effect of cysteamine seems to be specific to kindled seizures because the same dose does not suppress the clonic-tonic convulsion induced by
105 acute injection of pentylenetetrazol 1. Because the main effect of cysteamine is thought to be a depletion of somatostatin 2"7'23-26 and the disappearance of the antiepileptic effect of cysteamine is correlated with the recovery of somatostatin levels23, it is supposed that cysteamine reduces the somatostatin increase produced by kindling in cortex and some limbic areas 9A2'13. However, the dose of 200 mg/kg used in the previous reports appears to be higher than required purely to deplete brain somatostatin, since it is reported that this dose also influences the catecholamine content of the cortex and the limbic areas 7"26. Since catecholamines are also reported to be modulators of the kindling phenomenon 17, we tested a dose of cysteamine (100 mg/kg) which does not affect the cortical catecholamine content 7. The inhibitory effect of 100 mg/kg cysteamine on seizure stage of the limbic kindling was moderate compared with the effects described in previous reports using 200 mg/kg of cysteamine, in which it has been reported that cysteamine can completely suppress the motor seizure component 24 h after administration. The present results, however, indicated that 100 mg/kg of cysteamine still had statistically significant inhibitory effects both on the motor seizure duration and on the AD duration of limbic kindling 24 h after injection, suggesting that the antiepileptic effect of cysteamine may result from the depletion of brain somatostatin, rather than from a change in catecholamine levels. Further, cysteamine showed the same antiepileptic effect on cortical kindling and the effect was stronger than on amygdaloid kindling. Recently, Cottrell and Robertson reported that cysteamine induces myoclonic jerks in the hippocampal kindled rat, suggesting that cysteamine acts as a 'pro-convulsant' on the kindled rat 4. The same effect is also seen in the amygdaloid kindled rat (ref. 4, T. Higuchi, personal communication). We also observed the myoclonus inducing effect of cysteamine in the cortical kindled rat, which was very similar behaviorally and electroencephalographically to that seen in the amygdaloid kindled rat. Further, our results indicate that the myoclonus seen in both kindled groups coincided with cortical spiking, rather than amygdaloid spiking. These data suggest that the cysteamine induced myoclonus originates in the cortex for both kindled groups. At the same time as it induced the myoclonus, cysteamine also prolonged the
kindled motor seizure and AD durations. The latter effect on kindled seizures lasted at least 4 h after cysteamine administration. The same potentiating effect of cysteamine has also been reported for rapidly induced kindling-like seizures 16. These reports and the present results suggest that cysteamine has two different effects on kindled animals: an early onset, myoclonus inducing and kindled seizure potentiation effect, and a late onset and long-lasting kindled seizure suppression effect. The early effect of cysteamine does not correlate with the change of somatostatin concentration in the brain, as the maximum depletion of somatostatin is not completed until 1-4 h after cysteamine administration 7'23-25. Recently, Bahkit et al. reported that acute administration of cysteamine increases spontaneous release and facilitates potassium stimulated release from the hypothalamic slice in a calcium dependent manner 2. Cysteamine also increases the somatostatin content of the rat cerebrospinal fluid during the 40 min period after systemic injection, even when cortical somatostatin has been partly depleted 7. Further, somatostatin itself can produce epileptic E E G activity and seizure behavior when it is injected intracerebroventricularly8a5 or directly into the amygdala 22 or the hippocampus 21. Therefore, it is suggested that the increase of somatostatin release induced by cysteamine produces the E E G abnormality and the seizure-related behavior, myoclonus, in the kindled animal. It has previously been reported that both the cysteamine induced myoclonus and the kindlng stimulation at 4 h after the injection are important for the production of the antiepileptic effect of cysteamine seen 24 h later L6. Pitkanen et al. reported that the somatostatin levels in rat cerebrospinal fluid increase after clonic-tonic convulsions evoked by pentylenetetrazol TM or induced by chemical kindlng 19. Furthermore, by 4 h after cysteamine administration the [3H]phenylalanine incorporation into somatostatin is markedly reduced 25, suggesting that the inhibition of somatostatin synthesis is important. These reports suggest that the kindled seizure evoked 4 h after cysteamine injection could produce the release of somatostatin from various brain regions, resulting in a further depletion of somatostatin content. In conclusion, it is suggested that the cysteamine induced cortical somatostatin release, that is asso-
106 ciated with myoclonus, when added to the somatostatin release resulting from the kindled seizure elicited 4 h after cysteamine administration, encourages somatostatin depletion. These two early effects on somatostatin levels potentiate the depletion resulting from cysteamine's direct inhibition of somatostatin synthesis, thereby producing the late long-lasting inhibitory effect on kindled seizures.
REFERENCES 1 Assouline, G., Barkaie, E. and Gutnick, M.J., Cysteamine suppresses kindled seizures in pentylenetetrazol-kindled rats, Eur. J. Pharmacol., 106 (1985) 649-652. 2 Bakhit, C., Benoit, R. and Bloom, F.E., Effects of cysteamine on pro-somatostatin related peptides, Regul. Pept., 6 (1983) 169-177. 3 Burnham, W.M., Cortical and limbic kindling: similarities and differences. In K.E. Livingston and O. Hornykiewicz (Eds.), Limbic Mechanisms, Plenum, New York, 1978, pp. 507-519. 4 Cottrell, G.A. and Robertson, H.A., Induction and suppression of seizures by cysteamine in hippocampal kindled rats, Brain Research, 365 11986) 393-396. 5 Cottrell, G.A. and Robertson, H.A., Suppression of kindled seizures by cysteamine: dependence of injectionto-kindled seizure interval, Eur. J. PharmacoL, 134 (1987) 225-228. 6 Cottrell, G.A. and Robertson, H.A., Prevention of cysteamine-induced myoclonus blocks the long-term inhibition of kindled seizures, Brain Research, 412 (1987) 161-164. 7 Haroutunian, V., Mantin, R., Campbell, G.A., Tsuboyama, G.K. and Davis, K.L., Cysteamine-induced depletion of central somatostatin-like immunoreactivity: effects on behavior, learning, memory and brain neurochemistry, Brain Research, 403 (1987) 234-242. 8 Havlicek, V. and Friesen, H.G., Comparison of behavioral effects of somatostatin and beta-endorphin in animals. In E. Collu, J.R. Ducharme, A. Barbeau and J.G. Rochefort (Eds.), Central Nervous System Effects of Hypothalamic Hormones and Other Peptides, Raven, New York, 1979, pp. 381-41)2. 9 Higuchi, T., Sikand, G.S., Kato, N., Wada, J.A. and Friesen, H.G., Profound suppression of kindled seizures by cysteamine: possible role of somatostatin to kindled seizures, Brain Research, 288 11983) 359-362. 10 Higuchi, T., Kokubu, T., Sikand, G.S., Wada, J.A. and Friesen, H.G., A study of somatostatin receptors in amygdaloid-kindled rat brain, J. Neurochem., 43 11984) 1271-1276. 11 Higuchi, T., Yamazaki, O., Takazawa, A., Kato, N., Watanabe, N., Minatogawa, Y., Yamazaki, J., Ohshima, H., Nagaki, S., Igarashi, Y. and Noguchi, T., Effects of carbamazepine and valproic acid on brain immunoreactive somatostatin and gamma-aminobutyric acid in amygdaloidkindled rats, Eur. J. PharmacoL, 125 (1986) 169-175. 12 Higuchi, T., Kato, N., Noguchi, T., Friesen, H.G. and Wada, J.A., Kindling and somatostatin. In J.A. Wada (Ed.), Kindling 111, Raven, New York, 1986, pp. 349-360. 13 Kato, N., Higuchi, T., Friesen, H.G. and Wada, J.A., Changes of immunoreactive somatostatin and beta-endorphin content in rat brain after amygdaloid kindling, Life Sci., 32 (1983) 2415-2422.
ACKNOWLEDGEMENTS We thank Dr. W.C. A b r a h a m for helpful discussions. We also thank Ms. S. Mason for her excellent histological assistance, and Mr. B. Dingwall and R. Phillips for their electronic and technical help. The study was supported by the New Z e a l a n d Neurological Foundation.
14 McNamara, J.O., Kindling: an animal model of complex partial epilepsy, Ann. Neurol., Suppl., 16 (1984) $72-$76. 15 Nagaki, S,, Kato, N., Naruse, H., Takazawa, A., Ohshima, H., Higuchi, T. and Igarashi, Y., EEG changes and behavioral abnormalities elicited by intracerebroventricular administration of somatostatin and its analogs to rats, J. Jpn. Epil. Soc., 4 (1986) 75-81. 16 Perlin, J.B., Lothman, E.W. and Geary, W.A., Somatostatin augments the spread of limbic seizures from the hippocampus, Ann. NeuroL, 21 (1987) 475-480. 17 Peterson, S.L. and Albertson, T.E., Neurotransmitter and neuromodulator function in the kindled seizure and state, Prog. NeurobioL, 19 (1982) 237-270. 18 Pitkanen, A., Jolkkonen, J., Honkanen, K.-L. and Riekkinen, P., Effect of pentylenetetrazol-induced convulsions on somatostatin-like immunoreactivity in rat cerebrospinal fluid, Neuropeptides, 9 (1987) 19-24. 19 Pitkanen, A., Jolkkonen, J. and Riekkinen, P.J., Somatostatin-like immunoreactivity (SLI) in cisternal cerebrospinal fluid of rats kindled by pentylenetetrazol, Brain Research, 416 (1987) 180-182. 20 Racine, R., Burnham, W.M. and Livingston, K., The effect of procaine hydrochloride and diazepam, separately or in combination, on cortico-generalized kindled seizures, Electroencephalogr. Clin. Neurophysiol. , 47 11979) 204-212. 21 Rezek, M., Havlicek, V., Hughes, K.R. and Friesen, H., Central site of action of somatostatin (SRIF): role of hippocampus, Neuropharmacology, 15 (1976) 499-504. 22 Rezek, M., Havlicek, V., Hughes, K.R. and Friesen, H., Behavioral and motor excitation and inhibition induced by the administration of small and large doses of somatostatin into the amygdala, Neuropharmacology, 16 (1977) 157-162. 23 Sagar, S.M., Landry, D., Millard, W.J., Badger, T.M., Arnold, M.A. and Martin, J.B., Depletion of somatostatinlike immunoreactivity in the rat central nervous system by cysteamine, J. Neurosci., 2 (1982) 225-231. 24 Srikant, C.B. and Patel, Y.C., Cysteamine-induced depletion of brain somatostatin is associated with up-regulation of cerebrocortical somatostatin receptors, Endocrinology, 115 (1984) 990-995. 25 Szabo, S. and Reichlin, S., Somatostatin depletion by cysteamine: mechanism and implication for duodenal ulceration, Fed. Proc. Fed. Am. Soc. Exp. Biol., 44 (1985) 2540-2545. 26 Szabo, S., Horner, H.C., Maull, H., Schnoor, J., Chiueh, C.C. and Palkovits, M., Biochemical changes in tissue catecholamines and serotonin in duodenal ulceration caused by cysteamine or propionitrile in the rat, J. PharmacoL Exp. Ther., 240 11987) 971-978. 27 Weiss, S.R.B., Nguyen, T., Rubinow, D.R., Helke, C.J.. Narang, P.K., Post, R.M. and Jacobowitz, D,M.. Lack of effect of chronic carbamazepine on brain somatostatin in the rat, J. Neural. Transm., 48 (1987) 325-333.
99
BRE 14035
Myoclonus inducing and seizure modifying effect of cysteamine on cortical and amygdaloid kindled rats Akira Takazawa* and David K. Bilkey Department of Psychology and The Neuroscience Centre, Universityof Otago, Dunedin (New Zealand) (Accepted 24 May 1988)
Key words: Cysteamine; Seizure; Myoclonus; Somatostatin; Amygdaloid kindling; Cortical kindling
The effect of the somatostatin depleting substance, cysteamine (100 mg/kg, i.p.), on cortical and amygdaloid kindled seizures was investigated. Cysteamine was tested after the establishment of amygdaloid kindling (AM group) and at two different developmental stages of cortical kindling, namely 'focal-cortical' (FC group) and 'cortico-generalized' seizures (CG group). In control, non-kindled, sham operated animals, cysteamine did not induce any spike activity or myoclonus. However, in all kindled groups clustered spike bursting appeared in the cortex within 5-15 min of the injection. The kindled rats exhibited myoclonic jerks at 10 to 30 min after cysteamine injection, which coincided with the cortical spikes, and continued for about 40 min. In contrast, relatively small amounts of spiking were observed in the amygdala and this did not correlate with the myoclonus. At 4 h after cysteamine injection, the motor seizure and afterdischarge durations of the kindled seizure were prolonged in all kindled groups compared with preinjection levels. However, 24 h later the motor seizure duration and the afterdischarge duration were markedly reduced from the preinjection level in the AM and the CG groups and the tonic seizure component was suppressed in the FC group. This inhibitory effect on seizure activity lasted several days and gradually disappeared. These modifying effects of cysteamine were more marked in cortical kindled, than in amygdaloid kindled animals. The results suggest that the cortex is more sensitive to the effect of cysteamine than the amygdala, and that the effect of cysteamine on kindled seizures involves two phases. The first of these phases is excitatory and the second is inhibitory, presumably reflecting an early release-inducing and a later synthesis-inhibiting effect of cysteamine on brain somatostatin.
INTRODUCTION A n elevation in somatostatin concentration has been reported in the brain of amygdaloid kindled rats 13. This change in somatostatin content continued for at least 2 months after the cessation of kindling stimulation 13, suggesting a permanent change. In contrast, the affinity and number of somatostatin receptors are left unchanged in the cortex and the limbic areas of amygdaloid kindled rats in which somatostatin levels are remarkably elevated 1°. Further, Higuchi et al. have reported that cysteamine (2-mercaptoethylamine), which selectively and reversibly depletes somatostatin from the peripheral organs and the brain 2,7,23-26, retards the development of amygdaloid kindling 12 and temporally suppresses sei-
zure expression 24 h after administration in established amygdaloid kindled animals 9. It has also been reported that intracerebroventricular injection of somatostatin antibody suppresses amygdaloid kindled seizures 9 and that carbamazepine, which is known as the most effective anticonvulsant drug for limbic kindled seizures, induces a decrease of somatostatin content in the kindled it, but not in the naive 27, rat. These reports suggest that in the rat brain, endogenous somatostatin has an important role as a modulator of amygdaloid kindling. Recently, Cottrell and Robertson reported a further effect of cysteamine on the kindled animal 4. They demonstrated that cysteamine (200 mg/kg, i.p.) induces myoclonic jerks during the first hour after administration in amygdaloid and hippocampal
* Present address: Department of Psychiatry, Saitama Medical School, 38 Moro-hongo, Moroyama-machi, Iruma-gun, Saitama, Japan. Correspondence: D.K. Bilkey, Department of Psychology, University of Otago, P.O. Box 56, Dunedin, New Zealand. 0006-8993/88/$03.50 ~ 1988 Elsevier Science Publishers B.V. (Biomedical Division)
100
kindled rats but not in naive rats. This report, taken in conjunction with above mentioned findings, suggests that cysteamine has a proconvulsant, in addition to its anticonvulsant, effect in the limbic kindled animal and that these effects are time dependent. Cortical kindling differs from amygdaloid kindling in several respects. Somatostatin levels are elevated only in the neocortex of cortical kindled rats 12, whereas they are elevated both in the neocortex and in the limbic areas of the amygdaloid kindled rat llA3. It has also been reported that the development pattern of cortical kindling differs from amygdaloid kindling and that established cortical kindled seizures have two different components, early, brief clonictonic convulsions which are then followed by limbictype rearing and falling seizures 3. These reports suggest that the underlying mechanism in cortical kindling, at least partly, differs from amygdaloid kindling. In this report, we investigated the effects of cysteamine on the myoclonus induction and the kindled seizure expression and compared these effects for amygdaloid and cortical kindled animals. In addition, we recorded the E E G activity of the amygdala and the cortex in both amygdaloid kindled and cortical kindled animals during the occurrence of the cysteamine induced behavioral change. MATERIALS AND METHODS Male Sprague-Dawley rats, weighing 300-400 g ' at the time of surgery, were used. All rats were individually housed with a 12 h light-dark cycle. Food and water were available ad libitum. Under pentobarbital anesthesia (60 mg/kg, i.p.), each rat was implanted with tripolar electrodes in both left anterior cortex (2.5 mm anterior and 2.5 mm lateral to bregma, and 2.5 mm below the skull surface), and left amygdala (2.5 mm posterior and 4.8 mm lateral to bregma, and 8.5 mm below the skull surface). All coordinates were determined with the incisor bar 3 mm below the interaural line. The tripolar electrode consisted of 3 twisted Diamel-insulated Nichrome wires (0.18 mm diameter) with the uninsulated tips approximately 0.1 mm apart. Two screw electrodes were placed in the frontal and the occipital skull for grounding the animal and the reference electrode, respectively. All electrodes were attached by insu-
lated wires to small, gold-plated pins, which were inserted into a plastic socket and fixed on the skull by dental acrylic cement. At least 10 days recovery were allowed before the experiments began. Rats were divided into 3 groups, the cortical kindled, the amygdaloid kindled and the unstimulated control groups. Amygdaloid or cortical stimulation was delivered through two poles of the tripolar electrode while the E E G was recorded through the remaining pole and one of the screw electrodes in the skull. A kindling stimulus of biphasic square wave pulses at 100 Hz for 2 s was delivered once daily. The stimulus intensity was 400 p A for cortical kindling and 200 p A for amygdaloid kindling. Movement artifact was reduced by using a field-effect transistor in the cable end, and the signal was amplified, filtered (10-35 Hz half amplitude) and recorded throughout each kindling trial by a Grass model 79D electroencephalograph. The stage of behavior seizures 2°, the motor seizure duration, and the afterdischarge (AD) duration were observed and measured on all occasions. The test dose of 100 mg/kg of cysteamine (Sigma Inc.) was chosen, since it has been shown that this dose can deplete somatostatin content in naive rats to the same extent as the 200 mg/kg dose 7 that was used in previous reports 4-6,9. Unlike the 200 mg/kg dose, however, this dose does not affect the catecholamine content of the cortex 7. The effects of cysteamine were tested on amygdaloid kindling after animals showed 5 consecutive rearing and falling (stage 5) seizures (AM group). Cortical kindling has two different developmental stages, 'focal-cortical' seizures (FC), which are thought to represent a partially kindled stage and involve pure clonic-tonic motor convulsions, and later fully kindled 'cortico-generalized' seizures (CG), which include clonic-tonic convulsions followed by limbic-type, rearing and falling seizures 3. Therefore, cortical kindled animals were tested with cysteamine when they showed 5 consecutive pure clonic-tonic seizures (FC group) or 3 consecutive 'cortico-generalized' seizures (CG group) before testing. Myoclonic behavior and E E G activity were monitored for 90 min after cysteamine administration, and 4 h after the injection a kindling stimulation was delivered. From then on, the animals were tested daily with kindling stimulation until full recovery was seen.
101 All data were analyzed by A N O V A , two-tailed Student's t-test or p a i r e d t-test. Significant level was set at P < 0.05. W h e n all experiments were c o m p l e t e d , the rats were deeply anesthetized and their brains were perfused with 0.9% saline and 10% formalin, sliced into 40/~m sections and stained with thionin. The electrode positions were e x a m i n e d under a light microscope.
probably a form of 'partial generalized' seizure, since it cannot be elicited by cortical stimulation at threshold values until a good deal of kindling has taken place 3. The C G group of cortical kindled animals showed fully developed 'cortico-generalized' kindled seizures after 29.0 + 1.2 A D s (n = 6, range 24-32). There was no difference in the duration of A D between the A M and the C G group, while the A D duration of the F C group was significantly shorter than both the A M and the C G group after kindling establishment ( P < 0.01, see Table I).
RESULTS
Behavioral and electroencephalographic effects of cysteamine
The rate and p a t t e r n of amygdaloid and cortical kindling d e v e l o p m e n t was similar to that described previously 3'13'2°. A m y g d a l a stimulated rats devel-
Control, non-stimulated and sham o p e r a t e d animals (n = 4) which received 100 mg/kg of cysteamine showed no r e m a r k a b l e behavioral change t h r o u g h out the 90 min observation period. The E E G of control rats did not show spike activity in either the amygdala or the cortex during the preinjection period (Fig. 1 Aa). A p p r o x i m a t e l y 10 min after the
o p e d generalized stage 5 seizures after a m e a n of 7.5 + 0.8 A D s ( + S . E . M . ; n = 6, range 6 - 1 1 ) . The F C group of cortical kindled rats showed clonic convulsions at first stimulation and d e v e l o p e d c l o n i c - t o n i c convulsions after a mean of 6.2 + 1.2 A D s (n = 5, range 3 - 9 ) . The tonic phase of cortical kindling is
A A M
a ~-t
C :
.
--:~
:z::
-
; ; z
-
z
~ .
:
z
:
it
~1~I
CX
B
b
d
a
C
b
d T
~?
I l
I
_1
__1
Fig. 1. Examples of early resting EEG change in a control and a kindled rat after cysteamine injection (100 mg/kg, i.p.). A: in a control, non-kindled, sham operated rat there is no spike activity in either the amygdala (AM, upper trace of each recording) or the cortex (CX, lower trace of each recording) at the preinjection period (As). After cysteamine injection, moderate amplitude spike-like activity appeared in both regions (A b and Ac), in open channel recording (Ad) the spike-like activity was observed to be 0 activity. This was not observed during the preinjection period. B: an example from a cortical kindled rat (the cortico-generalizedgroup, see text for details) showing a very small amount of single spike activity during the preinjection period (B~). After cysteamine injection, clustered spike activity appeared and was seen predominantly in the cortex (Bb_d). Later on, myoclonic jerks (indicated by arrow heads) were exhibited, coincident with the cortical spikes (Bb_d). On the open channel recording (Bu), the cortical spike activity frequently rode on the 0 activity. Vertical bars indicate 5 mV in a, b and c, and 1 mV in d. Horizontal bars indicate 5 s in a and b, and 1 s in c and d.
102 cysteamine injection, moderate amplitude spike-like activity appeared in both regions although animals did not show any behavioral change (Fig. 1 Ab_a). The high speed, open channel recording (0.1-35 Hz half amplitude) showed that the spike-like activity was in fact 0 activity, which was seldom seen prior to the preinjection period (Fig. 1 Aa). The spontaneous behavior of kindled animals from any group could not be differentiated from control animals during the preinjection period. Within 10-30 min (mean latency 20.3 min) after the cysteamine injection, however, kindled rats exhibited myoclonic jerks. The myoclonic behavior was very similar in all 3 kindled groups. The appearance of myoclonic jerks lasted for a mean of 37 min (range 13-73 min). The E E G recording showed a very small amount of single spike activity during the preinjection period (Fig. 1 Ba), whereas clustered spike activity appeared within 5-15 min (mean 8.7 min) after the cysteamine injection (Fig. 1 Bb,c). This was predominantly in the cortex (Fig. 1 Bb), and at 5-20 min after the appearance of the spike activity, myoclonic jerks appeared, coincident with the cortical spikes (Fig. 1 Bb-d, indicated by arrow heads). On open channel recording, the cortical spike activity was observed to be true spiking that frequently rode on an underlying 0 activity (Fig. 1 Bd). This finding was c o m m o n to all kindled groups.
jerks during the observation period, the trend for a greater number of myoclonic jerks to occur in the FC group was not statistically significant (F2,1~ = 2.95, P = 0.08). The number of myoclonic jerks was not related to the amount of spike activity in any of the kindled groups.
Modifying effects of cysteamine on kindled seizures The seizure stage of the A M group was not affected 4 h after the cysteamine injection, but 24 h later 2 of 6 rats showed stage 5 seizures, 3 rats exhibited stage 4, and the remaining rat showed short A D activity but no seizure behavior. Within 5 days, all rats in the A M group recovered to show full stage
0
Total Amount of Spike Activity ( 00rain ) 100 200 300 I
i
FC
CG
÷
q
AM
~r J
Effect of cysteamine on the occurrence of myoclonic jerks and the EEG spike activity across kindled groups In the preinjection state, in all kindled groups, a very small number of spontaneous spikes appeared equally in both cortical and amygdaloid regions. After cysteamine injection, the spontaneous spiking increased at both recording sites. The number of amygdaloid spikes induced by cysteamine significantly differed between groups (F2.13 = 3.76, P < 0.05), i.e. a greater number of amygdaloid spikes were observed in the CG and the A M groups compared to the FC group (P < 0.05). The number of cortical spikes was not significantly different between groups (F2,13 = 0.90, P > 0.1). The increase in spiking was significantly greater in the cortex than in the amygdala for the A M group (P < 0.05, Fig. 2). The same tendency was also seen in the FC and the CG group of cortical kindled animals (FC group P = 0.08, CG group P = 0.08). Comparing the total number of myoclonic
400
I
Total
Numberof
50 !
Myoclonic
Jerks
100
(90min)
150 I
FC CG iiiiiiiiiiiiiiiiiiiiiii] AM
i j
Fig. 2. Total number of EEG spikes and myoclonic jerks induced by cysteamine injection (100 mg/kg) in each kindled group. In the graph for total number of EEG spikes, white and black columns indicate the spike number in the amygdala and the cortex, respectively. The number of cortical spikes was significantly greater than amygdaloid spikes in the AM group ('~¢, P < 0.05). The same tendency was seen in the FC and the CG groups but was not statistically significant. A greater number of amygdaloid spikes was observed in the CG and the AM groups compared with the FC group (+, P < 0.05). A greater number of myoclonic jerks was seen in the FC group, but this difference was not statistically significant. FC, the focal-cortical group of the cortical kindling; CG, the cortico-generalized group of the cortical kindling; AM, the amygdaloid kindled group.
103
4 ~0
~3 ~2 N
0
•
control t (~
1
2
3
4
5
6
days
Fig. 3. The mean ( - - ) and individual (0) seizure stage of amygdaloid kindled group are presented for control and days
injections. A l l kindled animals showed stable m o t o r seizure durations during the control period. The motor seizure duration was significantly different between groups during the control period (Fig. 4, F2,t4 = 46.80, P < 0.0001; A M vs C G P < 0.01, A M vs F C P < 0.0001, and C G vs F C P < 0.0001). F o u r hours after the injection, the m o t o r seizure duration did not significantly differ between groups (F2.14 = 2.53, P > 0.1). The duration was p r o l o n g e d in comparison with the control period, but this did not reach statistical significance except for the C G group ( P < 0.01, Fig. 4). In the F C group, 3 out of 5
0-5 after injection of 100 mg/kg of cysteamine. At day 0 cysteamine was injected 4 h prior to kindling stimulation (indicated by arrow).
300.
E
5 seizures (Fig. 3). T h e time course was very similar to the previous r e p o r t 9. The cortical kindled seizures included pure m o t o r convulsions 3, while the amygdaloid kindled seizures involved both focal limbic seizures and secondary generalized convulsions TM. To c o m p a r e the seizure severity of the m o t o r convulsions across groups, we m e a s u r e d the m o t o r seizure duration, i.e. the time of clonic and/or tonic c o m p o n e n t s of kindled s e i z u r e s . The control values were calculated after generating 3 - 5 consecutive kindled seizures before cysteamine sec 40'
._=
}a
,00
-10(
~,
300.
8
.=_
d .~ 2 0 O
.,e -100 ~rgr~r 0 control
tO-
1
2
;
4
;
6
days
Fig. 4. The effect of cystcamine on the motor seizure duration of the focal-cortical group of cortical kindled animals (11), the
cortico-generalized group of cortical kindled animals (&), and the amygdaloid kindled group (0). Cysteamine was injected 4 h before testing at day 0. Animals were subsequently tested every 24 h for 6 days. The vertical bar of each plot indicates standard error of mean. "A', P < 0.05; "i~r~, P < 0.01; ~--~--~, P < 0.001; compared with the control period of each group.
Fig. 5. The effect of cysteamine on the percent change of afterdischarge (AD) in the amygdala and the cortex. The symbols indicate the focal-cortical group of cortical kindled animals (11), the cortico-generalized group of cortical kindled animals (&), and the amygdaloid kindled group (0). Cysteamine was injected 4 h before testing at day 0. Animals were subsequently tested every 24 h for 6 days. The vertical bar of each plot indicates standard error of mean. ¢z, P < 0.05; ¢i'~, P < 0.01; ~-'~-'~, P < 0.001; compared with the control period of each group.
104 TABLE I
Afterdischarge duration (s) AM, afterdischarge duration in th e amygdala; CX, afterdischarge duration in the cortex; FC, focal-cortical group of cortical kindled animals; CG, cortico-generalized group of cortical kindled animals; A M , amygdaloid kindled group.
Control
Cystearnine 4h
AM
CX
FC 6.3+'1.2 a'b'** 6.7+-1.2 a'h'** CG 63.5+10.2 63.8+10.2 AM 68.9+11.4 71.1-+11.6
Day1
Day5
AM
CX
AM
23.1+-7.0~'***:~'** 109.3+6.0 94.2+13.4
25.8+7.5 a'***~b'** 109.6--+6.1 90.5+13.8
7.9+1.7 b'* 9.0+1.1 ~'* 3.8+1.2 :* 4.3+1.1 c* 34.5-+9.4 37.1-+11.9
CX
AM
CX
9.5+0.93"h'* 85.5-+21.9 74.7-+15.7
11.0+0.9 ~'~'* 86.7-+21.8 68.5+15.5
FCvsCG; b FCvsAM; c CGvsAM. *P<0.05; **P<0.01: ***P<0.001.
cases showed clonic-tonic convulsions followed by rearing (stage 4), which was never seen in the preinjection trial. The CG and AM groups, however, showed a prolongation of the motor seizure duration without a change in seizure pattern. In contrast, 24 h later the motor seizure duration of the AM and the CG groups was shortened compared to the preinjection period (AM P = 0.08, and CG P < 0.001, Fig. 4). Further, the motor seizure duration of the CG group was the same as the FC group, probably due to the disappearance of the limbic-type seizure component in all rats of the CG group. In the AM group, however, 5 out of 6 animals showed stage 4 or 5 seizures. There was no major change in the motor seizure duration of the FC group at this time (Fig. 4), but the tonic component of convulsions were almost completely inhibited. These inhibitory effects on the motor seizure disappeared gradually over several days. Four hours after the cysteamine injection, no myoclonic jerking was observed and about half of the animals in each group showed cortical spiking. Twentyfour hours later, both the spontaneous behavior and the EEG activity returned to preinjection level. There seemed to be no strict relationship between the number of myoclonic jerks or the E E G spike activity and the strength of the biphasic modifying effect of cysteamine on the kindled seizures.
Effect of cysteamine on the A D duration The effect of cysteamine on the AD duration of each group was very similar to the effects on the motor seizure duration shown in Fig. 4 (Fig. 5, Table I), except that the change in the AD duration of the AM
group was relatively marked compared to the effect on the motor seizure duration. The AD duration of both regions of the AM group was significantly prolonged at 4 h after cysteamine (P < 0.05), and at 24 h later was significantly shortened (in the amygdala P < 0.05 and in the cortex P < 0.01, Fig. 4). The later suppressing effect of cysteamine on the AD duration of the AM and the CG group lasted for several days. The change in the AD duration thus corresponded well to the change in the motor seizure duration of each group. The potentiating effect of cysteamine on the amygdaloid kindled seizure was also seen 1 h after administration. There was no significant difference in the extent of the effect 1 h after cysteamine administration and the effect at 4 h, both on the AD duration of the amygdala (Fl. 9 = 2.15, P > 0.1) and on the motor seizure duration (El, 9 = 0 . 1 8 , P > 0 . 1 ) .
Histology All amygdala electrodes were in the amygdala body, mainly in the basolateral nucleus of the amygdala, and all cortical electrodes were in the deep layers of the frontal and sensory-motor cortex. DISCUSSION
It has been reported that a'200 mg/kg dose of cysteamine inhibits seizure expression in amygdaloid or hippocampal kindled rats 4'5'9, and in animals chemically kindled by repeated pentylenetetrazol administration t. The effect of cysteamine seems to be specific to kindled seizures because the same dose does not suppress the clonic-tonic convulsion induced by
105 acute injection of pentylenetetrazol 1. Because the main effect of cysteamine is thought to be a depletion of somatostatin 2"7'23-26 and the disappearance of the antiepileptic effect of cysteamine is correlated with the recovery of somatostatin levels23, it is supposed that cysteamine reduces the somatostatin increase produced by kindling in cortex and some limbic areas 9A2'13. However, the dose of 200 mg/kg used in the previous reports appears to be higher than required purely to deplete brain somatostatin, since it is reported that this dose also influences the catecholamine content of the cortex and the limbic areas 7"26. Since catecholamines are also reported to be modulators of the kindling phenomenon 17, we tested a dose of cysteamine (100 mg/kg) which does not affect the cortical catecholamine content 7. The inhibitory effect of 100 mg/kg cysteamine on seizure stage of the limbic kindling was moderate compared with the effects described in previous reports using 200 mg/kg of cysteamine, in which it has been reported that cysteamine can completely suppress the motor seizure component 24 h after administration. The present results, however, indicated that 100 mg/kg of cysteamine still had statistically significant inhibitory effects both on the motor seizure duration and on the AD duration of limbic kindling 24 h after injection, suggesting that the antiepileptic effect of cysteamine may result from the depletion of brain somatostatin, rather than from a change in catecholamine levels. Further, cysteamine showed the same antiepileptic effect on cortical kindling and the effect was stronger than on amygdaloid kindling. Recently, Cottrell and Robertson reported that cysteamine induces myoclonic jerks in the hippocampal kindled rat, suggesting that cysteamine acts as a 'pro-convulsant' on the kindled rat 4. The same effect is also seen in the amygdaloid kindled rat (ref. 4, T. Higuchi, personal communication). We also observed the myoclonus inducing effect of cysteamine in the cortical kindled rat, which was very similar behaviorally and electroencephalographically to that seen in the amygdaloid kindled rat. Further, our results indicate that the myoclonus seen in both kindled groups coincided with cortical spiking, rather than amygdaloid spiking. These data suggest that the cysteamine induced myoclonus originates in the cortex for both kindled groups. At the same time as it induced the myoclonus, cysteamine also prolonged the
kindled motor seizure and AD durations. The latter effect on kindled seizures lasted at least 4 h after cysteamine administration. The same potentiating effect of cysteamine has also been reported for rapidly induced kindling-like seizures 16. These reports and the present results suggest that cysteamine has two different effects on kindled animals: an early onset, myoclonus inducing and kindled seizure potentiation effect, and a late onset and long-lasting kindled seizure suppression effect. The early effect of cysteamine does not correlate with the change of somatostatin concentration in the brain, as the maximum depletion of somatostatin is not completed until 1-4 h after cysteamine administration 7'23-25. Recently, Bahkit et al. reported that acute administration of cysteamine increases spontaneous release and facilitates potassium stimulated release from the hypothalamic slice in a calcium dependent manner 2. Cysteamine also increases the somatostatin content of the rat cerebrospinal fluid during the 40 min period after systemic injection, even when cortical somatostatin has been partly depleted 7. Further, somatostatin itself can produce epileptic E E G activity and seizure behavior when it is injected intracerebroventricularly8a5 or directly into the amygdala 22 or the hippocampus 21. Therefore, it is suggested that the increase of somatostatin release induced by cysteamine produces the E E G abnormality and the seizure-related behavior, myoclonus, in the kindled animal. It has previously been reported that both the cysteamine induced myoclonus and the kindlng stimulation at 4 h after the injection are important for the production of the antiepileptic effect of cysteamine seen 24 h later L6. Pitkanen et al. reported that the somatostatin levels in rat cerebrospinal fluid increase after clonic-tonic convulsions evoked by pentylenetetrazol TM or induced by chemical kindlng 19. Furthermore, by 4 h after cysteamine administration the [3H]phenylalanine incorporation into somatostatin is markedly reduced 25, suggesting that the inhibition of somatostatin synthesis is important. These reports suggest that the kindled seizure evoked 4 h after cysteamine injection could produce the release of somatostatin from various brain regions, resulting in a further depletion of somatostatin content. In conclusion, it is suggested that the cysteamine induced cortical somatostatin release, that is asso-
106 ciated with myoclonus, when added to the somatostatin release resulting from the kindled seizure elicited 4 h after cysteamine administration, encourages somatostatin depletion. These two early effects on somatostatin levels potentiate the depletion resulting from cysteamine's direct inhibition of somatostatin synthesis, thereby producing the late long-lasting inhibitory effect on kindled seizures.
REFERENCES 1 Assouline, G., Barkaie, E. and Gutnick, M.J., Cysteamine suppresses kindled seizures in pentylenetetrazol-kindled rats, Eur. J. Pharmacol., 106 (1985) 649-652. 2 Bakhit, C., Benoit, R. and Bloom, F.E., Effects of cysteamine on pro-somatostatin related peptides, Regul. Pept., 6 (1983) 169-177. 3 Burnham, W.M., Cortical and limbic kindling: similarities and differences. In K.E. Livingston and O. Hornykiewicz (Eds.), Limbic Mechanisms, Plenum, New York, 1978, pp. 507-519. 4 Cottrell, G.A. and Robertson, H.A., Induction and suppression of seizures by cysteamine in hippocampal kindled rats, Brain Research, 365 11986) 393-396. 5 Cottrell, G.A. and Robertson, H.A., Suppression of kindled seizures by cysteamine: dependence of injectionto-kindled seizure interval, Eur. J. PharmacoL, 134 (1987) 225-228. 6 Cottrell, G.A. and Robertson, H.A., Prevention of cysteamine-induced myoclonus blocks the long-term inhibition of kindled seizures, Brain Research, 412 (1987) 161-164. 7 Haroutunian, V., Mantin, R., Campbell, G.A., Tsuboyama, G.K. and Davis, K.L., Cysteamine-induced depletion of central somatostatin-like immunoreactivity: effects on behavior, learning, memory and brain neurochemistry, Brain Research, 403 (1987) 234-242. 8 Havlicek, V. and Friesen, H.G., Comparison of behavioral effects of somatostatin and beta-endorphin in animals. In E. Collu, J.R. Ducharme, A. Barbeau and J.G. Rochefort (Eds.), Central Nervous System Effects of Hypothalamic Hormones and Other Peptides, Raven, New York, 1979, pp. 381-41)2. 9 Higuchi, T., Sikand, G.S., Kato, N., Wada, J.A. and Friesen, H.G., Profound suppression of kindled seizures by cysteamine: possible role of somatostatin to kindled seizures, Brain Research, 288 11983) 359-362. 10 Higuchi, T., Kokubu, T., Sikand, G.S., Wada, J.A. and Friesen, H.G., A study of somatostatin receptors in amygdaloid-kindled rat brain, J. Neurochem., 43 11984) 1271-1276. 11 Higuchi, T., Yamazaki, O., Takazawa, A., Kato, N., Watanabe, N., Minatogawa, Y., Yamazaki, J., Ohshima, H., Nagaki, S., Igarashi, Y. and Noguchi, T., Effects of carbamazepine and valproic acid on brain immunoreactive somatostatin and gamma-aminobutyric acid in amygdaloidkindled rats, Eur. J. PharmacoL, 125 (1986) 169-175. 12 Higuchi, T., Kato, N., Noguchi, T., Friesen, H.G. and Wada, J.A., Kindling and somatostatin. In J.A. Wada (Ed.), Kindling 111, Raven, New York, 1986, pp. 349-360. 13 Kato, N., Higuchi, T., Friesen, H.G. and Wada, J.A., Changes of immunoreactive somatostatin and beta-endorphin content in rat brain after amygdaloid kindling, Life Sci., 32 (1983) 2415-2422.
ACKNOWLEDGEMENTS We thank Dr. W.C. A b r a h a m for helpful discussions. We also thank Ms. S. Mason for her excellent histological assistance, and Mr. B. Dingwall and R. Phillips for their electronic and technical help. The study was supported by the New Z e a l a n d Neurological Foundation.
14 McNamara, J.O., Kindling: an animal model of complex partial epilepsy, Ann. Neurol., Suppl., 16 (1984) $72-$76. 15 Nagaki, S,, Kato, N., Naruse, H., Takazawa, A., Ohshima, H., Higuchi, T. and Igarashi, Y., EEG changes and behavioral abnormalities elicited by intracerebroventricular administration of somatostatin and its analogs to rats, J. Jpn. Epil. Soc., 4 (1986) 75-81. 16 Perlin, J.B., Lothman, E.W. and Geary, W.A., Somatostatin augments the spread of limbic seizures from the hippocampus, Ann. NeuroL, 21 (1987) 475-480. 17 Peterson, S.L. and Albertson, T.E., Neurotransmitter and neuromodulator function in the kindled seizure and state, Prog. NeurobioL, 19 (1982) 237-270. 18 Pitkanen, A., Jolkkonen, J., Honkanen, K.-L. and Riekkinen, P., Effect of pentylenetetrazol-induced convulsions on somatostatin-like immunoreactivity in rat cerebrospinal fluid, Neuropeptides, 9 (1987) 19-24. 19 Pitkanen, A., Jolkkonen, J. and Riekkinen, P.J., Somatostatin-like immunoreactivity (SLI) in cisternal cerebrospinal fluid of rats kindled by pentylenetetrazol, Brain Research, 416 (1987) 180-182. 20 Racine, R., Burnham, W.M. and Livingston, K., The effect of procaine hydrochloride and diazepam, separately or in combination, on cortico-generalized kindled seizures, Electroencephalogr. Clin. Neurophysiol. , 47 11979) 204-212. 21 Rezek, M., Havlicek, V., Hughes, K.R. and Friesen, H., Central site of action of somatostatin (SRIF): role of hippocampus, Neuropharmacology, 15 (1976) 499-504. 22 Rezek, M., Havlicek, V., Hughes, K.R. and Friesen, H., Behavioral and motor excitation and inhibition induced by the administration of small and large doses of somatostatin into the amygdala, Neuropharmacology, 16 (1977) 157-162. 23 Sagar, S.M., Landry, D., Millard, W.J., Badger, T.M., Arnold, M.A. and Martin, J.B., Depletion of somatostatinlike immunoreactivity in the rat central nervous system by cysteamine, J. Neurosci., 2 (1982) 225-231. 24 Srikant, C.B. and Patel, Y.C., Cysteamine-induced depletion of brain somatostatin is associated with up-regulation of cerebrocortical somatostatin receptors, Endocrinology, 115 (1984) 990-995. 25 Szabo, S. and Reichlin, S., Somatostatin depletion by cysteamine: mechanism and implication for duodenal ulceration, Fed. Proc. Fed. Am. Soc. Exp. Biol., 44 (1985) 2540-2545. 26 Szabo, S., Horner, H.C., Maull, H., Schnoor, J., Chiueh, C.C. and Palkovits, M., Biochemical changes in tissue catecholamines and serotonin in duodenal ulceration caused by cysteamine or propionitrile in the rat, J. PharmacoL Exp. Ther., 240 11987) 971-978. 27 Weiss, S.R.B., Nguyen, T., Rubinow, D.R., Helke, C.J.. Narang, P.K., Post, R.M. and Jacobowitz, D,M.. Lack of effect of chronic carbamazepine on brain somatostatin in the rat, J. Neural. Transm., 48 (1987) 325-333.