A N INVESTIGATION OF 1 HE ROLE OF A M Y G D A LO ID c*-2
! ADRENOCEPTORS IN THE K IN D LIN G OF SEIZURES
' by
i Marc Roger Pelletier
B .A ., University o f Guelph, 1984 M .A ., Lakehead University, 1988
A Dissertation Submitted in Partial Fulfilm ent o f the A C C f * P T P I ) Requirements for the Degree o f
FACULTY 01 (.R A D IJA IL C fU O T.S DOCTOR OF PHILOSOPHY in the Department o f Psychology
r a ma L — We accept this dissertation as conforming
to the required standard
Dr, M, E. Corcoran, Supervisor (Department o f Psychology)
Dr. Ronald Skelton, Departmental Member (Department o f Psychology)
Dr, Esther Strauss, Departmental Member (Department o f Psychology)
Dr. George O. Mackie, Outside Member (Department o f Biology)
ternal Examiner (University o f British Columbia)
® MARC ROGER PELLETIER, 1992 University o f Victoria
A ll rights reserved, Dissertation may not be reproduced in whole or in part, by photocopying or other means, without the permission o f the author,
u .Supervisor: Dr, M. B. Corcoran
ABSTRACT
It has been reported previously that systemic administration o f clonidine, an agonist o f a-2 receptors for noradrenaline, significantly retards amygdaloid kindling, by delaying the emergence from partial seizure, Conversely, systemic administration o f «-2 antagonists has been reported to facilitate amygdaloid kindling, The
experiments 1 conducted attempted to discover whether oc-2 adrenoceptors in the amygdala participated in these effects, 1 examined the effect o f either systemic administration (i,p.) or intraamygdaloid infusions o f a variety o f noradrenergic drugs ou the kindling o f seizures with electrical stimulation o f the amygdala, Rats received either low-frequency stimulation o f the amygdala, to induce rapid kindling, or
conventional high-frequency stimulation, Drugs and electrical stimulation were administered once every h8 hrs, 1 observed a significant retardation of kindling in rats receiving i,p. injections o f clonidine (0,1 mg/kg) or unilateral infusions of clonidine in concentrations o f 10‘7 to 10'1 M , regardless o f the stimulation frequency. 'The prophylactic effect was due to a delay in the progression out o f partial seizure. I observed similar effects with infusions o f xylazine, also an a-2 adrenoceptor agonist, The effect was specific to the amygdala/pyriform region, because infusions o f clonidine dorsal lo the amygdala were without effect, Power spectral analysis o f the A D from Ihe stimulated and the contralateral amygdala during the initial occurrence o f bilateral AD failed to reveal differences attributable to clonidine, Therefore, clonidine might retard kindling by modifying the propagation o f AD from the
stimulated amygdala to a midbrain or pontine brainstem area critical, for the
expression o f generalized seizures. Clonidine had no effect on established generalized seizures, suggesting that it was producing a genuine prophylactic effect against
kindling. Unexpectedly, intraamygdaloid infusions o f either ida/oxan, yohimbine, or SK & F 104856, antagonists o f a-2 receptors, failed to accelerate kindling.
Simultaneous infusion o f idazoxan blocked clonidine’ s prophylactic effect, which suggests strongly that this effect was mediated at the «>2 adrenoceptor. Blockade o f amygdaloid a-l. adrenoce .s with corynanllvine failed to affect kindling.
I conclude that the population o f tv-2 adrenoceptors in the amygdala/pyriform region contributes to the antiepileptogenic effect observed after systemic
administration o f clonidine and that the facilitation o f kindling observed after systemic administration o f cv-2 antagonists reported previously may have been mediated by the blockade o f a population o f a -2 adrenoceptors in addition to, or outside of, the amygdala/pyriform region.
Examiners:
D r M . E. Corcoran, Supervisor (Department o f Psychology)
D r. Ronald Skelton, Departmental Member (Department o f Psychology)
D r. Esther Strauss, Departmental Member (Department o f Psychology)
D r. GcorgC'O. MackieyOutsicl'y'MymbcrJDcparlrncnl o f Biology)
T A B L E O F C O N T E N TS
Page
11tie Page. . . i
Abstract. . . . .. . .. . ... . .. . . ... n Table o f Contents... ... ... ... ...iv
List o f Tables, ... v iii List o f Figures. ... ... ... ... . .ix
Acknowledgements. ... ... ... ... xi
Dedication ... ..x ii Introduction Epilepsy ... , . ... ....1
Kindling ... 2
A brief history o f noradrenaline.,.,.. ... 7
Noradrenergic innervation ... ... ... ... J 0 Lifecycle of noradrenaline ... 12
Classification o f noradrenergic receptors... 20
Molecular biology o f tv-2 adrenoceptors ... .21
The amygdala... 25
Rationale., ... ....2 9 Experiments General Method ... 32 Experiment 1 - the effect o f systemic clonidine on low-frequency
amygdaloid k in d lin g ,,,...,... ... ...36
Method ... ,..3(i
R e s t i I t s , 36
Discussion... ... ... ... ... ...41 Experiment 2 - the effect o f intra-arnygdaloid clonidine infusions on low-frequency amygdaloid kindling. ... ... ... .44
M e th o d,..,... 44 Results... ... ... 46 Discussion... .. ... ... ,.,,.,...6 0 Experiment 3 - the effect o f clonidine infusions dorsal to the amygdala on
low-frequency amygdaloid kindling ... 66
Method . 66
Results ... 67
D iscu ssio n .,,..,,...,... 67
Experiment 4 - Rover Spectral A nalysis,.,. ... 69
Method... ,,.,,.7 0
Results ... 72
Discussion S3
Experiment 5 - the effect o f intraamygdaloid clonidine infusions on
high-frequency amygdaloid kindling... 85
M e th o d .,,... ... 85
Results ... 86
Discussion* ,92
VI
frequency amygdaloid kindling,, , , ... 94
M e t h o d m, , , , , , , , , I , , , , , , 95 Results ,,,,,,,,9 5 Discussion ... ...101
Experiment 7 - the effect o f intraamygdaloid idazoxan infusions on high-frequency amygdaloid kindling...103
M e th o d....,... 103
Results ... . ...104
Discussion,,.,... .,...,.,.1 0 9 Experiment 8 - the effect o f intraamygdaloid yohimbine infusions on high-frequency amygdaloid kindling, ... , ... 112
Method,... ... ,,... 112
Results ... . ... ...11?
Discussion ... ,,,.,.,1 1 8 Experiment 9 - the effect o f intraamygdaloid SK&F 104856 infusions on high-frequency amygdaloid kindling. ... 120
M ethod.,... 120
Results ... 121
Discussion... ...125
Experiment 10 - the effect o f intraamygdaloid clonidine/idazoxan infusions on high-frequency amygdaloid kindling, ... ,,.,....,,126
M e th o d .,,,..,... ... , , ... .,,126
Results ... ,.,,...1 2 7
VII
Experiment 11 - the effect o f intraamygdaloid corynamliine infusions on high-frequency amygdaloid kindling , ... ,,...,,..1.1.1
Method ...,1 ,14
Results, * . . . * * ... ......1,14
Discussion... ,.,.1,19 General Discussion,, ..., , ... ,,,,...1 4 0
a-2 agonists ... ,,..,..141
a-2 antagonists ... ... ... ,.., 14o Conclusion... .153 References... ...154 Appendix A ... 179 Appendix B ... 180 Appendix C ... ...,,,,..,...1 8 2 Appendix D , . . . . ... , , , , , ... 188 Appendix B . ... 197
viii L IS T O F TA B LES
Table 1: Systemic clonidine and low-frequency kindling... * ,40 Table 2: Intraamygdaloid clonidine infusions and low-frequency kindling,.58 Table 3‘ Intraamygdaloid clonidine infusions and high-frequency kindling.90 Table 4: Intraamygdaloid xylazine infusions and low-frequency kindling...99 Table 5; Intraamygdaloid idazoxan infusions and high-frequency kindling.107 Table 6; Intraamygdaloid yohimbine infusions and high-frequency kindling..., 116 Table 7; Intraamygdaloid SK&F 104856 infusions and high-frequency
k in d lin g .,,,, ... 124
Table 8; Intraamygdaloid infusions o f a clonidine/idazoxan cocktail on
nigh-frequency kindling...,, . . , ... 130 Table 9; Intraamygdaloid corynanthine infusions and high-frequency
IX
L IS T O F F IG U R E S
BlgS
Figure 1; Systemic clonidine and low-frequency kindling ... 39
Figure 2; Photomicrograph o f bilateral BLA chemitrode placements ,.,.,,,4 8 Figure 3: intraamygdaloid clonidine infusions and low-frequcney kindling ,... 51
Figure 4: Intraamygdaloid clonidine dose-response... ,53
Figure 5; Tripolar EEG - partial seizure, ... 55
Figure 6: Tripolar EEG - generalized seizure... ... ... ... ,,57
Figure 7; Percent absolute spectral power o f A D in stimulated amygdala... ,..,,.75
Figure 8: Percent absolute spectral power o f A D in contralateral amygdala... 77
Figure 9: Absolute spectral power (baseline E E G )..,... ... 79
Figure 10: Percent absolute spectral power for baseline EEG and EEG after kind ling ... ... ,,,82
Figure 11: Intraamygdaloid clonidine infusions and high-frequency k in d lin g .,... , , , , ... 89
Figure 12: Intraamygdaloid xylazine infusions and low-frequency kindling,,., ... 98
Figure 13: Et'raamygdaloid idazoxan infusions and high-frequency kindling ... , , , , , , , , , , ... ...106
Figure 14; Intraamygdaloid yohimbine infusions and high-frequency kindling ... ... ... ... Figure 15: Intraamygdaloid SK&F 104856 infusions and high-frequency kindling... ... ..., ... . Figure 16; Intraamygdaloid infusions o f a clonidine/idazoxan cocktail and
high-frequency kindling ... ...,,, Figure 17: Intraamygdaloid corynanthine infusions and high-frequency
Acknowledgements
I acknowledge the contribution to this dissertation o f Drs. Michael H.
Corcoran, Ron Skelton, George 0 . Maekie, Dorothy Paul, and Nancy Sherwood, I thank R. Duncan Kirkby, Chris Darby, Don Pierce, and Tom Gore for technical assistance. I thank also the members o f the Animal Care U nit and SinithKline Beecham for their donation o f SK&F 104856,
Xll
Dedication
I w ill begin my dissertation by introducing topics that arc germane to the appreciation o r the rationale and interpretation o f the series o f experiments described below. These topics include epilepsy, a pathological brain condition; kindling, an experimental representation o f human temporal lobe epilepsy; noradrenaline (NA), a putative neurotransmitter; noradrenergic receptors, binding sites fo r N A, with an emphasis on the ot-2 subtype; and the amygdala, a temporal lobe structure critical for the development o f seizure activity. I conclude this section by presenting a rationale fo r the experiments I conducted.
Epilepsy
Known as the “ sacred disease," epilepsy has been described since the time o f Hippocrates, approximately 460 - 374 B.C ., (Adams, 1929), More accurately, the epilepsies represent a diverse range o f spontaneously recurring self-sustained paroxysmal disturbances o f brain function with equally diverse etiologies and mechanisms. A hallmark o f the neural disturbance associated with epilepsy is excessive neuronal excitation, disinhibition, or both. Estimates o f the incidence o f this group o f clinical disorders range from 0.3 - 0.6 percent (Hauser, 1978) to approximately 2.0 percent (Hopkins, 1987a) o f the population at some stage in life, Categorization o f the epilepsies is difficu lt and requires reassessment as scientific advancements dictate (Aird, Masland, & Woodbury, 1989); however, a major distinction can be made between those epilepsies that have an identifiable focus, referred to as either focal or secondary, and those in which a focus is unidentifiable,
2
referred to as primary (Dichter & Ayala, 1987; Jefferys, 1990), The current international classification o f the epilepsies is presented in Appendix A.
Focal seizures can be induced experimentally both in vivo and in vitro by a variety Of procedures. Examples o f some experimental representations o f the epilepsies are presented in Appendix B. Kindling fc a phenomenon by which repeated, brief presentation o f an initially subconvulsant stimulus (electrical or chemical) to various brain regions results in a progressive and permanent enhancement o f seizure susceptibility. By virtue o f its focal origin and the progressive recruitment o f distant brain regions, kindling is thought to be
representative o f complex partial seizures (Cain, 1992), also referred to as secondary generalized epilepsy. The earliest examples o f a kindling-like effect were reported by Tatum and Seevers (1929) and Downs and Eddy (1932). These researchers reported
that repeated administration o f cocaine in the monkey, dog, and cat produced a progressive and persistent increase in epileptiform responses, Almost three decades later, a kindling-like phenomenon was reported by Alonso-DeFlorida and Delgado (1958), who described the electroencephalographic (EEG) profile in cats after repetitive elecrrical stimulation o f the amygdala. The term kindling was introduced by Goddard, McIntyre and Leech (1969), who also recognized the progressive nature, permanence, and potential heuristic o f this form o f plasticity in the brain, A point o f
3
historical interest warrants mentioning at this point, Goddard and his students were initia lly concerned with the effect that electrical stimulation o f the amygdala produced in the performance o f rats in various learning tasks, The convulsions that developed in the rats that had received many stimulations were initially considered to be both a source o f annoyance and a confounding variable (Racine, 1978),
In early studies directed at establishing the fundamental parameters o f kindling, Racine (1972a) demonstrated that when the EEG was recorded in a brain area stimulated with an electrical current o f sufficient intensity (typically 40 - 100 nA in the amygdala, personal observation), epileptiform afterdischarge (AD ) was evoked. M ore important, Racine (1972a) realized that AD was the critical factor determining the seizure susceptibility o f different brain regions. The rate o f kindling is typically represented as the number o f days o f AD to evoke a generalized seizure, The first brain area studied systematically was the amygdala (Cain, 1992), a mesial temporal lobe structure discussed in greater detail below. Kindling o f seizures in rats with amygdaloid stimulation is rapid (approximately 9 - 12 days o f AD ), reliable, and persistent (Cain, 1992).
Two principal features o f kindling that occur progressively during the kindling process are the reduction in the threshold current required to evoke AD , and the elaboration (in duration, complexity, frequency) and propagation o f AD throughout the brain, which is correlated with the development o f behavioral convulsions (Cain,
1992; Corcoran, 1988). In the rat, convulsions typically progress through 5 stages (Racine, 1972b): stage 1, mouth and facial twitches; stage 2, clonic head movements
4
(nodding;; stage 3, unilateral forelimb clonus; stage 4, clonic rearing; and, stage 5, clonic rearing with loss o f postural control (falling).
As stated above, epileptic seizures occur spontaneously and are self-sustained. Therefore, for kindling to qualify as an externally valid representation o f human epilepsy, kindling must possess the capacity to produce spontaneous, self-sustained seizures. There was initial difficulty demonstrating kindling-induced spontaneous seizures. This failure has been attributed to the premature cessation o f the stimulation regimen (Pine!, Mucha, & Phillips, 1975). For example, conventional kindling experiments are typically concluded after 3 to 10 generalized seizures have been evoked. I f the amygdala is considered to be the site o f stimulation, this would represent a total o f approximately 20 stimulations. In contrast to this, a mean o f 197 (range - 88 - 293) amygdaloid stimulations was required to produce spontaneous seizures in rats (Pinel & Rovner, 1978a). Kindling-induced spontaneous seizures have been observed in cats (Qotman, 1934; Shouse, King, Langer, Vreeker, King, & Richkind, 1990; Wada, Sato, 8c Corcoran, 1974), rats (Pinel et al., 1975; Pinel & Rovner, 1978a; Pinel 8c Rovner, 1978b), Senegalese baboons (Wada, Osawa, & Mizoguchi, 1975; Wada & Osawa, 1976), and dogs (Wauquier, Ashton, 8c Melis,
1979).
A role for interictal discharge (UD) has been suggested for the development o f spontaneous seizures. An increase in the incidence o f IID is thought to be important in the development o f spontaneous seizures (Pinel et al., 1975; Pinel 8c Rovner, 1978a; 1978b; Wada et al., 1974); however, no relation has also been reported
(Gotman, 1984), Kindling-induced spontaneous seizures are similar to kindled
seizures in that both are persistent (Pinel & Rovner, 1978b), but they d iffer somewhat in their behavioral expression; Spontaneous seizures arc more variable, with stage 1 seizures being the predominant class (Pinel et a l, 1975; Pine! & Rovner, 1978a; 1978b), There is evidence that kindling and kindling-induced spontaneous seizures may be governed by distinct mechanisms (Pinel, 1983; Pinel & Rovner, 1978a; 1978b),
During kindling I have observed on a few occasions spontaneous seizures (stage 3 - 4) in rats both in the home cage and during routine handling. In a more systematic attempt to produce spontaneous seizures with low-frequency amygdaloid stimulation, I evoked upwards o f 100 generalized seizures in a group o f rats
(unpublished observations). I failed to observe spontaneous seizures with this regimen o f stimulation; however, I did observe stages o f seizures more advanced than included in Racine’ s (1972b) 5 stage scale, I observed multiple stage 5 seizures, rolling over on to the back followed by tonic limb extension, and running fits, which I refer to as "agitated running." The bouts o f agitated running 1 observed were arrested briefly and intermittently by abrupt freezing, in which the rat maintained a fixed wide-eyed gaze. The rats typically ran in circles within the square testing cage, The occurrence o f multiple stage 5 seizures and running fits is similar to what has been described previously (Pinel & Rovner, 1978a; 1978b); however, the bouts o f running I observed were not accompanied by vocalization. Although I did not observe HD in the KHG recorded a few minutes prior to stimulation, I frequently observed postictal
6
discharges, usually accompanied by myoclonic jerks, which persisted fo r several minutes,
Kindling has been demonstrated in every species studied thus far, from amphibians to primates (Fisher, 1989), Several features o f kindling parallel human epilepsies, These include its (a) progressive nature, (b) occurrence o f spontaneously recurrent seizures, (c) permanency, (d) development o f m irror foci, and (e)
suppression by antiepileptic drugs (Sato, Racine, & McIntyre, 1990; Schmutz, 1987). On the basis o f these data, kindling is considered to be a valid representation o f human temporal lobe epilepsy,
Although the mechanisms of kindling are understood poorly, the kindling- induced neural reorganization (plasticity) is thought to involve multiple mechanisms, which include transient interactions with several neurochemicals, such as (a)
acetylcholine (ACh), (b) serotonin (SE), (c) glutamate (G LU), (d) gamma- aminobutyric acid (GABA), (e) cyclic nucleotides, and (f> N A (Fisher, 1989;
McIntyre & Racine, 1986), This dissertation concerns the modulation o f kindling by N A, and specifically the involvement o f noradrenergic a -2 receptors. I shall now describe briefly the historical events that led to the- acceptance o f N A as a putative neurotransmitter, I refer to Cooper, Bloom, and Roth (1991) extensively in the following sections describing the history and lifecycle o f NA. Therefore, unless specifically referenced, the reader should assume that Cooper, Bloom, and Roth (1991) is the source o f the information.
7
A b rie f history o f noradrenaline
The pressor effect o f suprarenal extracts was initially demonstrated by Oliver and Schafer (1895). The active compound was named epinephrine1 by Abel in 1899 and was subsequently synthesized by both Slolz and Dakin (Ilartung, 1931). Barger and Dale (1910) studied the pharmacological properties of a seres o f related synthetic amines and termed their action sympathomimetic. It was later observed that both cocaine and dennervation o f effector organs reduced the responses to ephedrine and tyraminc but enhanced the effect o f NA (Hoffman & Lefkowitz, 1990). It was thus suggested that adrenaline acted directly on the effector cell while the action o f ephedrine and tyramine were indirect, on the nerve endings. The discovery that reserpine depletes tissues o f N A by Bertler, Carlsson, and Rosengren (1956) was followed by reports that tyramine, in addition to other sympathomimetic amines, was without effect on the tissue from animals treated with reserpine, thereby providing convergent evidence that these amines act by releasing endogenous NA (Burn & Rand, 1958).
In 1946, Euler in Sweden and shortly thereafter Holtz in Germany independently identified the presence o f N A in adrenergic nerves. Although the technology did not exist to verify the claim, Euler predicted that N A was concentrated in the nerve terminal from which it was released and behaved as a neurotransmittcr,
'The terms epinephrine and norepinephrine are synonymous with the terms
adrenaline and noradrenaline, respectively (i.e., they are terms that describe the same neurochemical). For the purpose o f clarity, the terms adrenaline and noradrenaline w ill be used in the remainder o f this dissertation.
8
NA became established as the neurotransmitter o f adrenergic nerves in the periphery,, and Holtz reported that NA was a constituent o f mammalian brain. This hypothesis was met with resistance, and the accepted belief at the time was that the presence o f NA in the brain reflected the degree o f vasomotor innervation to the cerebral blood vessels. In 1954, Vogt demonstrated that NA was not uniformly distributed in the brain and that the distribution did not correlate with the density o f vascularization. Thus, the regional specificity o f NA suggested a possible role as a central
neurotransmitter.
Dahlstrom and Fuxe (1964), using the Falk-Hillarp histofiuorescence
technique, were the first to demonstrate the existence o f catecholaminergic7 neurons in the brain, although 10 years earlier Eranko, and also H illarp and Hokfelt, had reported the existence o f catecholaminergic cells in the adrenal medulla. The description o f the central catecholaminergic pathways was made possible by studies that utilised lesions o f the ascending monoaminergic pathways (Ungerstedt, 1971) to cause backup o f NA to detectable levels in noradrenergic axons, the more sensitive glyoxyiic fluorescence technique (Lindvall & Bjorklund, 1974), and the retrograde
2Catecholamine is a generic term that refers to organic compounds that contain a catechol nucleus (a benzene ring with two adjacent hydroxyl substituents) and an amine group (Cooper et al., 1991). Catecholamines generally include
dihydroxyphenylehtylamine (dopamine; D A ) ,and its metabolic products, N A and adrenaline. NA is the transmitter of most sympathetic postganglionic fibers and o f certain pathways in the brain (see below). DA is the predominant transmitter o f the mammalian extrapyramidal system and o f several mesocortical and mesolimbic neuronal pathways, Adrenaline is the major hormone o f the adrenal medulla, and exists in a species-specific proportion at sympathetic terminals, Adrenergic neurons are also present centrally in close association with noradrenergic neurons (Cooper et al., 1991).
transport o f horseradish peroxidase (Mason & Fibigcr, 1979),
The actions o f sympathomimetic amines, endogenous catecholamines and drugs that mimic their actions, can be described under six broad categories (Hoffman & Lefkowitz, 1991). These categories include a) peripheral excitation o f smooth muscle (salivary and sweat glands, vascular supply o f skin and mucous membranes), b) peripheral inhibition o f smooth muscle (wall o f the gui, bronchial tree, vascular supply o f skeletal muscles), c) cardiac excitation (increased heart rate and force o f contraction), d) metabolic actions (increased glycogenolysis in liver and muscle, liberation o f free fatty acids from adipose tissue), e) endocrine actions (modulation o f secretion o f insulin, renin, and pituitary hormones), and 0 central actions (respiratory stimulation, increased wakefulness, psychomotor activity, and reduced appetite), A hypothesis concerning the physiological function o f central noradrenergic neurons, based upon unit discharge in the noradrenergic cells o f the locus coerulcus (LG) in freely moving mammals and primates, suggests that the LC and its projections determine the brain’ s global orientation concerning both external and internal (visceral) events, analogous to a central autonomic system, The next section describes the noradrenergic innervation o f the neuroaxis.
10
Noradrenergic innervation
Noradrenergic neurons are localized in two major clusters in the CNS, a) the LG (A 6 o f Dahlstrom & Fuxe, 1964), and b) the lateral tegmentum ( A l, A2, A5, and A7 o f Dahlstrom & Fuxe, 1964; see also Loy, Koziell, Lindsey, & Moore, 1980; Mason, 1984; Storm-Mathisen & Guldberg, 1974), The LC in the rat is situated in the ventrolateral corner o f the basal central grey o f the fourth ventricle, approximately 1,2 mm posterior to lambda (Maeda, Kojima, A ia i, Fujiyama, Kimura, Kitahama, & Geffard, 1991), The LC is so named because its cell bodies contain a blue pigment that is visible in higher primates, Estimates o f the number o f noradrenergic neurons in the LC in the rat range from a few hundred (Shepherd, 1988) to approximately
1500 on either side. The existence o f a small number o f cells bodies, which project axons with a highly divergent arborization in target fields, is a distinctive
morphological feature o f aminergic neurons (Maeda et a l, 1991),
Electrophysiologically, LC neurons are characterized by tonic pacemaker activity with basal firing rates o f 0.3 to 4,0 Hz (Alreja & Aghajanian, 1991). The tonic release o f NA by LC cells is dependent upon endogenous cyclic adenosine monophosphate (cAMP) and involves the phosphorylation o f CAMP-dependent protein kinase A (Alreja & Aghajanian, 1991).
The predominant projection pattern from cells within the L C is in the rostral direction; however, a discrete mass o f cells distributes their axons from the LC in the underlying reticular formation. Fibers from the LC form five major noradrenergic
11
tracts, a) central tegmental tract (dorsal bundle), b) central grey dorsal longitudinal fasciculus, c) ventral tegmental tract (fasciculales with medial forebrain bundle), d) cerebellar cortical tract (via superior cerebellar peduncle), and e) mesencephalic/spinal cord tract,
Lateral tegmental noradrenergic cell bodies are scattered throughout the lateral tegmentum. The axons arising from these cells generally fasciculate with those from the LC. Fibers from more posterior tegmeatal levels contribute prim arily to
descending tracts whereas more anterior levels contribute to ascending tracts (forebrain, diencephalon; especially the hypothalamus).
Most o f the noradrenergic fibers in the neuroaxis arise bilaterally from the LC, mostly from the A6 cell group (Fallon & Ciofi, 1992). Approximately 15 percent o f the noradrenergic fibers arise from more caudal cell groups, A l - A5 (Fallon & C iofi,
1992). These bifurcating axons innervate the cerebellum and, via the dorsal bundle, wide areas o f the forebrain including the cortex, hippocampus, septum, and amygdala (Mason, 1984).
Every area o f the amygdala receives noradrenergic input. The densest concentration o f noradrenergic fibers terminates in the medial central nucleus and basal nuclei (Fallon & Ciofi, 1992). One interesting feature o f the noradrenergic input into the amygdala, which deviates from the pattern o f noradrenergic innervation in other forebrain areas, is the lack o f collateralization o f the fibers. This suggests that there might be local and unique noradrenergic modulation o f amygdala functions (Fallon & C io fi, 1992).
12
The amygdala also receives input From other monoaminergic neurons. The densest monoaminergic innervation appears to be the medial central nucleus, where there is a convergence o f all four monoamines (DA, N A , adrenaline, SE). Second only to the medial central, nucleus are the basal nuclei, which receive a moderately dense convergence o f D A , N A and SE innervation (Fallon & C iofi, 1992),
There is sparse and localized (medial central nucleus, medial, basal nuclei) adrenergic innervation o f the amygdala. The cells o f origin o f the adrenergic inputs are from the C l region o f the ventrolateral medulla and the C2 region o f the
dorsomediai medulla. Axons from these cells enter the amygdala via the ventral amygdalofugal bundle.
I shall now describe the lifecycle o f N A , which includes its synthesis and termination o f action.
Lifecycle o f N A
The synthesis o f NA from tyrosine was initially demonstrated in the adrenal medulla by Bfaschko in 1939 and subsequently confirmed by Nagatsu and colleagues in 1964 (Hoffman & Lefkowitz, 1990). Tyrosine is present in the blood stream, derived from dietary phenylalanine prim arily by hepatic phenylalanine hydroxylase, and accumulates in the brain and sympathetically innervated tissue via an active transport (uptake) mechanism. A t present there are no drugs that antagonize tyrosine uptake. Conversion o f L-tyrosine to 3,4 ;dihydroxyphenytalanine (DOPA) by tyrosine
hydroxylase is the rate lim iting step in the catecholamine biosynthetic pathway, Alpha-methyl p-tyrosinc (AMPT) is an effective inhibitor o f ’yrosine hydroxylase and is used clinically for inoperable NA-secreting pheochromocytomas and malignant hypertension. DCPA is converted into DA by the removal o f carboxyl groups with the enzyme DOPA-decarboxylase3, which requires pyrkioxal phosphate (vitamin 1%) as a cofactor. In noradrenergic neurons and adrenal chromaffin cells, DA is jS- hydroxylated by the enzyme dopamine-jS-hydroxylase (DBH) into N A , This enzyme utilizes molecular oxygen and requires ascorbic acid as a cofactor, DBH is not substrate specific and oxidizes almost any phenylethylamine to its corresponding phenyJethanolamine, For example, DOPA-deearboxylase can convert the drug meliiyldopa to a-methyldopamine, which is converted by DBH into the false
transmitter a-methylnoradrenaline, Also, tyramine can be hydroxylated by DBH to octopamine (the phenol analog o f N A ), which is present in only minute amounts in mammals yet may be the primary noradrenergic transmitter In some invertebrates, The activity o f DBH can be inhibited by copper chelators such as
diethyldithiocarbamate or FLA-63, which results in a reduction o f N A but an increase in D A , its metabolites, or both. In the adrenal medulla, and to a much lesser extent in the brain, N A is N-methylated by the enzyme phenylethanolamine-N-inethyl* transferase into adrenaline, N-metlrylation o f N A requires S-adenosyl methionine as the methyl donor,
3DOPA-decaHi0xylase acts on all naturally occurring aromatic L-amino acids (e.g., histidine, tyrosine, tryptophan, phenylalanine, and 5-hydroxytryptophan) and more appropriately should be referred to as L-aromalic amino acid decarboxylase,
14
Free intraneural NA participates in the control over its own synthesis via negative feedback inhibition o f tyrosine hydroxylase. Complementary to this mechanism is the observation that depletion o f NA (as a consequence o f release induced by neuronal depolarization) initiates an acceleration o f N A synthesis. This post-stimulation increase in the activity o f tyrosine hydroxylase persists beyond the duration o f the stimulation (maximally induced by 10 minutes o f continuous
stimulation). During this activation period tyrosine hydroxylase exhibits an increase in affinity for the pteridine cofactor and a decrease in affinity fo r the negative feedback properties o f D A and NA. Calcium-calmodulin-dependent phosphorylation is a candidate mechanism for this form o f short-term use-dependent plasticity o f tyrosine hydroxylase. Prolonged depolarization may amplify the action o f tyrosine hydroxylase via the formation o f new enzyme molecules, a process called transynaptic induction.
The cell bodies o f central catecholaminergic neurons contain relatively low concentrations o f transmitter, whereas the varicosities display concentrations an order o f magnitude greater. The axons are highly branched and largely unmyelinated and contain extremely low concentrations o f transmitter. Most o f the transmitter is contained within specialized subcellular organelles, referred to as electron dense core vesicles, or amine granules. The amine granules contain adenosine triphosphate (ATP) as well as DBH. The molar ratio o f ATP to transmitter (1:4) facilitates binding o f the transmitter within the vesicle: Anionic phosphate groups o f ATP form a salt link with the Cationic transmitter. There is strong evidence that the vesicles are
15
formed in the cell body a..u later transported to different locations in the neuron. Most notable among the drugs that interfere with catecholamine storage are alkaloids derived from the plant Rauwolfia serpentina, such as rescrpine and tetrabenazine. These drugs block uptake into vesicles, disrupt binding o f amines, and result in excess availability. Reserpine causes persistent damage to the vesicles, whereas the effects o f tetrabenazine are reversible.
Release o f N A from central neurons is believed to he similar to what has been observed in the periphery, where the primary mechanism is calcium-dependent
exocytosis. There is evidence that newly synthesized NA is released preferentially, suggesting that more than one transmitter pool exists, The local concentration o f catecholamine participates in the modulation o f its own release via presynaptic autoreceptors (ot-2 in the case o f N A; agonists decrease and antagonists increase release). Suggestions as to the mechanism o f autoreceptor function include inhibition o f voltage-gated calcium channels, interruption o f action potential propagation along the varicosity, opening o f potassium channels resulting in hyperpolarization, and inhibition o f adenylate cyclase (via a Gi protein) resulting in a reduction o f CAM P and calcium. In addition to autoreceptors, prostaglandins (E series), vasoactive amines, polypeptides (e.g., angiotensin II) and acetylcholine (Ach) may also participate in the regulation o f catecholamine release.
In contrast to the possibilities for the suppression o f NA release described above, stimulation o f /32 receptors may facilitate N A release. Stimulation o f presynaptic receptors could result in coupling to a Gs protein, thereby activating
16
adenylate cyclase, resulting in an increase in CAMP, calcium, and N A release. In the periphery angiotensin II has also been implicated in the /32 effects on N A release by acting as a retrograde messenger: Postsynaptic (32 Stimulation would induce the synthesis o f angiotensin II, which ^ould diffuse back across the extracellular space, bind to presynaptic angiotensin I I receptors, and stimulate N A release, The presence o f both inhibitory and facilitatory release mechanisms on the same presynaptic nerve terminal could provide sensitive and responsive control o f activity-dependent
transmitter release. Because these mechanisms are presumed to be initiated by ligand- reeeptor interactions, the relative density o f the receptors responsible for either
inhibition or facilitation o f NA release would influence the ultimate response o f a particular neuron (see Appendix C for a description o f the two-state receptor theory).
Deactivation o f released N A occurs as a function o f the operation o f a
metabolizing system, which comprises the interaction between a transporter (uptake,) and an intracellular enzyme, monoamine oxidase (Trendelenburg, 1991). The two components o f this system w ill be discussed in turn, I refer frequently to
Trendelenburg (1991) in the following discussion o f N A uptake. Therefore, unless specifically referenced, the reader should assume that Trendelenburg (1991) is the source o f the information.
Uptake, o f NA into the presynaptic terminal represents the initial and most important mechanism in terminating the action o f released N A. The carrier has binding sites for sodium, chloride, and a substrate, and is mobile when all three binding sites are free, The carrier loses m obility when it binds sodium but regains its
17
effectiveness when, in addition to sodium, the carrier also binds chloride and NA, Many more carriers are immobilized by bound sodium on the outside compared to the inside o f the axonal membrane due to the inward-directed sodium gradient. This creates an asymmetrical distribution o f carrier, with the greater concentration residing in the extracellular space (VmnxlN is greater than Vmw0UT). The affinity of N A for the carrier is increased when the carrier has bound sodium (K,MlN is less than K m0UT). The loaded carrier has a positive charge, which means that the resting membrane potential favours inward over outward transport. Therefore, the asymmetry o f carrier
distribution, its affinity for N A , and the membrane potential all contribute to a transport system that is skewed in favour o f inward versus outward transport. Uptake, is stereospecific (preference given to L-N A ), and dependent upon the extracellular concentration o f chloride and sodium ions (blocked by inhibition of sodium, potassium-activated ATPase).
The half-life for NA in the biophase (the extracellular space in proximity to receptors) is short, with a transport cycle for uptake, o f 2,5 Hz. In the periphery, uptake, varies directly with the density o f sympathetic innervation and vascularisation, The greatest uptake and binding occur in the spleen and heart.
Drugs such as cocaine, amphetamine, desipramine, amitryptyline, and other tricyclic antidepressants effectively block uptake, both peripherally and centrally (Hoffman & Lefkowitz, 1990). The net result o f greater catecholamine availability is potentiated transmission. Chronic treatment with antidepressant drugs results in reduced responsiveness o f central 0 adrenoceptors but enhanced responsiveness o f or-1
18
adrenoceptors.
As suited above, the metabolic degradation o f catecholamines in mammals occurs principally from the actions o f two enzymes, monoamine oxidase (M AO ) and catechol-O-methyltransferase (COMT). M A O converts amines into their
corresponding aldehydes, which are in turn oxidised by the enzyme aldehyde dehydrogenase into the corresponding acid. M A O is present both intracellularly, where it is prim arily associated with the outer membrane o f mitochondria but also perhaps microsomes, and extracellularly. The intracellular concentration o f M A O appears to be most important for amine metabolism; there M A O can act on free form amines in the axoplasm, amines that have been taken up by the axon and not yet bound in a vesicle, or on amines released from vesicles and not yet passed through the axonal membrane. M AO exists in two forms, designated type A and B. The distinction is made on the basis o f substrate specificity and the sensitivity to inhibition by a particular inhibitor. The A type demonstrates a specificity for N A and SE and is inhibited by clorgyline. The B type has a substrate preference for 0-phenylehtylamine and benzylamine and is inhibited by deprenyl.
COMT is a relatively non-specific enzyme, is magnesium-dependent, and catalyses the exchange o f methyl groups from S-adenosyl methionine to the m-
hydroxyl group of catecholamines. COMT is present in the cytoplasm o f most animal tissue and is abundant in the liver, kidney, brain, and sympathetically innervated organs. The termination o f the action o f catecholamines by M A O and COMT is more pronounced centrally compared to their effect in the periphery.
19
Having discussed the historical events that lead to the acceptance o f NA as a putative neurotransmitter, noradrenergic innervation in the ncuroaxis, and the lifecycle o f N A , I shall now introduce the historical and current classification scheme of
noradrenergic receptors (adrenoceptors), a topic o f particular importance to researchers concerned with the pharmacological basis o f noradrenergic drugs,
Classification o f noradrenergic receptors
20
Ahlquist (1948) differentiated noradrenergic receptors into a and 0 subtypes based on the rank order o f responses to various amines in different preparations (Maze & Tranquiili, 1991; Ruffolo, Nichols, Stadel, & Hieble, 1991). This
dichotomy was subsequently confirmed by other researchers (Hoffman & Lefkowitz, 1990). Lands, Arnold, M cA uliff, Luduena, and Brown (1967a) and Lands, Luduena, and Buzzo (1967b) refined this characterization pharmacologically and identified a-1, a-2, 0-1, and 0-2 subtypes (Maze & Tranquiili, 1991; Ruffolo et al., 1991)
Paton & Vizi (1969) identified a subtype o f adrenoceptor that regulated transmitter release, It was inferred that this receptor was located presynaptically. This led to a subdivision o f a adrenoceptors based on synaptic location, with a - l receptors considered to be postsynaptic and a-2 receptors considered to be
presynaptic. However, Wikberg (1979) reported the existence o f a-2 adrenoceptors that were postsynaptic and extrasynaptic, and that were not associated with transmitter release. Therefore, classification based upon anatomic location is untenable. Use o f more selective a adrenoceptor antagonists allowed for definitive pharmacological classification o f a-2 and a-1 adrenoceptors. A t a-1 receptors, prazosin is more potent than yohimbine, whereas at a-2 receptors yohimbine is more potent (Cheung, Barnett, &. Nahorski, 1982). More recent convergent evidence from three
independent lines o f investigation (pharmacology, mechanisms o f signal transduction, and molecular cloning o f the cDNA encoding the receptors) suggests that the
traditional classification system o f adrenoceptors needs to be reevaluated (Bylund, 1988). For example, a proposed classification o f a-2 adrenoceptors identifies up to four distinct a -2 isoreceptors, namely A, B, C, and D (Maize & Tranquiili, 1991),
For readers interested in a molecular analysis o f a -2 adrenoceptors, 1 provide a b rie f discussion o f the molecular biology o f a-2 adrenoceptors in the following section.
Molecular biology o f a -2 adrenoceptors
The transmembrane signalling o f a -2 adrenoceptors involves the coupling o f at least three separate components: the receptor protein, guanine nucleotide binding proteins (G proteins), and effector mechanisms. I shall discuss each o f these components in turn.
Molecular structure o f the a-2 receptor protein: a-2 adrenoceptors are members o f the G protein-coupled family o f membrane receptors (Gilman, 1987; R uffolo et al., 1991), and range in size from 415 to 480 amino acids in length. They possess seven hydrophobic transmembrane (TM ) segments, approximately 24 amino acid residues in length, which form alpha helices. The hydrophilic segments form loops that project either into the cell interior or exterior space. Ligand-binding sites are presumed to be associated with the T M segments, because proteolytic digestion o f the hydrophilic regions does not affect binding. TM domains fold around to make a pocket in which the ligand binds, and negatively charged aspartic acid residues within
22
the third T M segment channel bind positively charged ligands into the centre o f receptor, making T M segment 3 crucial for agonist binding (Matsui, Lefkowitz, 4 Caron, & Regan, 1989). Domains of the cytoplasmic side o f T M segments 5 to 7 exist as a contact point for a G protein, thus providing a means for signal
transduction. Modification o f T M domains, particulary 7, determines antagonist binding specificity (Kobilka, Kobilka, Daniel, Regan, & Caron, 1988),
Guanine nucleotide binding proteins: The ability o f a -2 adrenoceptors to rapidly stimulate an effector system is mediated by a fam ily o f membrane-bound guanine nucleotide binding proteins, G proteins. G proteins are heterotrimeric, with subunits designated as (in order o f decreasing mass) alpha, beta, and gamma (Gilman,
1987; Ross, 1989), Differences in the alpha subunit provide heterogeneity to the more than 10 G proteins thus far identified and serve to classify the G proteins into major classes (e.g., Gs, Gi, Go, Gk), The beta and gamma subunits are usually Closed associated and are d ifficu lt to separate. The alpha subunits have molecular weights varying between 39 and 46 kD, a single high-affinity binding site for guanine nucleotides, and intrinsic guanosine triphosphatase (GTPase) activity; they are
substrates for adenosine diphosphate-ribosylation by various toxins (e.g., pertussis, cholera, botulinum).
G Protein Cycling: The following steps describe the process o f G protein coupling after the binding o f an a-2 agonist to the adrenoceptor (Maze & Tranquiili,
1991).
receptor or the effector, and the alpha subunit o f the G protein is bound by GDP, 2.) The a-2 agonist binds to the receptor and produces a conformational change in the receptor, which makes contact with the G protein,
3.) The affinity o f GDP for the alpha subunit o f the G protein is decreased, and in the presence o f magnesium ions, GDP is replaced by GTP.
4.) The GTP-bound alpha subunit dissociates from its q dimer and couples to the effector. The affinity o f the receptor fo r the agonist decreases, and the agonist leaves the binding site.
5.) The intrinsic GTPase in the dissociated alpha subunit is now activated and hydrolyses the bound GTP into GDP, releasing inorganic phosphate.
Effector mechanisms: A t least five effector mechanisms associated with a-2 adrenoceptors have been identified, It appears likely that any one a-2 adrenoceptor can couple to more than one effector system (Maze & Tranquiili, 1991; Rtiffulo et al., 1991),
1.) Inhibition o f adenylate cyclase: Binding o f an a-2 agonist to the receptor produces a conformational change in receptor, which allows for contact with the Gi protein, and results in the inhibition o f adenylate cyclase, Reduced accumulation o f cAM P results in attenuated stimulation o f cAMP-dependent protein kinases and thus decreased phosphorylation o f target regulatory proteins. Although the inhibition o f adenylate cyclase after agonist binding is common to all a-2 adrenoceptors, a decrease in cAM P production is insufficient to mediate a-2 adrenoceptor effects. For example, artificially elevating cAM P levels in chick dorsal root ganglion cells fails to diminish
24
the ability o f NA to suppress voltage-dependent calcium conductance (Holz, Rane, & Dunlap,, 1986),
2.) Acceleration o f sodium/hydrogen ionic exchange: This antiporter effect results in the alkalinization o f the cell interior as demonstrated in NG 108-15 cultured platelet cells (Isom, Cragoe, & Lirubird, 1987). Alkanization may stimulate phospholipase A2 and initiation o f the arachidonic acid pathway (elaboration o f autacoids, e,g,, thromboxane A2),
3.) Activation o f potassium channels: The opening o f outwardly directed potassium channels hyperpolarizes membranes, which results in the suppression o f neuronal
firing (Aghajanian & VanderMaelen, 1982) and release o f N A (Schoffelmeer & Mulcier, 1984) and perhaps also co-released neurochemicals such as neuropeptide Y in medullary and pontine neurons, somatostatin and enkephalin in sympathetic neurons, and neurotensin in the LC. This effector mechanism is mediated by Gk (Codina, Yatani, Grenet, Brown, & Birnbaumer, 1987).
4.) Inhibition o f voltage-sensitive calcium channels: Suppression o f calcium channels (W illiams & North, 1985) may block the entry o f calcium into nerve terminals, thereby blocking fusion o f transmitter-containing vesicles with the synaptic membrane. This effector mechanism is mediated by a G protein, possibly Go (Heschler, Rosenthal, Trautwein, & Schultz, 1987).
5.) Phosphatidyl inositol turnover: The coupling with the membrane-bound enzyme phospholipase C via an unidentified G protein results in the hydrolysis o f phosphatidyl inositol bisphosphate into diacylglycerol and inositol triphosphate. This second
i - J
messenger system has been implicated (Lynch, Clements, Errington, & Bliss, 1988; Stezler, Feasey, Mone'a, Sincini, ten Bruggencate, & Noble, 1989) as a
possible
mechanism o f 'ong-term potentiation (LTP) and may be modulated by a-2 adrenoceptors (Maze & Tranquiili, 1991).
The amygdala
Because my research concerned kindling o f seizures by electrical stimulation o f the amygdala, a brief review o f the amygdala’ s anatomy and function is warranted, and is presented below .
The mammalian amygdala is located in the rostromedial temporal lobe, The amygdala comprises several nuclei, which can be divided into two principal groups; the superficially located corticomedial nuclear group and the deeply situated
basolateral nuclear group. The corticomedial nuclear group comprises the cortical nucleus (anterior, posterior, periamygdaloid cortex), the medial nucleus, the central nucleus, and the nucleus o f the lateral olfactory tract (Krettek & Price, 1978b). The anterior amygdaloid area and the amygdalohippocampal area are also often included in the corticomedial group. The basolateral nuclear group comprises the accessory basa* nucleus (basomedial), magnocellular (anterior), parvicellular (posterior), and the dorsolaterally situated lateral nucleus (Krettek & Price, 1978b), Recent
cytoarchitectural and histochemical studies suggest that each o f the amygdaloid nuclei may be subdivided further (McDonald, 1992),
26
Each subdivision o f the amygdala has a unique pattern o f extrinsic connections with a variety o f regions in the brain. The two major extrinsic projections from the amygdala are the ventral amygdalofugal pathway and the stria terminalis (Amaral, Price, Pitkanen, & Carmichael, 1992). The axons o f neurons from the basolateral nuclei form the more direct ventral amygdalofugal pathway and terminate in the hypothalamus, preoptic region, septum, midbrain tegmentum, and periaqueductal grey matter. Neurons in nuclei from the corticomedial group contribute axons to a looping band o f fibers, the stria terminalis, which terminates in the hypothalamus and other forebrain structures. These two tracts are both primarily subcortical. fiber systems, and do not appear to be discrete bundles: fibers from one bundle have been observed to fasciculate with the other (Amaral et al. 1992).
Amygdaloid afferents include those from the thalamus, midbrain,
.hypothalamus, raphe nuclei (SE containing cells), and the LC. The primate amygdala also receives massive input from the neocortex, including unimodal, relatively
processed sensory input from the visual (e.g., anterior portions o f area TE), auditory, and somatosensory systems. Unlike the modalities mentioned above, olfactory inputs to the amygdala possess less highly processed information (e.g., olfactory bulbs) and terminate in superficial amygdala structures (Amaral et al., 1992). The cortical connections to the amygdala are largely reciprocal, but the amygdala returns a more widespread projection. Therefore, while the amygdala receives relatively processed sensory information, it can potentially modulate sensory processing at early stages (Amaral et al., 1992). Deviating from reciprocity of cortical connections is the visual
07
A* /
input* which terminates prim arily in the lateral nucleus, with the return projection originating in the basal nuclei. This suggests intrinsic processing o f visual
information, The amygdala also receives and sends out input to polysensory regions o f the temporal, insular, cingulate, and frontal lobes.
In addition to the extrinsic connections is a robust complement o f important intrinsic connections between amygdaloid nuclei (Amaral et al. 1992), The intrinsic flow o f information through the amygdala is primarily unidirectional and proceeds from the lateral to the medial.
A ll amygdaloid nuclei possess at least two morphologically distinct cell types (McDonald, 1992), The cell type that predominates in the amygdala is the pyramidal- like projection neuron that possesses dendritic spines, The second most abundant cell type is the non-pyramida! spine-sparse neuron.
In 1939, Heinrich Kluver and Paul liucy reported dramatic behavioral changes in monkeys after bilateral removal o f the temporal lobe. Referred to as the Kluver- Bucy syndrome, it was initia lly characterized by overattentivcness, hypcrorallty, psychic blindness, sexual hyperactivity, and emotional changes (Ono & Nishijo, 1992; Rolls, 1992). Destruction o f the amygdala was implicated in the emotional changes observed in the syndrome, and the changes appeared to be species specific; monkeys became tame and placid whereas cats were rendered savage, Early studies (e.g,, MacLean & Delgado, 1953) that reported the effects o f electrical stimulation o f nuclei w ithin the amygdala described a variety o f behaviors that were produced including facial automatisms (licking, biting, blinking), both clonus and tonus o f limbs, and
28
sympathetic responses (pupil dilation, increased respiration). These behaviors were dependent upon the area stimulated, MacLean and Delgado (1953) may have used stimulation parameters sufficient to kindle seizures, because their descriptions o f the behavioral effects o f amygdaloid stimulation are similar to what is currently used to characterize partial seizures.
The amygdala is a component o f the neural circuit that is critical for processing information concerning certain forms o f learning (Chapman, Kairiss, Keenan, Sc Brown 1990; Kentridge, Shaw, Sc Aggleton, 1991; Peinado-Manzano,
1990), including classical conditioning o f the nictitating membrane response (Weisz, Harden, Sc Xiang, 1992), passive avoidance (McGaugh, 1990), and both cued and contextual fear conditioned associations (Hatfield, Graham, &. Gallagher, 1992; Phillips Sc LeDoux, 1992), Lesioning the amygdala interferes with these forms o f learning,
There is an emerging consensus that the amygdaloid complex is critical fo r an organism to appreciate the significance o f environmental stimuli (Amaral et al.,
1992), Consistent with this role, the amygdala also appears to be a prominent structure in establishing and maintaining current emotional states (Rolls, 1992).
These features o f amygdala function make it particularly relevant in the expression of social behavior as well as the modulation o f memory storage and retrieval. The mnemonic contribution o f the amygdala lies in its participation in the cognitive evaluation o f events and assignment o f the emotional valence o f information to be remembered. (Rolls, 1992). This is an attractive hypothesis i f one considers the
29
important role o f the amygdala in emotion as well as its projection, via the entorhinal cortex, to the hippocampus, a temporal lobe structure that is critical for certain forms o f memory.
Rationale
To this point I have reviewed epilepsy, kindling, NA, and the amygdala. These topics are all interrelated in my dissertation research. The purpose o f
introducing these topics was to provide the reader with a basic understanding o f the essential elements o f my research. I w ill now present the rationale for conducting my dissertation research.
One o f the most consistent findings in the kindling literature is that treatments that mimic or modify noradrenergic transmission can influence kindling profoundly (Corcoran Sc, Weiss, 1990). Depletion o f N A after central infusions o f the neurotoxin 6-hydroxydopamine (6-OHDA) results in a significant acceleration o f the acquisition o f generalized seizures induced by electrical stimulation o f limbic structures (Arnold, Racine, & Wise, 1973; Corcoran, Fibiger, McCaughran, Sc Wada, 1974; Corcoran Sc Mason, 1980; McIntyre Sc Edson, 1982; McIntyre, Saari, & Pappas, 1979) or
neocortex (Altman Sc Corcoran, 1983). However, depletion o f NA after kindling does not affect established generalized seizures (Westerberg, Lewis,, Sc Corcoran, 1984). These data indicate that the modulatory role o f NA is restricted to the early stages o f the kindling process: NA acts specifically to delay the emergence o f
30
generalized seizures from partial seizure (Corcoran, 1988).
Convergent evidence that further documents noradrenergic modulation o f kindling has come from recent studies4 demonstrating that the density o f a-2 receptors fo r NA is increased during and after kindling (Chen, V ig il, Savage, & Weiss, 1990; Chen, Weingardt, & McNamara, 1990; Jimenez-Rivera e ta l., 1989), that single neurons in the LC are activated by lim bic AD (Jimenez-Rivera & Weiss, 1989), that kindling can be delayed by electrical stimulation o f the LC (Weiss, Lewis, Jimenez-Rivera, V ig il, & Corcoran, 1990), that lesions o f the LC facilitate kindling (N ’ Gouemo et al., 1990), and that 6-OHDA-induced facilitation o f kindling can be offset by transplantation o f fetal LC neurons into the amygdala/pyriform cortex (Barry et al., 1989) and the hippocampus (Bengzon, Kokaia, Brundin, & Lindvall, 1990).
Gellman, Rallianos, and McNamara (1987) reported that systemic
administration o f the a-2 noradrenergic agonist clonidine (0.01 - 0.2 mg/kg; i.p .) produced a dose-dependent retardation o f kindling with electrical stimulation o f the amygdala. They also reported that kindling was dose-dependently facilitated by systemic administration o f the a-2 noradrenergic antagonists idazoxan, yohimbine, and rauwolscine (0.1 - 10.0 mg/kg; i.p ,), but that kindling was unaffected by
^Depletion o f N A by 6-OHDA has also been reported to potentiate other forms o f seizures including audiogenic seizures (Bourne, Chin, & Picchioni, 1977), seizures evoked by alcohol withdrawal (Chu, 1978), and seizures produced by
electroconvulsive shock (Browning & Maynert, 1978; London & Buterbaugh, 1978). Pharmacological and electrophysiological studies using mutant mice with epileptic phenotypes (e.g., ddY. qk, tg) have also contributed to the current understanding o f the role o f the noradrenergic system in the modulation o f seizure susceptibility (Helekar St Noebels, 1991; Tsuda et al., 1990),
administration o f selective antagonists o f a-1 receptors or nonselectivc antagonists o f /3 receptors. I also have observed that systemic administration o f idazoxan (1.0 mg/kg; i.p,) significantly facilitates kindling produced by electrical stimulation o f the amygdala or the perforant path (unpublished observations). The a-2 adrenoceptors
mediating noradrenergic modulation o f seizure susceptibility arc presumably located postsynaptically, because agonists continue to exert antiepileptogcnic effects even when presynaptic noradrenergic terminals have been eliminated by pretreatment with N-(2~chloroethyl)-N-ethyl-2-brOmcbenzylamine (DSP4) or 6-OHDA (Bcngzon et al.,
1990; Gellman, et al,, 1987; McIntyre & tjiu g n o , 1988).
One vital question concerning the Influence o f NA during kindling has not been addressed: Where in the nervous system does N A act at a-2 receptors to antagonize kindling? Because infusions o f 6-O HDA into the amygdala itself can facilitate amygdaloid kindling (M cIntyre, 1980), and because o f the high density o f a- 2 receptors in the amygdala/pyriform region (Unnerslall, Kopajtic, & Kuhar, 1984), I hypothesized that a critical population o f a-2 adrenoceptors that influence kindling resides in the amygdala/pyriform region. I therefore examined the effects o f
intraamygdaloid infusions o f the a-2 adrenoceptor agonist clonidine on the kindling o f seizures with electrical stimulation o f the amygdala. Repeated central administration o f clonidine was required; therefore, to minimize potential neurotoxic effects I initially utilized low-frequency stimulation, which kindles generalized seizures quite rapidly (Corcoran & Cain, 1980). I report here that clonidine administered
32
infusions o f clonidine significantly retard the kindling o f seizures with both low- frequency and conventional high-frequency electrical stimulation, and that this effect is localised to a population o f a-2 adrenoceptors in the amygdala/pyriform lobe. I provide additional evidence in support o f the idea that stimulation o f a-2
adrenoceptors mediates the observed retardation by demonstrating that kindling is retarded by intraamygdaloid infusions o f xylazine, another a-2 adrenoceptor agonist, and that the retardation produced by clonidine is abolished when clonidine and
idazoxan are infused simultaneously. In contrast to my hypothesis, however, I found unexpectedly that intraamygdaloid infusions o f a-2 antagonists failed to accelerate kindling,
GENERAL M ETHOD
Procedures common to all experiments are described in this section and deviation from these procedures is described in the appropriate section. I used male Long-Evans hooded rats (Charles River, St, Constance, Quebec), weighing 270 - 450 g at time o f surgery.
Surgery: Rats were anesthetized with sodium pentobarbital (65 mg/ml) and received bilateral implantation o f either bipolar electrodes, bipolar chemitrodes, or tripolar chemitrodes into the basolateral amygdala (B LA ) using conventional
stereotaxic techniques, Relative to bregma, the stereotaxic coordinates were: AP - 2,8, M L + 5 .2 and D V -8.6 mm (Paxinos & Watson, 1986), with the inci*. j r bar at - 3.9 mm. The electrodes consisted o f twisted nichrome wire, 127 ^m dia. The chemitrodes consisted o f twisted nichrome wire electrodes epoxyed to 23 gauge
stainless steel guide cannulas. The tip o f the electrode* and the injection cannula, extended 0.5 - 1.0 mm beyond the guide cannula. Obterators were kept in the guide cannula between infusions. The electrode wires were attached to female Amphenol pins and inserted into a plastic headplug. Dental acrylic secured the headplug
assembly to four jew eller’ s screws anchored in the skull; one o f the screws served as ground.
Afterdischarge threshold: After a 10 day postoperative period the afterdischarge threshold (A D T) was determined for each rat. Stimulation was delivered using a Grass S88 stimulator and consisted o f constant-current biphasic square wave with a pulse width o f 1.0 ms, frequency o f 60 pulse pairs per sec (pps), and train duration o f 1.0 sec. Stimulation was initially administered at 20 fiA. (base to peak) and was increased in 10-juA steps once per min until AD was evoked (Pinel, Skelton, & Mucha 1976). Rats were randomly assigned to treatment groups. Group size was in part dependent upon attrition o f subjects due to uncontrollable events including sickness, electrical failure, and incorrect electrode placement,
Low-frequency kindling: Rats in the low-frequency groups received kindling stimulation at a frequency o f 3 pps and an intensity o f 1200 nA, an intensity selected on the basis o f p ilo t work demonstrating that it reliably evoked AD during low- frequency stimulation o f the amygdala. The maximum duration o f the stimulation train was 60 sec; stimulation was terminated i f a stage 4 seizure (Racine, 1972b) occurred. Stimulation was delivered to all rats once every 48 hrs,
