Title: Neuroanatomical and pharmacological correlates of the behavioral manifestations of intraventricular administration of kainic acid in the rat /
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098847/00001
 Material Information
Title: Neuroanatomical and pharmacological correlates of the behavioral manifestations of intraventricular administration of kainic acid in the rat /
Physical Description: vi, 90 leaves : ; 28 cm.
Language: English
Creator: Lanthorn, Thomas Herbert, 1950-
Publication Date: 1978
Copyright Date: 1978
Subject: Kainic acid -- Physiological effect   ( lcsh )
Neuroanatomy   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 84-89.
Statement of Responsibility: by Thomas Herbert Lanthorn.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098847
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000095745
oclc - 06349345
notis - AAL1176


This item has the following downloads:

PDF ( 3 MBs ) ( PDF )

Full Text







Words cannot express the heart-felt thanks I give to all those

who have made this life possible. Nevertheless, thank you.

There are a few people who must be singled out from the many

because of their direct contribution to insuring that this dissertation

was a successful event.

First, SueAnne, from whom all life draws its meaning; my friend

and my wife.

Dr. Robert L. Isaacson; who always seemed to understand my needs

and to give direction when I didn't see any. I could not do better

than to have half his breadth of foresight.

Dr. Carol Van Hartesveldt; a true critical scientist worthy of

being imitated, to whom this final written work owes much.

The rest of my committee; Drs. Adrian Dunn, Charles Vierck, L.

James Willmore, and Marc Branch, who received my work with an interest

I hope to deserve.

Mrs. Virginia Walker; who, in the final analysis, made this

dissertation a reality--a right hand and a friend.

And last, SueAnne, who put up with more than could be expected of

anyone, who brought me joy and beauty, and to whom the rest of my life

is dedicated.

ACKNOWLEDGEMENTS ... . . . . . . . . . ii

ABSTRACT . . . . . . . . . . . . v

INTRODUCTION . . .......... . . . . . . 1

Glutamate . . . . . . . . .. . . .. 1

Kainate . . . . . . . .... . . . 7


Introduction . . . . . . . . . . . . 18

Materials and Methods .... . . . . ..... 18

Results. . . . . . . . . ... ... ... . 21

Discussion ....... . . . .. ...... . 24


Introduction .. . . . . ..... . . . . 26

Methods and Materials. . . . . . . ... ...... 26

Results. . . . . . . . ... ......... 27

Discussion . . . . . . . . . . . . 28



Introduction . . . ... . . . . . . . 33

Methods and Materials . .. . . . . . . ... 33

Results . . . . . ..... . . 33

DISCUSSION ... . . . . . . . . ... 34

RECEPTOR BLOCKAGE. . . . . . ... .. . . . . 36


Introduction . . . . . . . . .. .. 36

Methods and Materials . . . . . . . . ... 37

Results. . . . . . . . . . ....... .37

Discussion ..... . . . . . ..... 37


Introduction . . . . . . . . .... 39

Methods and Materials . . . . . . . .... .40

Results . . ... . . . . . . . 41

Discussion . . . . . . . . ... .. . . 44


Intraventricular versus Intracisternal Injection . . .. 46

Kainate-Induced Lesion versus Kainate WDS . . . ... 51

Selective Lesions of Hippocampal Subfields . . .... . 53


Introduction . . . . . . . . .. .. .. . 63

Opiate Withdrawal . . . . . . . . . . . 64

Serotonin . . ... . ............ . 68

Methionine-Enkephalin . . . . . . . ... .. 71

Ketocyclazocine. ....... . . . . .... 73

Sodium Valproate .................. . . 75

Ice Water .... . . . . . . . .. .... 77

GENERAL DISCUSSION .. ....... . . . . . 79

Summary .... . . . . . ...... . 79

Implications . . . . . . . . . . . .... 80

REFERENCES. .. . . . . . . . . . .... 84

BIOGRAPHICAL SKETCH ..... . . . . . .. . 90

Abstract of Dissertation Presented to the Graduate Council of the
University of Florida in Partial Fulfillment of the Require-
ments for the Degree of Doctor of Philosophy



Thomas Herbert Lanthorn

December, 1978

Chairman: Robert L. Isaacson
Major Department: Psychology

Kainic acid (kainate) is believed to be an agonist of the putative

neurotransmitter, glutamate. Intraventricular injection of kainate

was found to induce dose-related changes in the behavior of rats and

to result in selective lesions in the brain. One constellation of

behaviors was induced by doses that did not appear to be neurotoxic.

This set of behaviors included 'wet-dog' shakes (WDS), diarrhea and

excessive salivation. It appeared to be the same constellation of

behaviors that occurs during morphine withdrawal in the rat. This set

of behaviors was chosen for further study because they were considered

to be the most likely to have resulted from activation of normal

mechanisms of neurotransmission.

A series of studies investigated the neuroanatomical and pharma-

cological correlates of this behavioral manifestation of kainate

administration. The evidence from studies which examined the effects

of restricted ventricular injection of kainate and those which investi-

gated the effects of selective lesions of the brain on kainate-induced

behavior suggest that kainate induces this constellation of behaviors

by actions on the CA3 and CA4 pyramidal neurons of the hippocampus.

A lesion, induced by kainate itself, was found to prevent the

occurrence of this constellation of behavior by subsequent injections

of kainate. The effect of this lesion on the occurrence of these be-

haviors induced by agents other than kainate was examined. The data

suggests the existence of two distinct mechanisms for the induction

of these behaviors. One mechanism reacts the same as morphine with-

drawal to pharmacological manipulations, and the other, sensitive to

kainate, reacts in an opposite manner to the same manipulations.

Evidence is presented to suggest that the kainate-sensitive mechanism

is mediated by a second opiate receptor, the kappa (K)-receptor.



Evidence for a neurotransmitter role

A strong case can be made that glutamate is a neurotransmitter in

the central nervous system. lontophoretically applied glutamate can

depolarize neurons (1-4). On vertebrate neurons a hyperpolarizing

effect has been reported only once--in cerebellar slices (5). Near-

ly every neuron tested has been found to be sensitive to glutamate.

This led some investigators to propose that glutamate may be a non-

specific membrane depolarizing agent (1,2). Several experimental

results are not consistent with this hypothesis. It has been reported

that iontophoretically applied glutamate depolarizes neuronal soma or

dendrites, but not axons (3). The depolarizing effect on soma or

dendrites is restricted to extracellular application; intracellular

ejection has little effect (6). Furthermore, glutamate does not de-

polarize most olfactory bulb cell bodies (7), indicating that it is not


In addition, some neurons are more sensitive to glutamate than

others. McLennan et al. (8) probed the thalamus with micropipettes and

found regional differences in the amount of response to glutamate. These

differences appeared to be coexistant with known nuclei, that is, neurons

of the ventrolateral nucleus were more responsive to glutamate than the

neurons of adjacent nuclei.

Neurons have been shown to possess uptake mechanisms for glutamate

(9-12). In fact, two mechanisms appear to exist. One is a glutamate

specific, Na+-dependent, mechanism present in synaptosomes. The other

is a nonspecific, non-Na+-dependent mechanism. The glutamate-specific

sodium-dependent mechanism has a much higher affinity than the non-

specific binding site. The sodium-dependent mechanism corresponds to

a transmitter reuptake mechanism (a specific, fast-transport mechanism),

while the nonspecific binding site is probably just the general uptake

mechanism present in all cells.

A third binding site for glutamate also exists. It has an even

higher affinity for glutamate than the specific reuptake mechanism, but

is not sodium dependent. It is very specific for glutamate (versus GABA

or glycine, for example) and most of its specific activity is in synapto-

somes. This probably represents a postsynaptic glutamate receptor (12-15).

Looking at just one area, the mossy fibre bundle, for one example, a

high concentration of glutamate and glutamate dehydrogenase, its synthetic

enzyme, is found (16). The specific, Na+-dependent transport mechanism

is also found in this region of the hippocampus (17). Stimulation of the

mossy fibre bundle excites the pyramidal neurons postsynaptic to it (8),

and also results in a large increase in extracellular (presumably, re-

leased) glutamate (16). The excitation of the post-synaptic neurons is

rapid in onset and very powerful; effects which are mimicked by ionto-

phoretic application (19). Thus, many of the usual criteria for defining

a substance as a neurotransmitter have been satisfied at the mossy fibre

bundle terminal field. The major piece of data missing is to show that

the effects of tract stimulation can be blocked by glutamate receptor

antagonists. This criteria has been met in the corticostriatal tract (20)

and the perforant pathway (21), but other criteria have not been satisfied

in these terminal fields. Therefore, by all the usual criteria, though

not all have been shown at one synapse, glutamate is a neurotransmitter.

It seems fairly reasonable then to carry out research as if it were a

transmitter, always cognizant that it may fail at some point to fit the


Use of glutamate as an experimental tool

Although glutamate is the endogenous compound of interest, several

factors make it undesirable for experimental use. First of all, the

detection and measurement of glutamate directly involved in neurotrans-

mission is severely hampered because glutamate is a natural amino acid

used by all cells of the body, both as a constituent of protein and as

part of the Krebs cycle, the energy mechanism of all neurons and glia

(22). Thus the brain contains a large amount of glutamate which is not

directly related to neurotransmission.

The brain goes to extraordinary lengths to keep the extracellular

concentration of glutamate low. The only areas which allow passage of

measurable amounts of glutamate from the blood into the nervous system

are the circumventricular organs which do not possess the primary blood-

brain barrier--capillaries whose lining cells have tight junctions be-

tween them (23,24).

Within the brain itself, extracellular concentrations of glutamate

are kept very low by active uptake processes. The intracellular concen-

tration is quite high and this imbalance contributes to the polarization

of neuronal membranes. In an attempt to study the distribution of

glutamate synapses, McLennan (25) injected tritiated glutamate into the

brains of rats. The autoradiographs showed that essentially all the

radioactivity (glutamate) ended up in glial cells. In a study of the

metabolism of glutamate by the brain, Mao, Guidotta, and Costa (26)

injected glutamate into the lateral ventricles, killed the animals as

soon as five minutes after the injection, and assayed the brains for

increases (relative to controls) in glutamate and various metabolites.

Even though the amount of glutamate injected was equal to about 30% of

the total normal glutamate content of the whole rat brain, they could

find no increase in glutamate content after only five minutes. Metab-

olites, like glutamine, were greatly increased. These two studies under-

score the fact that the brain itself has very powerful mechanisms for

sequestering and catabolizing glutamate. Furthermore, such mechanisms

also occur in the glial cells so that neuronal mechanisms need not be

affected. Thus, the direct injection of glutamate is not likely to have

much direct effect upon neurons unless very high local concentrations

can be achieved, as they are in iontophoretic application.

In an effort to overcome this handicap, large amounts of glutamate

have been administered by various routes. The results from the injection

of glutamate, by any route, seem to fall into two classes. The first

category consists of little or no effect. The second category consists

of convulsions, toxic behaviors and neural degeneration (26,31). For

example, the intraventricular injection of 1 mole of glutamate had little

effect on the general behavior or operant responding of rats. It general-

ly lowered the rate of responding but did not alter the acquisition,

pattern, or extinction of the operant response. Ten moles of glutamate

resulted in tonic seizures (32). Thus, exogenous glutamate is not a

very useful neuromodulator. The toxic, convulsive dose is very high

relative to behaviorally active doses of other transmitters, suggesting

that convulsions are not the normal function of glutamate.

The neural degeneration induced by systemically injected glutamate

is notable since it selectively destroys the inner layers of the retina

and those parts of the brain immediately adjacent to circumventricular

organs (CVO), such as the arcuate nucleus of the hypothalamus and the

area postrema (24). Thus peripheral injections of glutamate may be

useful as a tool for inducing certain lesions. The pattern of degenera-

tion after systemic or direct, intracranial injection is also interest-

ing. The lesion is restricted to those neurons whose soma or dendrites

are within the injected area. Axons passing through and afferent termi-

nals are unaffected (23,24,33,34). Thus,glutamate may be useful as a

tool for inducing lesions in areas containing major fiber bundles, such

as the striatum and the lateral hypothalamus. While it may become use-

ful for producing lesions, glutamate is of little use in understanding

the normal function of glutamate as a neurotransmitter.

Even if glutamate did induce interesting behavioral changes there

would still be a problem in interpretation. There appear to be two re-

ceptors with which glutamate can interact. One is also well-suited for

aspartate, another excitatory aminoacid, which is slightly shorter than

glutamate. The glutamate molecule, itself, can fold to some extent and

fit this same receptor. However, there also appears to be another recep-

tor which is best suited for a fully extended molecule of glutamate and

for which aspartate is too short. The first receptor may be termed the

general excitatory amino acid receptor (GEAAR) and the second, the gluta-

mate-preferring receptor (35-37). (See Figurel). The first receptor

would have specificity bestowed on it in the brain because either a

glutamergic or aspartergic terminal would synapse with it. However, in-

terpretation of the results of exogenously applied glutamate would be

limited to actions possibly mediated endogenously by glutamate.









/ I I
\/ 11





Figure 1

Structures of Aspartate, Glutamate and Kainate


Kainate as a glutamate agonist

Recently, a number of chemicals have been recognized to be analogues

of glutamate and tests have shown some to be agonistic and a few others

antagonistic by iontophoretic application. The agonists have gained

particular attention because some of them are as much as one thousand

times more potent than glutamate itself. In particular, domoate,

quisqualate, and kainate, are very potent glutamate agonists on mammali-

an neurones (35,38,39).

Kainate is of particular interest. By far the most attractive

attribute of kainate was that it appeared to be a very poor substrate

for the glutamate reuptake mechanism (12,40). One report showed that

an excess of kainate strongly displaced glutamate from the high affinity,

Na+-independent binding site in the synaptosome fraction of locust neuro-

muscular junction, but had almost no effect on the binding of glutamate

to the lower affinity, Na+-dependent binding site (12). The two sites

probably correspond to the postsynaptic receptor and the neuronal reup-

take mechanism, respectively. This noninteraction with the fast trans-

port mechanism would result in kainate being active for a prolonged

period of time. Iontophoretic studies of kainate have shown that it is

one to two orders of magnitude more powerful than glutamate (35,38,39).

This could be accounted for by the paucity of kainate uptake though no

direct evidence is available.

Recently this simple agonistic mode of action of kainate has been

challenged. Singh, McGeer, and McGeer (41) reported that after kainate

injection, the Na+-dependent reuptake of glutamate was increased, an

effect which seemed to be dissociable, temporally, from postsynaptic

receptor degeneration. This dissociation suggested a direct effect of

kainate upon the presynaptic elements. In a following study, McGeer,

McGeer, and Singh (42) destroyed the corticostriatal tract, presumably

the glutamergic input to the striatum, prior to striatal injection of

kainate. This loss of presynaptic input eliminated the neurotoxic

effect of kainate in the striatum. Further in vitro data showed that

kainate, at fairly high concentrations (mM), strongly inhibits the Na+-

dependent glutamate reuptake mechanism (42,43). This suggests that

kainate's effects at these doses may be due to presynaptic actions, such

as blockade of reuptake or increased release of glutamate.

Like glutamate, the effect of kainate on mammalian neurons is de-

polarization. Hyperpolarization has never been reported as a result of

kainate application. The depolarization occurring after application of

kainate is long-lasting. Excessive or prolonged depolarization is neuro-

toxic and kainate is indeed a very potent neurotoxin (44).

Systemic (oral, sc, or ip) administration of kainate has three

notable effects--convulsions, neural degeneration, and death. All three

effects occur at the same doses so that these effects cannot be sepa-

rated. A dose of 0.15 mmol/kg (30 mg/kg) of kainate induces convulsions

and neural degeneration in mice. This makes kainate two hundred times

more potent than glutamate in inducing neurotoxicity (24). The pattern

of neuronal destruction succeeding kainate is identical to that induced

by systemic glutamate. The kainate lesion, not unexpectedly, extends

slightly further out from the CVOs than glutamate-induced degeneration.

Since death often results after a neurotoxic dose of kainate, systemic

administration of it is not a very useful experimental tool. One study,

however, is interesting. Polc and Haefely reported that 0.3 mg/kg

kainate, iv, increased monosynaptic, though not polysynaptic, reflexes

[45), that is, increased the amount of muscle contraction induced by

submaximal stimuli. This may be consistent with the hypothesis that

glutamate is a transmitter of primary sensory afferents to the spinal


The binding of 3[H]-kainate to the brain has been studied by Simon,

Contrera, and Kuhar (46). Their studies show that kainate binds to

grey matter, but not white matter, and most strongly to the crude synap-

tosomal fraction. In terms of structures, the striatum had the most

dense binding. Cerebral cortex, hippocampus and cerebellum had similar

densities of binding which were about half that of the striatum. The

midbrain, thalamus and pons-medulla all were similar and were one-fifth

to one-fourth of the density shown by the striatum.

The binding of kainate was most easily displaced by kainate and

quisqualate, suggesting a similarity of these two neuroexcitatory agents.

Of all the other agents examined glutamate was the most effective agent

in displacing kainate. Glutamic acid dimethyl ester (GDME) and glutamic

acid diethyl ester (GDEE) were also reasonably effective displacing agents

However, aspartate was a rather poor displacing agent. Other putative

neurotransmitters, specifically glycine, GABA, serotonin, norepinephrine,

dopamine, and acetylcholine, were completely ineffective even when the

concentration of these chemicals was a couple of thousand times that of

kainate. These results indicate that kainate and glutamate bind at the

same site, a site which is quite specific for their chemical structure.

Kainate is even more interesting as a glutamate agonist because of

its chemical structure. As shown in Figure 1 it is locked into its con-

figuration by the presence of a ring structure and some bulky side-chain

groups. Thus, it is an analogue of the extended conformation of gluta-

mate (35). The functional realization of this specificity has been shown


in the differential sensitivity of Renshaw and non-Renshaw spinal inter-

neurones to aspartate and glutamate in cats. Various studies have shown

that glutamate is more concentrated in the dorsal horn, while aspartate

is concentrated in the more ventral regions. This has suggested to some

authors that primary afferent terminals release glutamate while spinal

interneurones use aspartate (47). In the cat, at least, Renshaw cells

do not appear to receive primary afferent terminals, while most other

spinal interneurones do (48,49). The Renshaw cells appear to be more

sensitive to iontophoretically applied aspartate than glutamate, while

the converse is true of the other spinal interneurones (50). Kainate

applied to these same neuron populations is more effective on spinal

interneurones other than Renshaw cells. The difference is greater with

kainate than with glutamate suggesting that the difference is in fact due

to the presence of two different receptors, the GEAAR and the glutamate-

preferring receptors. In the rat, Renshaw cells appear to receive pri-

mary afferent terminals and the differential sensitivity to glutamate and

aspartate is not seen (51).

Aspartate was a poor antagonist of kainate binding (46), a finding

which is consistent with the hypothesis that kainate specifically inter-

acts with the glutamate-preferring receptor. Furthermore, two agents

known to antagonize responses to aspartate, diaminopimelic acid (DAP) and

diaminoadipic acid (DAA), were completely ineffective in displacing

kainate from its binding site (46).

Use of kainate as an experimental tool

A relative mountain of studies of the effects of intracranial in-

jection of kainate have appeared in the last two years. Three research

groups are responsible for the vast majority of these studies: E.G.

McGeer, P. L. McGeer, T. Hattori et al.; J. T. Coyle, R. Schwarcz et al.;

and J. W. Olney et al.

The primary locus of these studies has been kai.nate injection into

the striatum of rats. With concentrations of kainate ranging from 2.5 -

15 mM (0.5 1 pl) the following have been consistently reported. Within

threeto five days after unilateral injection, neurons of the injected

striatum degenerate (52,56). When 10 mM (1 pl) of kainate is used, over

90% of the neurons totally degenerate. The degeneration can be seen

within a couple of hours of the injection. Histological examination has

revealed several important facts. The first elements affected are

neuronal soma and dendrites. The axons of the degenerating neurons begin

to degenerate in due course (56). However, up to three weeks after in-

jection, the longest period studied, axons of passage and afferent termi-

nals are unaffected (57). This is particularly interesting in the caudate

because the internal capsule fibers course through it. These would be

destroyed by most lesion techniques. Hattori, McGeer, and McGeer (58)

labeled the axon terminals of the corticostriatal tract with 3[H]-proline

and then injected kainate into the striatum. They found that the

labeled terminals were unchanged by the injection while the relevant post-

synaptic structures degenerated. This result is doubly interesting be-

cause of the evidence that glutamate is the transmitter released by

these terminals.

Biochemical markers of synaptic integrity have also been analyzed.

The results are consistent with the histological data. Neurons of the

striatum are believed to use two transmitters, ACh and GABA. Intrinsic

neurons use ACh and GABA, while the principal efferent neurons use GABA.

As expected, the kainate-induced lesion greatly reduces the content of

these transmitters, their synthetic enzymes, and their specific reuptake

mechanisms. GABA content is also reduced in the substantial nigra which

is consistent with the striatonigral tract being GABAergic. ACh content

of the substantial nigra is unaffected (53,55).

Several other transmitters are afferent to the striatum. Dopamine

input to the striatum is believed to arise from neurons in the pars

compact of the substantial nigra. After injection of kainate into the

stratum, neither the content nor the specific reuptake mechanism for

dopamine is changed. The synthesis of dopamine is, by distinction, great-

ly increased. This appears to be due to an increased number of tyrosine

hydroxylase molecules (53). These results are consistent with unaffected

presynaptic terminals and a loss of postsynaptic receptors.

Serotonin input to the striatum comes from neurons of the raphe'

nuclei. After a kainate-induced lesion, the activity of the specific

reuptake mechanism for serotonin is unchanged, while receptor binding is

decreased (59). These results are similar to those seen with dopamine.

The status of glutamate in the striatum is less clear. One possible

glutamergic input is the corticostriatal tract. After lesion in the

striatum with kainate, glutamate receptor binding is decreased. However,

the glutamate content is unchanged and the reuptake mechanism is slightly

increased (63). The study by Hattori, McGeer,and McGeer (58), mentioned

earlier, showed that the corticostriatal terminals stayed intact and had

the same concentration of vesicles after kainate as in controls. This

evidence strongly suggests that glutamate in the striatum arises from

extrinsic neurons.

Lesion in this tract selectively reduces the glutamate content of the
striatum (relative to other amino acids) and the terminals of this tract
possess a glutamate reuptake mechanism. Stimulation of the corticostri-
atal tract excites striatal neurons and this stimulation-induced excita-
tion can be blocked by glutamic acid diethyl ester (GDEE), a glutamate
receptor blocker (20,60-62).


These biochemical and histological changes are very similar to

those found in human patients who have died of Huntington's chorea. The

kainate-induced lesion may provide an animal model of Huntington's chorea

and research is moving in that direction.

Finally, lesion in the basal ganglia by kainate (5 mM, 1 pl) lowers

the met-enkephalin content by 50% (64). This is half the dose of kainate

that destroyed over 90% of the striatal neurons. This suggests that the

enkephalinergic input to the striatum arises from basal ganglia neurons.

It has been suggested that these are located in the globus pallidus.

In the doses used to produce most of these biochemical changes (10 mM,

1 pl) destroy over 90% of the striatal neurons. However, in most cases

synaptic integrity, as measured by biochemical methods, is reduced by

60-60%. This difference is disconcerting, but may be partly explained:

by a recent report by Olney and de Gubareff (57). They found, by EM

examination, that although neurons had totally degenerated, the postsynap-

tic densities were still attached to many of the intact presynaptic termi-

nals. These densities, which probably contain the receptors, appeared to

be intact up to three weeks after kainate injection. Longer postinjection

periods were not examined. If the receptors are intact, then binding

could still occur and explain the presence of binding in the absence of

postsynaptic neurons.

In their first report on biochemical changes McGeer and McGeer (55)

reported that 2.5 nmol ( > 0.5 pg) induced significant changes, but that

1 nmol (0.21 pg) did not. Coyle and Schwarcz (52) reported that they

tried 0.1 10 pg of kainate. The lowest dose for which data was reported

was 0.5 pg and it induced minimal changes. These studies suggest the possi-

bility that some concentrations of kainate may not be neurotoxic.

Unilateral injection of kainate, in neurotoxic concentrations, into

the striatum results in changes in behavior. In rats, grooming stops and

diarrhea occurs nearly immediately (55). The animals turn contralaterally

and go into tonic-clonic convulsions. With more than 2 jg (10 mM, 1 pl),

death occurs frequently (53,55). The contralateral turning occurs for

several hours and is replaced by ipsilateral turning. The timing of this

change correlates well with neuronal degeneration and may represent

excited and then inhibited (absolutely) glutamoceptive neurons in the

caudate (53). Only one study has examined the effects of bilateral in-

jection. Neurotoxic doses (6 mM, 1pl) were injected into each caudate.

The rats were tested later for changes in general activity, and acquisi-

tion and memory of a step-down passive avoidance task. No changes were

seen in spontaneous locomotor activity. However, kainate-injected animals

were significantly impaired in both acquisition and one-day retention of

the passive avoidance task (65).

Specificity of kainate to glutamate receptors has been tested by

injecting various agents prior to or along with kainate. Neither agonists

nor antagonists of dopamine, acetylcholine, GABA, or serotonin, nor chron-

ic morphine treatment protected against kainate toxicity. Some analogs of

kainate and glutamate did, e.g., diethyl kainate, OC-methylglutamate,

diaminopropionic acid, and diaminobutyric acid (66).

Several studies have examined the effects of injections into the

substantial nigra (53,67). Histological examination reveals that the pars

reticulata is the primary site of degeneration. Biochemical markers of

presynaptic activity indicate that GABA is unaltered. Thus, GABA is

probably an extrinsic transmitter; from the striatum. Acetylcholine is

also unchanged. Dopamine content is reduced, though only about 40-50%

at a high concentration of kainate (12 mM, 0.3pl). This suggests that

nondopaminergic, glutamoceptive neurons exist in the substantial nigra,

primarily localized in the pars reticulata.

Injection of kainate into one substantial nigra results in ipsilateral

turning by the animal for several hours. This is succeeded by contra-

lateral turning in concert, at least temporally, with neuronal degenera-

tion. These results are consistent with the effects of unilateral elec-

trolytic lesion and opposite to the results with 6-OHDA-induced lesion.

The kainate effects are not affected by 6-OHDA lesions. This strongly

suggests that a nondopaminergic, glutamoceptive neuron population also

controls turning in rats, independently of dopaminergic neurons, and

opposite in behavioral results (67).

Only one study has examined the effects of thalamic injection. The

injection of 5 mM kainate (1 1p) was directed into the ventrolateral

nucleus. Behaviorally, these unilateral injections result in the

cessation of grooming, the appearance of circling, and various convul-

sive behaviors. Biochemical changes were relatively minor, consisting

of a 28% decrease in glutamic acid decarboxylase (GAD) activity in the

thalamus and no change in choline acetyltransferase (CAT) or tyrosine

hydroxylase (TH) activity. No change in the activity of these enzymes

was seen in either the striatum or substantial nigra subsequent to thalam-

ic injection (55).

Herndon and Coyle (68) injected kainate (10 mM, 1 pl) into the

cerebellum. They found, histologically, that all the neurons of the

cerebellar cortex degenerate except the granule cells. This result is

crucial to the claim that kainate acts via glutamoceptive neurons. The

granule cells give rise to the parallel fibers which are the primary

excitatory input to the cerebellar neurons. The boutons of the parallel

fibers synapse on the Purkinje, Golgi, stellate and basket neurons, that

is, all the other neurons of the cerebellum. However, no granule-granule

synapse has been reported. Several lines of evidence on various mutants

and viral infections have correlated loss of granule cells with a specif-

ic reduction in glutamate content. Thus it appears that the granule

cells are glutamergic, but not glutamoceptive. Since kainate selectively

spares the granule cells, the claim that it acts via glutamate receptors

is strengthened.

Schwarcz and Coyle (54) reported that when they injected the striatum

with kainate (10 mM, 1 pl) they also found that the CA3/4 pyramidal neurons

of the hippocampus underwent degeneration. This selective degeneration

was studied directly by Nadler, Perry and Cotman (69). They injected

0.1 3.0 g (0.5 15 nmol) of kainate into or near the hippocampus.

As little as 1.5 nmol (0.3 yg) caused many neurons of CA3/4 to degenerate,

but CA1 and dentate granule cells were spared. CA3/4 neurons differ from

the others in one respect by virtue of their innervation by the mossy

fibre bundle. Earlier it was suggested that the evidence is sufficient

to claim that mossy fibers are glutamergic, as is the perforant pathway.

At higher doses, and correlated with total loss of CA3/4 neurons, CA1

neurons degenerate. The dentate granules degenerate if the injection is

made directly into the hippocampal formation, but not by ventricular in-

jection. The results indicate that it is possible to selectively destroy

CA3/4 of the hippocampus at concentrations which are not known to be

toxic to any other neurons. The results also show that very low concen-

trations of kainate are selectively toxic to certain CA3/4 neurons. In-

jection of 0.1 pg (0.5 nmol) induced degeneration of CA3a pyramidal neurons

only (69).


DeGubareff and Olney (70) injected 20 nmol of kainate into the gen-

eral ventricular circulation. The volume was unspecified. They reported

that the subjects (mice) died within one hour. Degeneration was reported

in primary sensory-receiving neurons of the spinal cord, some cerebellar,

and some hippocampal neurons. The fact that kainate spread so far so

quickly suggests that a large volume was injected. This report indicates

the sensitivity of the hippocampus and cerebellum which contain major

fiber tracts suspected to be glutamergic.



The behaviors reported in these studies of kainate that were just

reviewed result from administration of neurotoxic doses of kainate. It

is unclear whether they tell us much about behaviors in which glutamate

normally participates. However, injection of agonists of other suspected

transmitters have resulted in enhancement or reduction of particular

behaviors. One example are the turning behaviors induced by unilateral

striatal injection of various agents. Such agonist-induced behaviors

can be useful for examining neural circuits and may also indicate neural

mechanisms underlying behavioral problems. Thus, it seemed useful to

determine if kainate would induce a particular constellation of behaviors,

especially at nontoxic doses.

Materials and Methods

The subjects for this study were 56 male, Long-Evans hooded rats

and six male Sprague-Dawley albino rats. All weighed from 200-300 grams.

Surgical procedures were performed while the subjects were anesthe-

tized by sodium pentobarbital (55 mg/kg), administered intraperitoneally.

The head was shaved and the animals secured in a stereotaxic device. The

head was tilted such that the nose pointed upward five degrees. This

allowed the use of the atlas of the rat brain by Pellegrino and Cushman

(71). The skin was split with a scalpel and the fascia reflected from

the skull with a blunt scalpel handle. The skin and fascia were maintained

in a retracted position by means of S-shaped pins attached by rubber

bands to the stereotaxic machine.

The stereotaxic was used to locate the part of the skull above the

left lateral ventricle. The coordinates, relative to bregma, used were

2.4 mm posterior and 4.7 nn lateral. A hand-held drill with a one milli-

meter diameter bit was used to drill through the skull. The bit extended

one millimeter out of the drill chuck which allowed it to drill through

the skull without slipping into cortex and possibly damaging cortex.

Three holes were drilled, one at the coordinates given above and two more,

one on each side of the skull to hold supporting screws. After the holes

were drilled, bone chips were removed with the aid of microdissecting

forceps and a dissecting microscope until the dura matter was unobstruct-

ed and flat across the bottom of the hole. The dura and pia matters, in

the hole placed above the lateral ventricle, were slit using the beveled

edge of a 27 gauge needle.

Two small screws were attached to the skull and then a polyethylene

cannula guide, riding on the electrode holder of the stereotaxic instru-

ment, was lowered through the hole placed above the left lateral ventricle

to a depth of five millimeters from the dorsal surface of the skull. This

placement was calculated to put the tip of the guide into the lateral

ventricle just antero-lateral to the posterior portion of the hippocam-

pus. The guide was constructed from polyethylene tubing that had an

outer diameter of about 0.7 millimeter. A metal wire which fit snugly

into this size tubing was threaded through a length of the tubing. This

combination was placed over a low flame until the plastic softened. This

softened plastic was then compressed so that it bulged out and formed a

ridge around the tubing. This process was repeated every few centimeters

along the tubing. The tubing was then cut into pieces such that five

millimeters of tubing was left below the ridge and the total length of

the piece was 11 millimeters long. The ridge served as a guard against

the guide slipping too far into the brain.

Once the guide was in place in the brain, it and the two screws

were covered by cranial cement to form a smooth, solid cap. The loose

skin was pulled together around the cap and held together by wound clips.

A stylet made from a 27 gauge needle was placed in the guide to keep it

from becoming clogged prior to use. Bicillin (60,000 U) was administered

intramuscularly to reduce the possibility of infection. The subject was

then removed from the stereotaxic and allowed to recover in its home cage

for seven to ten days before testing.

Just prior to testing, the appropriate solution of kainate was

prepared by dissolving it in saline. The solution was brought up to a

pH of 7.4 with sodium hydroxide. The subjects were transported to the

observation area in their home cages. Injections were performed by gently

hand-restraining the subjects and, after removing the stylet, inserting

a 27 gauge needle into the cannula guide.

Eleven and one-half millimeters of this needle were allowed to pass

into the guide so that a small portion extended past the guide itself.

The 27 gauge needle was attached to one end of a length of polyethylene

tubing which was filled with kainate solution. The other end of this

tubing was attached to a microliter syringe.

Two and one-half microliters of kainate solution, of different con-

centrations, was injected over a 30-second period. Kainate solutions

used had concentrations of 0.006, 0.03, 0.1, 0.2, 0.8, 2, 20 and 200

millimolar. After the injection was complete, the needle was removed

and the subject was immediately placed in an individual observation box

(30 x 30 x 20 centimeters). Each. box was illuminated by a 7.5 watt white

light bulb. The rats were observed from an adjacent room through a one-

way mirror. White noise (65 decibels) was introduced into the experiment-

al room to mask extraneous noise. The subject's behaviors were observed

for one hour and any unexpected behaviors recorded and, where possible,


At the end of behavioral testing, the subjects were given an over-

dose of sodium pentobarbital and two and one-half microliters of India

ink was injected in the same manner as kainate had been. The animals

were then intracardially perfused with 0.9% saline, followed by 10%

formalin. The brain was removed, sectioned and the presence of ink in

the various ventricles was determined.


Rats receiving large concentrations of kainate (20 or 200 milli-

molar) underwent fits of leaping and running, convulsions and death.

Injection of lower concentrations (0.8 and 2 millimolar) of kainate was

followed by convulsions, barrel-rolls and contralateral turning. Con-

centrations from 0.006 0.2 millimolar induced 'wet-dog' shaking (WDS)

and diarrhea. At the 0.2 millimolar concentration of kainate, wet-dog

shaking was overshadowed by myoclonic jerks, especially prominent in the

front legs. The magnitude of these convulsive bodily reactions made it

difficult to identify 'wet-dog'shaking apart from the other convulsive

movements and therefore WDS was not quantified at this dose. Following

injections of the 0.1, 0.03, and 0.006 millimolar kainate solutions, WDS

were not accompanied by clonic jerks. Kainate at a concentration of

0.006 millimolar induced a mean of 5.5 'wet-dog' shakes during the one-

hour observation period. Rats injected with 0.03 millimolar kainate pro-

duced a mean of 40 WDS in the hour, while 0.1 millimolar kainate

resulted in a mean of 75 WDS in one hour. Diarrhea became very pro-

nounced at this last dose level. Excessive salivation, chewing and

teeth chattering were also seen at these three concentrations, but their

occurrence was not as consistent as WDS and diarrhea. Also, especially

with 0.1 millimolar kainate, some rats seemed to expel a large amount of

moisture which manifested itself in the fogging up of the observation

boxes. Injection of pH-corrected saline (the vehicle solution) into the

lateral ventricle produced a consistent pattern of effects: short bouts

of exploration and grooming for 20-30 minutes, followed by sleep. Very

few WDS were observed, mean of 0.9 in the hour period, in vehicle-

injected animals. The three groups receiving the lowest concentrations

of kainate differed significantly from each other and from vehicle-inject-

ed rats in terms of the number of WDS produced (p's <.05 or less; Mann-

Whitney U-test). These data are summarized in Table 1.

Six Sprague-Dawley albino rats were injected with 0.25 nmol of

kainate. Their behavior was quantitatively and qualitatively similar

to the Long-Evans hooded rats at the same dose.

Study of the distribution pattern of India ink showed that 2.5 1l

of ink injected intraventricularly generally filled most of the left

lateral ventricle, but only rarely entered the body of the right ventricle.

It was present in the intraventricular foramen of Munro, the third ven-

tricle, cerebral aqueduct, and fourth ventricle. In some animals it was

present in the cisterna magna, in the area above the colliculi, and

around the lateral edge of the thalamus.







* C
2- 0





0 D


O >i



0' f
+^ S


















C -

a J


O *
*- CL

c" C

O s-

0 J


U -a
0 0 5

0=E o
'-0 Q(


n1 L
C S-
0 S-


U 0

0- u



:3 r
C '-

4* 0

4- 0.

I nC

(0 -

30 Li 4-

S 4


i- r-!

oc Ln
Sc o


n O-

u 2

0 -0C
0U -0
-a s


00 0

0 C
t 0

0a >

3 0





-a C





o 0.e



The results indicate that kainate is a very potent behaviorally

active agent. As little as 1.3 ng/pl of injection solution induced a

significant modification of behavior. The doses of kainate used here

induced three concentration-dependent sets of behavior. At the high

concentrations examined, kainate administration resulted in fits of

leaping and running, convulsions, and death. The middle range of con-

centrations induced contralateral turning, barrel-rolls and convulsions.

Administration of the lower doses was associated most consistently with

wet-dog shaking and diarrhea.

Leaping and running fits and death, as seen after large doses of

kainate, are behaviors which are associated with general neurotoxins such

as quabain (72), copper (72) and high doses of glutamate (28). Since

these concentrations of kainate are much higher than those necessary to

destroy neurons in any area tested, it is suggested that these doses,

associated with toxic behaviors and death, are acting as general neuro-


Contralateral turning and convulsions have been reported by Schwarcz

and Coyle (53) and McGeer and McGeer (55) following intrastriatal injection

of kainate. The effective concentrations are in the same range as those

which induce the same behaviors following intraventricular injection. The

distribution of India ink suggests that kainate solution reached the ven-

tricular surface of the striatum. Thus, the occurrence of these behaviors

after intraventricular injection of kainate may be a consequence of direct

striatal stimulation.

Low concentrations of kainate induce a set of behaviors which appears

to be very similar to morphine withdrawal (the morphine abstinence syn-

drome)(73). Jumping is a behavior commonly observed during withdrawal in

rats dependent on large amounts of morphine (73). It was not observed

in rats administered kainate and observed in small, closed-top boxes.

However, jumping was seen when a couple of kainate-injected subjects were

transferred to a large, open-top field. Thus, many of the symptoms of

morphine withdrawal can be elicited by acute administration of kainate

in opiate-naive rats.

The capability of kainate to induce WDS places it among a rapidly

growing list of substances which can induce WDS in opiate-naive animals.

Natural situations such as skin contact with cold water or xylene are

followed by wet-dog shaking. Chemicals which can bring on this kind of

shaking include TRH thyrotropicc hormone) (given intraperitoneally, ip,

or intracerebroventricularly,icv) (74,75), 5-HTP (given ip)(76), o<,B and

y-endorphin (given icv)(77), met- and leu-enkephalin (given icv)(78),

high doses of morphine (given icv)(79), AG-3-5 (l-[-hydroxyphenyl]-4-[3-

nitrophenyl]-1,2,3,6 tetrahydropyrimidine-2-one) (given ip)(80), sodium

valproate (given ip) (81), Sgd 8473 (o<-[(4-chlorobenzylideneamino)-oxy]-

iso-butyric acid) (given ip)(82), somatostatin (given icy) (83), RX 336-M


ip) (84) and theophylline (given ip) (85).



The aim of the dose-response studies was to discover behavioral

manifestations of nonneurotoxic doses of kainate. Therefore it is

necessary to examine the brain itself to determine what consequences

various doses of kainate have. Concentrations of kainate of two nano-

moles/microliter, or more, result in potently convulsive behaviors and

often in death. These toxic behaviors represent abnormal neuronal func-

tioning and so histological examination was restricted to lower concen-


Methods and Materials

Histological alterations produced by unilateral injection of kainate

were studied in 24 male, Long-Evans hooded rats. Cannulas were implanted

in the same manner as previously described and the animals were allowed

to recover seven to ten days before receiving kainate. Kainate solutions

were prepared and injected in the manner already described.

Seven rats were injected with a 1 nmol/pl solution into the left

lateral ventricle as described above. Two of the seven received 1 pi,

two received 2 pl, and two were injected with 5 l1 of the kainate solution.

The other member of each group was sacrificed 24 hrs after injection. The

seventh rat was administered 0.5 pl and it was sacrificed 24 hrs after in-

jection. Seven other rats received 2 pl containing a total of 0.2 nmol

of kainate. Two of these were sacrificed 24 hrs afterwards, while the

other five were sacrificed after one week. Ten rats were injected with

2 pl of 0.05 nmol/pl kainate solution and sacrificed one week later.

All subjects were sacrificed by an overdose of sodium pentobarbital

(Nembutal). This was followed by intracardial perfusion with 0.9% saline

and then a 10% formalin solution. The brains were removed and placed in

10% formalin. The brains were subsequently embedded in celloidin and cut

into 30 pm coronal sections. Every fifth section was saved, mounted on

slides and stained with thionin for microscopic examination.


Examination of the sections revealed that a 1 nmol/pl solution is

consistently neurotoxic to some areas of the brain within reach of the

injection. Most prominent of the affected areas is the CA3 and CA4 pyrami-

dal neurons of the hippocampus. Just three hours after injection these

neurons on the injected side appear deflated and twisted. By 24 hrs

degeneration is quite evident, involving all the neurons in these two

fields of Ammon's horn on the side of the injection.

The CA3 and CA4 fields of the contralateral hippocampus appear un-

affected. CA1 neurons on the injected side appear pale compared to the

contralateral CA1 neurons, but structurally well-formed. Neurons of the

dentate gyrus seem unaltered. A small cluster of large pyramidal neurons

in the dorso-lateral extreme of Ammon's horn also appear to be insensitive

to this concentration of kainate. These neurons may represent the CA2

field of the hippocampus.

At the larger injection volumes, neuronal degeneration also appeared

in the cortico-medial division of the ipsilateral amygdala. There is also

loss of deep pyramidal neurons of the neocortex near the injection site,

but whether this is due to the kainate or the cannula itself was not


A lower concentration of kainate (0.2 nmol in 2pl) produced neuronal

degeneration in three out of seven rats. The loss of neurons was


restricted to area CA3a in dorsal hippocampus. This area is immediately

adjacent to the tip of the implanted cannula. In most cases the degenera-

tion extended only a short distance from the level of the cannula in the

anterior-posterior plane. The appearance of degeneration in these animals

did not correlate with the number of wet-dog shakes observed, nor with the

appearance of convulsive behavior.

The lowest dose of kainate used (0.1 nmol in 2 pl) produced no ob-

servable neuronal alterations in the hippocampus or any other structure

in the brain.

As reported in many studies (53,55), kainate is a neurotoxic agent.

The neurotoxicity appears to be both selective and dose-dependent. The

CA3-4 region of the hippocampus appears to be extremely sensitive to

kainate, even at a concentration of 0.1 nmol/pl. The extreme sensitivity

of the CA3-4 region was first noted by Coyle and Schwarcz (52) who re-

ported destruction of the neurons in this hippocampal area following

intrastriatal injection of kainate. The CA3 and CA4 fields of the

hippocampus were the only extrastriatal damage observed by these investi-

gators. This selective neurotoxicity was also investigated by Nadler,

Perry, and Cotman (69). They reported that intraventricular injection

of 0.5 nmol/pl 2 nmol/pl kainate resulted in the selective destruction

of the CA3-4 region of the brain. No disturbances were seen in any other

parts of the brain when examined in the presence of a Nissl body stain or

silver stain for degenerating fibers, by these investigators.

The present study indicates that this part of the hippocampus is

even more sensitive to the neurotoxic effects of kainate than previously

reported (69). A 0.1 nmol/pl solution was found to destroy some neurons

in this region in about half the animals examined, suggesting that by this

method of administration, 0.1 nmol/pl represents an LD50 (lethal dose,

half the time) for these neurons. Injection of kainate in a lower con-

centration seemed to be without neurotoxic effect, as examined by a Niss]

body stain, even. though this dose was effective in inducing WDS and other

behaviors associated with the injection of kainate.

The number of WDS which occurred following the injection of a 0.1

nmol/pl solution did not appear to be correlated to the presence or

absence of degenerating neurons. This may further indicate that WDS and

neurotoxicity are not directly correlated phenomena.

The primary hippocampal neurons destroyed by these neurotoxic doses

of kainate are large pyramidal neurons. However, a small cluster of

large pyramidal neurons in the dorsolateral extreme of Ammon's horn

appear to be insensitive to these doses. Since the CA1 field of the

hippocampus consists primarily of medium-size pyramidal neurons, this

kainate-insensitive cluster of large pyramidal neurons may represent the

CA2 field of the hippocampus.


Administration of 2.5 pl of kainate solution into the lateral

ventricle of rats results in changes in their ongoing behaviors. There

seem to be three sets of dose-dependent kainate-sensitive behaviors.

Concentrations of 0.006 0.1 nmol/pl induce a constellation including

wet-dog shakes, diarrhea, etc. that is quite similar to morphine with-

drawal (73) or endorphin- or enkephalin-induced behavior (77 ). Higher

concentrations, up to about 5 nmol/Pl, induce myoclonic jerks and con-

tralateral turning, while even higher concentrations manifest them-

selves as running fits and death.

Simply because of the problem of keeping subjects alive, further

investigation of the highest concentrations of kainate does not seem

advantageous. Furthermore, leaping and running fits are a highly abnormal

behavior which makes them, at least to this investigator, intrinsically

less interesting. Since lower doses are neurotoxic, it is likely that

these doses are at least as toxic, adding to the feeling that this is a

very abnormal situation and less directly related to the mundane activi-

ties of the animal.

The histological evidence indicates that the intermediate doses,

associated with contralateral turning and convulsions, are neurotoxic. A

1 nmol/pl solution consistently obliterates the CA3-4 field of the.

hippocampus, while even 0.1 nmol/pl resulted in more restricted destruc-

tion in about half the animals examined. This belies that the effects

of these doses are also correlated with highly unusual neuronal function-

; 30

It is interesting that similar behaviors can be obtained from direct

intrastriatal injection of similar concentrations, but the fact that

other groups are working on this type of injection suggests that work

on some other behaviors would be useful and possibly most helpful.

Kainate can induce wet-dog shakes, etc. at doses which do not appear

to be neurotoxic by light microscopic examination of brains stained with

Nissl-substance stains (to allow examination of cell bodies) by either

this investigator or Nadler, Perry and Cotman (69) or in material stained

with silver to identify degenerating axons (Nadler, Perry and Cotman).

Even though some structural or functional alterations may have occurred

that were not identified by these techniques, the behaviors induced by

these doses stand as the most likely possibility for a behavior normally

mediated by glutamate.

The constellation of behaviors is interesting in its own right

because it appears to be very similar to the constellation of behaviors

observed in opiate withdrawal in rats. Any chance to shed some light on

the mechanisms of opiate dependence and withdrawal is worth following up.

Therefore, the behaviors resulting from administration of kainate

in concentrations of 0.1 nmol/pl or less will be intensely investigated.

Several questions will be emphasized.

1) Is kainate-induced behavior glutamate-mediated behavior?

2) Where does kainate act to induce wet-dog shakes, etc.?

3) How does kainate-induced behavior relate to similar behaviors

induced by other agents and situations, especially morphine


4) Does this behavioral constellation represent a response to

normally encountered environmental situations?


Of the behaviors in this constellation, wet-dog shakes appear to be

the most consistent response. It is also the most quantal response and

the frequency of the occurrence of wet-dog shakes seems to be directly

related to the concentration of kainate (within a range of 0.006 0.1

nmol/pl). Therefore, wet-dog shakes will most often be the behavior

used to test the efficacy of experimental manipulations, though the

presence of the other associated behaviors will be recorded when possible.



In the dose-response studies, behavior was observed for one hour.

This time period was selected simply because many of the obvious altera-

tions in the behavior of the subjects had disappeared by that time and

the subjects had begun to sleep. However, in order to manipulate a

specific behavior, in particular kainate-induced WDS, it is useful to

know the exact time course of the behavior of interest. This is important

in order to devise the most effective test period.

Methods and Materials

The subjects for this study were eight male, Long-Evans hooded

rats weighing 200-250 grams. Surgical and injection procedures were the

same as already described. A dose of 0.2 nmol of kainate (in 2 pl), the

most effective concentration in inducing WDS, was administered to six

rats. Two rats served as controls and were injected with 2Pl of the

vehicle solution (pH-corrected saline). The observation period was ex-

tended to 3.5 hrs.


The majority (79%) of kainate-induced WDS occurred during the first

hour following its administration. Another 18% occurred during the

second hour and by 2.5 hrs, all WDS had occurred. At this time, all

subjects were asleep and except for brief periods of wakefulness, re-

mained asleep for the remainder of the observation period. One saline-


injected rat emitted one shake during the second half-hour. The other

control subject showed no WDS. The data are shown graphically in Figure 2.


Most wet-dog shakes induced by kainate occur in the first hour after

injection. This suggests that a one-hour observation period is probably

sufficient for testing. This result also suggests that the effects of

this concentration of kainate do not persist or progress into toxic

behaviors as might be expected if permanent changes in the brain were




0- kainate injected
40\ vehicle injected



1 2 3 4 5 6 7

1/2 hr. periods

Figure 2

Time Course of Kainate-Induced Wet-Dog Shakes



It appears from the data examined above that kainate can result

in alterations in behavior at doses which are not neurotoxic. Tenta-

tively, then, we may hypothesize that this change in behavior is due

to neuronal stimulation and furthermore to glutamate receptor stimu-

lation. This hypothesis may be tested by studying the effect of a

glutamate receptor antagonist on kainate-induced behavior.

Glutamic acid diethyl ester (GDEE) has been reported to antagonize

glutamate-induced neuronal excitation (86,89 ). Furthelrore, it has

been shown, by iontophoretic and systemic administration, to reduce the

effects of stimulation of fiber tracts thought to release glutamate

(87, 89 ). However, there have been reports that GDEE is not particu-

larly effective or specific (90 ). In general though, the most recent

reports (e.g., 88 ) show GDEE to be a very effective antagonist, which

is selective to aspartate and glutamate, but which shows little ability

to separate these two amino acid putative neurotransmitters. A large

part of the earlier controversy probably stems from the speed with which

GDEE undergoes hydrolysis to form glutamate itself (87 ). Therefore,

if GDEE is properly stored (dry and cold) and used soon after being

placed in aqueous solution, it appears to be a very effective antagonist

of glutamate and aspartate.

Segal (89) reported that systemic administration of 200 mg/kg

(1 mmol/kg) moderately reduced the field potential associated with

perforant pathway stimulation--thought to be mediated by glutamate.

Stone (87 ) reported that 30Q and 60Q mg/kg reduced the effects of

corticostriatal tract stimulation. This tract also contains fibers

believed to release glutamate.

Methods and Materials

Eighteen male, Long-Evans hooded rats were used as subjects for

this study. Surgical and i ntraventricular injection procedures were

the same as in preceding studies. GDEE was removed from a freezer and

dissolved in saline just prior to each use. Ten rats were injected with

2 mmol/kg GDEE (2 ml/kg), intraperitoneally, 5 min prior to intraventric-

ular injection of kainate (0.25 nmol in 2.5 pl). Eight animals received

saline ( 2ml/kg) instead of GUEE. All sugjects were observed for 1 hour,

but for six animals in each group separate determinations of the WDS

in the first and second 30-min periods were made.


Rats pretreated with 2 mmoles/kg of GDEE had an overall reduction

of 56% of WDS relative to their controls (p <.02). As evaluated by

those subjects where separate evaluations of the first and second 30

min periods were made, this effect occurred because of a reduction of

WDS in the first 30-min period. For these GDEE-treated animals the

median WDS for the first 30-min period was 21, while their controls

exhibited 42 (p <.005). During the second 30-min period the median

WDS of the GDEE-treated group was 27 compared to the median of the con-

trol animals of 36. This difference was not significant.

Stone (87 ) and Segal C89 ) reported that GDEE was effective when

administered intraperitoneally. This route was chosen for this study


because the time to effect had been reported and there was concern about

injecting two yoluynes of liquid into the ventricles. On the other hand,

this meant allowing GDEE to spread throughout the central nervous system.

The result from this experiment indicates that blockade of glutamate

and aspartate receptors; reduces the effect kainate has on overt behavior.

This is consistent with the expectation that kainate affects behavior by

stimulating glutamate receptors. However, it is also possible that

GDEE blocks excitatory amino acid receptors at some point on the output

side of the site of kainate's action. Thus, this result can only be

taken as suggestive, though it does indicate that a hypothesis in which

kainate does not act via glutamate receptors will not gain easy support

from the data.



There exist doses of kainate which induce WDS, but do not produce

noticeable neural degneration. Degeneration is seen in some animals

injected with a 0.1 nmol/pl solution, but the occurrence of degenera-

tion does not appear to be related to the number of wet-dog shakes

recorded. These findings suggest that WDS induced by kainate are not

a result of denerative changes in neurons.

The actions of kainate in inducing WDS appear, at the light micro-

scopic level, to be dissociable from kainate's neurotoxic effects.

However, it is possible that permanent subcellular or functional al-

terations are produced by the WDS-inducing actions of kainate. This

possibility may be examined by giving kainate on multiple occasions and

examining if consistent alterations in the response to kainate admini-

stration are found as a consequence of prior administrations. The dose

of kainate employed should be in the nonneurotoxic range.

The elicitation of a unique and consistent set of behaviors by low

doses of kainate suggests that kainate may be stimulating a particular

part of the nervous system at these doses. Slightly higher concentra-

tions of kainate have been found to induce neural degeneration in a

specific field of neurons. It is possible that these two events are


It has been suggested that prolonged, excessive depolarization is

the means by which glutamate and kainate induce neuronal death--excito-

toxic effects (24 ). Depolarization mediated by stimulation of gluta-

mate receptors would also be expected to be the mechanisms by which

kainate induces behavior. It is possible that the concentrations of

kainate which induce wet-dog shakes do so by depolarizing a particular

group of sensitive neurons and that raising the concentration of kainate

increases and/or prolongs that depolarization to a toxic level. If this

were the case, then the specific neurotoxicity induced by minimally

toxic doses of kainate may be on the same neural field by which kainate

induces WDS. This possibility can be tested by first inducing a lesion

with kainate itself and subsequently examining the ability of kainate to

produce WDS.

Methods and Materials
The subjects for this experiment were 23 male, Long-Evans hooded

rats. They all weighed 200-250 grams at the beginning of the experiment.

Surgical and injection procedures were again the same as described earlier.

Five rats made up Group 1 and were not implanted. They were used to

measure the behavior of intact animals in the experimental situation.

Five other rats (Group 2) were implanted with a cannula in the left ven-

tricle and were injected only with the vehicle solution on each test day.

The test days for all groups were four days apart. Group 3 consisted of

five rats implanted with a cannula. This group of rats was treated

according to the following schedule:

Test Day 1 1.5 nmol kainate (in 1.5 1p solution)
Test Days 2-4 0.25 nmol kainate (in 1.5 pl solution)
Test Day 5 vehicle (1.5 pl)

Eight rats made up Group 4. They were cannulated in the same manner

as groups 2 and 3 and were tested according to the following schedule:

Test Days 1-6 0.25 nmol kainate (in 1.5 pl solution)
Test Day 7 vehicle (1.5 pl)
Test Days 8,9 0.25 nmol kainate (in 1.5 pl solution)

Six weeks separated Test Days 7.and 8, but all other Test Days

were separated by four days.

In addition to wet-dog shakes (WDS), the occurrence of stretching

and yawning was also recorded for all the groups during the one-hour

observation periods. Observation began immediately after the injections

were completed.


Very few WDS were seen at any time in either intact or saline-

injected rats. The average median score for the intact animals was 1.0,

while it was 1.17 WDS in an hour for the saline-injected rats. The

median number of WDS did not change significantly across days for either

group, nor were there any significant differences between the two con--

trol groups. The mean and median number of wet-dog shakes for each

group is shown in Table 2.

The number of wet-dog shakes was not recorded on the first test

day for those animals administered 1.5 nmol of kainate (Group 3). All

rats given 1.5 nmol manifested numerous myoclonic jerks and some tonic-

clonic convulsions. On Test Days 2-4, when 0.25 nmol of kainate was in-

jected, the number of wet-dog shakes stayed at a relatively consistent,

low level. The mean median number of WDS for all four groups is given

in Table 2. The number of wet-dog shakes evoked on Days 2-4 were not

significantly different from either control group on the corresponding

test day. Furthermore, the number of wet-dog shakes produced on Days

2-4 were not different from that produced by vehicle injection on Test

Day 5.

The median number of wet-dog shakes produced by injection of 0.25

nmol kainate on Day 1 (Group 4) was 74.5. This was significantly more

than produced by intact, uninjected animals or by vehicle injection

(ps < .01). On the second test day the median number of WDS was 89.

The Group 4 scores on Day 2 were not significantly different from their

Day 1 scores. They were significantly different from the control groups

(ps-< .01). The number of wet-dog shakes produced on Day 2 in Group 4

was also significantly higher than that induced by similar injection in

Group 3 (p <.01). On Test Days 3-6 the number of WDS evoked steadily

decreased, though it always remained significantly higher than the con-

trol groups (ps < .05). In order to examine whether the decrease was

due to tolerance or habituation, or due to neuronal destruction, six

weeks were allowed to pass without the subjects being exposed to injec-

tion or the observation apparatus. Two more test sessions, four days

apart, were then carried out. The median WDS produced on these days was

2 and 3 respectively, which was not significantly different from the num-

ber produced by the saline injection on Test Day 7.

The number of stretches and yawns evoked by either control group was

never significantly different from the other on any day. The number of

stretches and yawns recorded for all four groups is compiled in Table 3.

As can be seen, the number of stretches and yawns rises quickly and

reaches a plateau of about 8 by the third test day.

Stretches and yawns in Group 3 rats were not counted on Test Day 1.

On Test Days 2-4, when they received 0.25 nmol of kainate, they did not

differ from either control group, and also reached a plateau of about 8.

Injection of vehicle solution on Test Day 5 did not induce more or less

SY than the kainate injection of Test Day 4.

On all test days when they received kainate injections,except Day 6,

Group 4 animals evoked significantly less SY than either control group

(ps < .05, or less). On Test Day 7 when they were injected with the






















Cl- C\ a- ^

0 0 0 0
(3 (D C3
o a 0 0
C- C- C- C-
cc (c c 0














I i


C0 00


r- C

CL ". mL C.

S- S- C- s-
Lc C3 cc C

C --

0- 0

r- 0




vehicle they produced a median of 9 SY during the observation period,

well within the plateau established by the other groups.


Doses of kainate which have been shown to be neurotoxic to the

CA3-4 field of the hippocampus almost totally prevent subsequent doses

of kainate from inducing WDS. Thus, the effects of kainate in inducing

this behavior and producing neuronal degeneration are linked. This data

indicates that these two effects are mediated by the same neural circuit

and possibly even the same neurons, though this cannot be determined from

this data.

The ability of the kainate-induced lesion to prevent kainate-induced

behavior is further confirmed by the occurrence of a normal amount and

pattern,across days, of stretching and yawning following kainate adminis-

tration in lesioned subjects. In animals given only the lower dose of

kainate, the normal appearance of stretching and yawning is suppressed

until Test Day 6 when the number of WDS induced by kainate is much re-

duced relative to earlier administrations. To summarize, the ability of

kainate to suppress stretching and yawning is lost when the animal is

pretreated with a neurotoxic dose of kainate.

The number of WDS induced by injection of kainate after one previous

injection of kainate, also at the lower dose (Group 3), was not signifi-

cantly different from the number produced by the first injection. This

could suggest that no significant alterations in function of the nervous

system had resulted from the first application of kainate. Furthermore,

stretching and yawning were effectively suppressed by the second applica-

tion indicating that no loss of potency had occurred in regard to this

behavior as a consequence of the previous dose.

However, subsequent injections of kainate resulted in reductions in

the amount of WDS produced and a reduced ability to suppress stretching

and yawning. The reduction could be the result of many factors such as

tolerance or habituation. It could also be the result of destruction.

In order to separate these possibilities, the subjects in Group 4 were

not exposed to the experimental situation, the injection procedure, or

kainate for six weeks. Retesting indicated that the ability of kainate

to induce WDS was still reduced. This suggests that a permanent change

in the nervous system had occurred.

To try to increase the chances of selectively affecting the hippocam-

pus the volume of injection solution had been reduced to 1.5 1p. How-

ever, the amount of kainate used to induce WDS, i.c., 0.25 nmol, was kept

the same. This raised the concentration as compared to previous studies,

from 0.1 to 0.167 nmol/pl. Furthermore, at the time this study was begun,

the ability of a 0.1 nmol/Pl solution to lesion some neurons had not been

recognized. These facts suggest that the low dose used here was probably

capable of inducing partial destruction of CA3 and CA4 each time it was



Intraventricular versus Intracisternal Injection


Since intraventricular administration of low doses of kainate

induce specific constellations of behaviors, it is probably reasonable

to suspect that it is acting on specific neural circuits. Thus, it may

be possible to define specific anatomical loci of kainate's action

which result in behavioral change. In particular, the structure by

which kainate induces WDS should be identifiable.

Bloom, Segal, Ling, and Guillemin (77 ) reported that intraventricular

(lateral ventricle) injection of the endorphins or enkephalins resulted

in the occurrence of wet-dog shakes in rats. However, intracisternal

(cisterna cerebellomedullaris) injection of these agents did not induce

wet-dog shaking. The ventricular space of the brain is comprised of

four ventricles, the cisterna magna (or cisternacerebellomedullaris),

several interconnecting foramen and the cerebral aqueduct. Most en-

dogenous cerebrospinal fluid (CSF) is produced in the two lateral ven-

tricles and from there flows through a large foramen (the foramen of

Munro) underneath the septal area, into the third ventricle. From the

third ventricle, the CSF passes through the narrow cerebral aqueduct, in

the midbrain, and enters the fourth ventricle. At the juncture of the

pons and medulla the fourth, ventricle loses its lateral walls which in-

clude the cerebellar peduncles. Posterior to this point the ventricular

space is known as the cisterna magna. These above-named spaces comprise

the most conmonly used ventricular spaces. Flow in these spaces is

most restricted in the cerebral aqueduct. Drugs injected into the

cisterna cerebellomedullaris diffuse easily into the fourth ventricle.

However, given a reasonable volume of injection fluid and reasonable

injection pressure, it is less likely that the drug will pass forward

through the cerebral aqueduct. Similarly, solutions administered in

either of the lateral ventricles should flow freely into the third

ventricle, but will encounter resistance at the cerebral aqueduct. Thus,

the data of Bloom et al. (77 ) suggests that endorphins induce WDS by

actions on forebrain structures.

However, their intracisternal injections were performed while the

subjects were under the influence of ether, while intraventricular in-

jections were performed in unanesthetized animals. It was suggested

that the ether may have been responsible for the inability of endorphins

to induce WDS when injected via the intracisternal route, but no control

was performed.

This possible interaction with ether can be controlled for, at least

to the extent of giving ether to animals injected intraventricularly.

Since kainate also results in WDS after intraventricular injection, this

same experiment, using kainate, could provide evidence on the general

locus of kainate's action and also provide a comparison of kainate- and

endorphin-induced WDS.

The parameters of reasonable volume and pressure are unanswered questions.
However, Bass ( 91 ) has estimated total CSF volume in the rat to be 250 pl
and its production rate to be about 1 pl/min and therefore injection
volumes should be kept small in relation to 250 pl and injection rates
around 1 pl/min.

In order to try to strengthen the belief that the injected solution

is hindered in its passage by the cerebral aqueduct, India ink was

injected in the same way as kainate and its localization examined.

Methods and Materials

Male, Long-Evans hooded rats weighing 200-250 grams were permanent-

ly implanted with plastic cannula guides stereotaxically guided to end

in the left lateral ventricle. The stereotaxic coordinates of the

cannula guide were -2.SAP, 4.5L, and 5.5V, Pelligreno and Cushman (71 ).

Kainate (0.25 nmoles) in 2.5 pl of saline or 2.5 pl saline was injected

through this guide via a 27-gauge hypodermic needle. Intracisternal

injections (2.5 and 5 pl) of kainate (0.1 mM) were performed in rats

anesthetized with ether. The rats were held by the ear bars of a

stereotaxic instrument with their noses downward to gain most effective

access to the cisterna magna. Injection was made through a 27-gauge

hypodermic needle attached to a microliter syringe. The needle had a

ball of solder attached to it about 4-5 mm from the tip to insure that

the needle could not slip beyond the ventricular space. The solutions

used for intracerebral injection were adjusted to a pH of 7.4.

When behavioral testing was completed, subjects were anesthetized

with sodium pentobarbital and injected with India ink in the same amount

and location as the kainate. They were then intracardially perfused

with 0.9% saline followed by 10% formalin. The brains were removed,

sliced, and examined for the presence of ink in the ventricular system,

and gross destruction of neural structures.

Seven rats were injected with 0.25 nmoles of kainate (in 2.5 pl

saline) icy. Six rats received 0.25 nmoles of kainate in 2.5 pl and

12 rats 0.5 nmoles in 5 pl, ic. Five additional rats were injected

intraventricularly with 0.25 nnoles of kainate while anesthetized

with ether.


In previous work we observed WDS in rats injected with saline,

icy. Saline produced less than 1 per hr which did not differ from

WDS observed in uninjected rats. Data from these saline-injected

rats were used for the determination of baseline levels of WDS in this


Icy injections of 0.25 nmoles of kainate in unanesthetized rats

induced WDS at a mean rate of 102.6 per hr. Icy injection in ether-

anesthetized rats induces WDS at a mean rate of 95.6 per hr. These

were both different from saline-injected animals (ps .001), but not

different from each other.

Intracisternal injections of 0.25 and 0.5 nmoles produced means of

7.2 and 7.8 WDS in an hour and medians of 3.5 and 6, respectively. The

number of WDS seen in those rats administered 0.25 nmoles of kainate

was not significantly different from saline-injected rats. Those ani-

mals receiving 0.5 nmoles of kainate did show more WDS than the saline-

injected animals (p < .01).

The two icy-injected groups were combined into one and this was

used for comparison with the ic-injected groups. Statistical tests

confirmed that icy injections produced more WDS than either ic injection

(ps < .001). These results are shown in Figure 3.


There was no significant difference in the number of WDS occurring

following intraventricular administration of kainate with or without

ether anesthesia. Thus, at least by this route of administration, ether



.- 50-

(a) (b) (c) (d)

Figure 3
Effectiveness of Injection Route in Inducing Wet-Dog Shakes
(a) kainate, icy; (b) kainate, icy with ether; (c) kainate,
ic 0.25nmol; (d) kainate, ic 0.5 nmol. Shaded area repre-
sents range of WDS in saline-injected rats.

does not seriously affect kainate-induced WDS. On the other hand,

intracisternal injection of kainate was virtually ineffective in in-

ducing WDS. These results suggest that kainate produces WDS by ac-

tions on forebrain structures.

This conclusion is strengthened by the examination of India ink

injections. Intraventricular injections of 2.5 pl of ink filled the

injected lateral ventricle and the whole third ventricle. On some

occasions, intraventricular injections of ink resulted in its presence

in the cerebral aqueduct and the fourth ventricle. Ink in these areas

was diluted (lighter in color) compared to that found in areas rostral

to the cerebral aqueduct.

On the other hand, intracisternal injection of India ink resulted

in the blackening of the fourth ventricle and adjacent portions of the

cisterna magna. In a few cases, when 5 pl was injected, the cerebral

aqueduct was colored. In one case, ink was also present in the posterior

third ventricle. These injections indicate that the cerebral aqueduct

forms a partial barrier to the flow of substances injected into the CSF

and strengthens the suggestion that the WDS seen after administration of

kainate are the result of actions on forebrain structures.

Ink did not enter the body of the contralateral lateral ventricle.

This distribution is not surprising since the exogenous flow of CSF is

from each lateral ventricle into the third ventricle. If the ink had

tried to flow into the opposite lateral ventricle it would have had to

turn nearly 3600 and flow against the exogenous current, both of which

it is unlikely to do.

Kainate-induced Lesion versus Kainate WDS


The hippocampus appears to be extremely sensitive to the neurotoxic

properties of kainate. Coyle and Schwarcz (52 ) reported that follow-

ing intrastriatal injections of kainate, they found selective destruc-

tion of the CA3 and CA4 fields of the hippocampus. Nadler, Perry, and

Cotman (69 ) found that intraventricular injections of less than 3 nmol/

p1 induce lesions which appear, in both Nissl- and silver-stained

material, to be selective for the CA3 and CA4 fields of the hippocampus.

This neurotoxic effect was observed after as little as 0.5 nnol/pl(lpl)

kainate. We have confirmed this selective lesion using Nissl-stained

material and observed that as little as 0.1 nmol/pl (2ul) kainate in-

duced neural destruction in about half the animals so injected. Other

reports (70 ,92 ) also attest to the extreme sensitivity of the hippo-


Low doses of kainate induce a highly specific behavioral response.

Slightly higher doses result in a highly specific lesion. These two

occurrences may be linked. It was hypothesized that kainate induces

WDS by its action on neurons of the CA3 and CA4 fields of the hippocampus.

Since the hippocampus lines the lateral ventricle, this hypothesis is

consistent with the results of the study just reported. This hypothesis

was investigated by injecting neurotoxic doses of kainate into rats to

destroy the CA3 and CA4 neurons and subsequently examining the ability

of kainate to induce WDS. This experiment was reported earlier in

Chapter VII.


The result of this test was that the initial dose of 1 nmol/pl

kainate, a dose associated with total destruction of CA3 and CA4, pre-

vented the appearance of WDS following subsequent injection of the lower

dose. This result is consistent with the hypothesis that kainate acts

on CA3 and CA4 neurons to induce WDS. However, it does not fully address

the question. It is possible that kainate acts on other neurons, e.g.,

entorhinal or septal neurons, and in this case the CA3 and CA4 neurons

form the output pathway. It is also possible that the initial kainate

altered, but did not destroy, neurons other than CA3 or CA4 neurons and

that this alteration was responsible for the observed reduction in WDS.

Selective Lesions of Hipocamal Subfields


The previous study strongly suggests that the CA3 and CA4 fields

of the hippocampus are part of the neural circuit by which kainate in-

duces WDS.

The aims of this study are 1) to show that the destruction of CA3-4

by itself can prevent kainate-induced WDS, 2) to investigate the im-

portance of various inputs to CA3 and CA4 to kainate-induced WDS, and

3) to investigate the importance of various outputs from CA3 and CA4

to kainate-induced WDS.

First, it is necessary to selectively lesion CA3 and CA4 to insure

that the reduction of WDS following kainic acid-induced lesions is attrib-

utable to the destruction of CA3 and CA4.

Then, assuming that CA3 and CA4 lesions do prevent WDS following

injection of kainate, it is most important to determine if kainate-in-

duced behavior is a result of kainate's actions on neurons afferent to

regio inferior and not directly on the pyramidal neurons themselves.

Finally, it is useful to try to understand what pathway out of the

hippocampus is necessary for the realization of WDS. This data would

help suggest how kainate-induced opiate withdrawal-like behavior is

related to true opiate withdrawal.

In order to carry out this study it is necessary to identify the

input and output pathways of the hippocampus. The hippocampus is com-

prised of two interlocking lobes, the dentate gyrus and Ammon's horn.

Ammon's horn has been further divided into four fields, CA(Cornu Ammonis)

1-4 (93). The primary bases for these distinctions are the size of the

principal neurons (the pyramidal neurons), their packing density (or how

restricted they are to lamellar organization) and the mixture of differ-

ent size pyramidal cells. In particular, CA1 is comprised only of

medium-size pyramidal neurons tightly organized into one layer. On the

other hand, CA3 is composed only of large pyramidal neurons organized

into one thin layer, while CA4 is composed of a scattered pattern of

large pyramidal neurons. Field CA2 is an area between CA1 and CA3

where there is a mixture of medium and large pyramidal neurons. These

fields can also be examined in terms of input and output connections.

On this basis it is difficult to distinguish between CA3 and CA4, and

CA2 appears simply to be an overlapping of CA1 and CA3 (94). It has been

suggested that at our present level of understanding it is sufficient to

divide Ammon's horn into two fields, regio superior and regio inferior

(95). These correspond to CAl-2 and CA3-4, respectively. Since they

also correspond to kainate-insensitive and kainate-sensitive areas, in

regard to the toxic effects of a 1 nmol/pl solution, this simplified

description will be used in looking at input and output connections of

the hippocampus.

Since the focus of this study has already become set on the regio

inferior we can narrow our original goals to affecting its input and out-

put. Input to CA3-4 pyramids comes from two directions, anterio-lateral

and postero-medial. The antero-lateral intrusion comes in the form of

the fornix which contains fibers from the septal area, hypothalamus and

other areas (96 ). The fornix just ahead of the hippocampus is a fiber

bundle separate from any of the structures it innervates or whose axons

make it up. Thus, at this level the fornix can be discretely destroyed.

Two complications arise, one of which can be rectified. The fimbria,

which contains the axons of CA3-4 neurons (that is, the output of regio

inferior), also travels in the fornix (96). Thus, destruction of the

fornix destroys both input and output of CA3-4. However, the ability to

discretely destroy a fiber bundle is too good an opportunity to pass up,

and much easier than trying to destroy all the structures which make up,

or are innervated by, the fornix.

Axons of CA3-4 neurons in the fimbria also pass through the hippocamp-

al cormissure psalteriumm) into the contralateral fimbria (96 ). This

occurs just anterior to the hippocampus. Since this constitutes an out-

put pathway it is important to destroy it. To accomplish this it is

necessary to destroy the fimbria-fornix just anterior to the hippocampus.

To do this without damaging cell bodies of the hippocampus is probably

most easily accomplished by cutting the fornix-fimbria with a knife.

The other major inputs to CA3-4 are the mossy fibre bundle, the

perforant pathway and the dorsal fornix/cingulum (96 ,98 ) The mossy

fibre bundle arises from the dentate gyrus and synapses exclusively

with CA3-4 neurons. The perforant pathway,which originates in entorhin-

al cortex, terminates on the dentate gyrus granule cells and also sends

fibers over and past the dentate gyrus to synapse with the apical den-

drites of the CA3 and CA4 pyramidal neurons. The dorsal fornix carries

fibers from various locations including the locus coeruleus and the

raphe nuclei ( 97 98 ). These fibers enter the hippocampus from the back

taking a pathway between CA1 and the dentate gyrus. They distribute to

all regions of the hippocampus.

To eliminate the mossy fibre bundle the dentate gyrus must be des-

troyed. This will also strongly reduce the effect the perforant path-

way has on the CA3-4 neurons because a large number of perforant path

fibers terminate on neurons of the dentate gyrus and excite those neurons

which, in turn, excite the CA3-4 neurons. If the lesion used to damage

the dentate gyrus is allowed to extend slightly dorsal to the dentate

gyrus into the superficial molecular layer of CA1 almost all the perfor-

ant path fibers can be destroyed. This manuveur will also eliminate

many of the fibers from the dorsal fornix that would have terminated in

CA3 or CA4. Thus, an extensive lesion of the dentate gyrus can destroy

most of the input to the CA3-4 neurons from the postero-medial direction.

Regio superior acts as an output pathway for regio inferior. Axon

collaterals of region inferior neurons, the Schaeffer collaterals, synapse

on CAl-2 neurons ( 95). The CAI-2 neurons send axons out to synapse in

the subiculum. Destruction of CAl-2 can provide information about the

pathway out of the hippocampus necessary for the realization of kainate-

induced WDS. It can also help suggest the function of the fornix. If

destruction of either the fornix or regio superior prevents kainate-

induced WDS, then the most reasonable explanation would be that the

fornix is important as an input to regio inferior. This is so because

CAl-2 appears to act only as an output pathway for CA3-4, not as an in-

put ( 99).

Therefore, four kinds of lesions should suffice to indicate how

kainate-induced WDS are produced. In summary these are:

1) selective destruction of CA3-4

2) destruction of the fornix (input and output)

3) large dentate gyrus lesion (input)

4) CAl-2 destruction (output)

Methods and Materials

The subjects for this study were 31 male, Long-Evans hooded rats.

At the time of surgery they weighed from 250-375 grams. Surgical pro-

cedures were performed while the subjects were anesthetized by chloral

hydrate (400 mg/kg; 500 mg/ml) which was administered intraperitoneally.

The head was shaved and the animals secured in a stereotaxic device.

The head was tilted upward five degrees to allow the use of the atlas of

the rat brain by Pelligreno and Cushman. The skin was split with a

scalpel and the fascia retracted from the skull with a blunt scalpel

handle. The skin and fascia were maintained in a retracted position

by means of S-shaped pins attached by rubber bands to the stereotaxic


The stereotaxic was used to loacte the areas of the skull above the

structures of interest. A hand-held drill with a one mm diameter bit

was used to drill through the skull. The bit extended only one mm out

of the drill to protect against damaging cortex. After the holes were

drilled, bone chips were removed with the aid of microdissecting for-

ceps and a dissecting microscope until the dura matter was unobstructed

and flat across the bottom of the hole. In those holes where electrodes

or cannulae were to be lowered, the dura and pia matter were split using

the beveled edge of a 27 gauge needle.

Lesion devices were directed at four areas. In ten animals the for-

nix-fimbria was cut. A small knife, 2.5 mm wide, made from a razor

blade was used to make the cut. This knife was directed at the fornix-

fimbria just anterior to the hippocampus and posterior to the psalterium.

The knife was placed at the following coordinates of the Pelligreno and

Cushman atlas--P0.2mm, L 0.5mm, V 4.5 mm.

In seven animals, destruction of CA3 and CA4 was attempted. In

these animals, a 400 pm stainless steel wire insulated except for the

final 0.5 mm, was inserted into the hippocampus at the following loca-

tions: P1.5, L2.5,V4; P2.8, L3.7, V4; P4, L5.2, V5; P4, L5.2, V7. At

each of these locations 1.5 mA of DC current, produced by a Grass lesion

maker, was passed for 15 sec. A banana plug inserted into the rectum

formed the other pole of the circuit.

Electrolytic lesion of the dentate gyrus was attempted in seven rats.

Current (1.5 mA) was passed for 15 sec at the following sites: P1.5, L1,

V3.6; P2.8, L2.3, V3.6; P4, L3.5, V4.3; and P4, L5, V7.5.

Lesions were directed at the CA1 field of the hippocampus in seven

rats by passing 1.5 mA of current for 10 sec at each of the following

locations: P1.8, L1, V2.8; P2.8, LI.2, V2.8; P2.8, L2.5, V2.3; P4, L2.7,

V215; and P4, L5, V4.2.

Immediately following the creation of lesions, two small screws were

placed in holes drilled into the skull. A stainless steel cannula guide

was then lowered through a hole drilled at P2.4, L4.75 to a depth of

5 mm from the skull. This placement left the tip of the guide in the

left lateral ventricle just dorso-lateral to the ventral hippocampus.

The cannula guide was constructed from a 23 gauge (0.61 mm dia.) syringe

needle. The needle was broken 5 mm down from the hub by bending it to

cause metal fatigue. The hub, which is larger than the 1 mm skull hole,

served as a guard to prevent the guide from going too far into the brain.

Once the guide was in place, it and the two screws were surrounded

by dental cement to form a solid cap. The skin was then placed over this

cap, but around the hub, and closed with wound clips. Bicillin, im,

(60,00OU) was administered. The subject was removed from the stereotaxic

and allowed to recover in its home cage for seven to ten days before


Just prior to testing, a 50 pM solution of kainate was prepared by

dissolving it in saline. The solution was brought up to a pH of 7.4

with NaOH. The subjects were transported to the observation area in

their home cages. Injections were performed by gently hand-restraining

the subject and inserting a 30 gauge needle into the cannula guide.

When the guides are constructed the implanted end becomes slightly con-

stricted. Because of this, only the beveled tip of the 30 gauge needle

(< 1mm) entered the lateral ventricle. The close fit also ensured the

absence of backflow of solution into the guide. The 30 gauge needle

was attached to one end of a length of polyethylene tubing which was

filled with kainate. The other end was attached to a microliter syringe.

Two microliters of the kainate solution (0.2 nmol, 43 ng) was injected

over a 30 sec period. The rats were observed, as described in Chapter

II, for one hour. The number of wet-dog shakes was recorded, as well

as the number of tonic-clonic convulsions and the general deportment of

the subject.

At the end of behavioral testing, the subjects were given an over-

dose of chloral hydrate, and intracardially perfused with saline followed

by 10% formalin. The brain was removed and stored in cold 10% formalin.

At a later time, the brains were frozen and cut into 30 pm sections.

Brains possessing electrolytic lesions were cut coronally and every

eighth section saved and mounted. Brains in which a knife cut was made

were cut horizontally and every fourth section through the fornix re-

tained and mounted.


The fornix-fimbria knife cuts proved almost impossible to verify

histologically. Therefore these animals will not be presented. In

regard to animals receiving electrolytic lesions, those in which the

cannula obviously ended up outside the ventricle were thrown out prior

to analysis (n = 2). One additional animal was not subjected to the

analysis because of widespread, unexplained damage throughout the third


The results from the remaining subjects were analyzed in two ways.

In the first, all those subjects sustaining any CA3 damage at all were

put into one group. Other groups consisted of the subjects in which

hippocampal damage was restricted to the dentate gyrus, the CA1 field

or to a combination of these fields. Another group was comprised of

subjects sustaining damage only in the cortex overlying the hippocampus.

These groups were compared to 14 nonlesioned rats. The second method for

examining the data consisted of determining the lesions sustained by

those subjects showing less WDS following application of kainate than

any of the nonlesioned subjects.

There were eleven rats which sustained at least some damage to CA3

or CA4 along with varying amounts of damage to other structures. The

number of WDS induced in these animals by kainate ranged from zero to 143.

The mean number of WDS was 19.8, while the median was one. Two rats sus-

tained hippocampal damage restricted to CA1. These two subjects produced

17 and 88 WUS in response to kainate administration for a mean (and

median) of 52.5. One rat had hippocampal damage limited to the ventral

blade of the dentate gyrus. That rat was observed to evoke 14 WDS and

three tonic-clonic convulsions. Two subjects sustained damage to both

CA1 and the dentate gyrus. These lesions, combined with kainate, re-

sulted in 125 and 173 WDS. Two rats were found to have damage limited

to the overlying cortex and were observed to respond to kainate adminis-

tration with 30 and 75 WDS. This works out to be a mean of 52.5 WDS.

Finally, injection of kainate into 14 nonlesioned animals produced a

range of five to 112 WDS--a mean of 45.5 and a median of 42.

Statistical analysis was possible only to compare rats with at

least some CA3 damage to those having no lesions at all. The result of

a Mann-Whitney U test revealed a significant change (reduction) in the

WDS produced by the group sustaining CA3 damage (p <.01). While no

analysis can be performed for the remaining groups it is worth noting

that either cortical or CA1 damage alone yielded means of 52.5 WDS, not

far from the 45.5 of the nonlesioned rats. Combined damage of CA1 and

the dentate gyrus yielded a mean of 149, far above that seen in nonlesioned


The smallest number of WDS observed after kainate application in

nonlesioned rats was five. Nine rats given electrolytic lesions pro-

duced four or less WDS. All nine sustained damage of CA3. The damage

was generally extensive to the cellular field and/or the fimbria. Damage

to any other particular region within or outside the hippocampus was not

consistently seen in these nine subjects.


The results of the lesion experiments are anything but clean. How-

ever, both in terms of the number of WDS produced by CA3-damaged rats and

the lesions sustained by animals showing subnormal numbers of WDS, damage

to regio inferior is strongly correlated with a reduction in the ability

of kainate to induce WDS.


Animals not sustaining CA3 damage showed no obvious trend toward

a reduced number of WDS. Any such trend was, in fact, in the opposite

direction--towards increased numbers of WDS.



In Chapter II it was stated that WDS are a symptom of morphine

withdrawal in the rat. Furthermore, as the name suggests, WDS are

a common response of dogs and most other mammals, including rats

(73), to immersion in water. More recently, a number of chemicals

have been found to induce WDS in opiate-naive, dry rats. These were

listed in Chapter II. In Chapter VIII it was reported that kainate-

induced WDS were similar to endorphin- and enkephalin-induced WDS in

at least one respect. They are all induced by intraventricular, but

not intracisternal administration.

It is useful to further compare agents which will induce WDS. It

is possible that all of these agents act on the same site or the same

neural circuit, but this hypothesis should be subject to testing.

Since kainate-induced WDS can be prevented by pretreatment with a

selective neurotoxic dose of kainate, this same pretreatment can be

used to compare kainate-induced WDS to those induced by other agents.

If all WDS-inducing agents are acting on the same neural circuit, then

this manipulation should also prevent WDS induced by agents acting on

neurons afferent to, and including, those destroyed by the pretreatment.

Similarly, manipulations which affect the production of WDS by various

agents should have similar effects on kainate-induced WDS. These possi-

bilities were examined in the following series of studies.


Opiate Withdrawal


Introduction. Naloxone is recognized as an effective opiate antago-

nist, with little or no opiate agonist properties at commonly used doses.

In animals made dependent upon morphine, administration of naloxone (by

various routes) will result in the appearance of the morphine abstinence

syndrome ( 73). The intensity of this syndrome, when induced by naloxone,

may be greater than that seen during simple morphine abstinence. How-

ever, naloxone, administered to an opiate-naive animal does not induce

the withdrawal behaviors.(73).

If kainate were acting by stimulating the endogenous mechanism for

the abstinence behaviors, then naloxone should have either no effect

or enhance kainate-induced WOS.

Methods and materials. Surgical and intraventricular injection

procedures were the same as described in Chapter VIII. Either naloxone

or saline was injected i.p., 5 min prior to icy injection of 0.25 nmoles

of kainate in 1.5 l saline. Three groups of naloxone-treated animals

receiving different doses were used along with separate saline control

groups. Ten animals received 1 mg/kg naloxone (2.5 ml/kg), and seven

received a similar volume of saline. Nine animals received 2 mg/kg

naloxone (5 ml/kg), while six received saline and served as controls.

Six rats were injected with 4 mg/kg naloxone (5 ml/kg) and six animals

were saline controls.

Results. The two lower doses of naloxone (1 and 2 mg/kg) did not

alter the number of WDS produced by the icy injection of kainate. The

4 mg/kg dose of naloxone reduced the kainate-induced WDS by 88%. The

mean WDS of the 4 mg/kg naloxone-treated group was 12.2 and the mean

of the control group was 103 per hr.

Discussion. Naloxone antagonizes kainate-induced WDS. This suggests

that kainate-induced WDS are mediated by opiate receptor stimulation.

This is in contrast to morphine withdrawal where loss of opiate receptor

stimulation appears to be the precipitating factor. However, naloxone

prevents endorphin-induced WDS (77), again indicating that kainate and

endorphins have similar actions. Furthermore, kainate- and endorphin-

induced WDS were only antagonized by rather high doses, greater than

1 mg/kg, of naloxone.


Introduction. Once the morphine abstinence syndrome has been pre-

cipitated, it can be antagonized by administration of morphine. Thus,

if kainate directly induces the morphine abstinence syndrome, morphine

would be expected to reduce the behavioral manifestations of kainate.

At a moderate dose, morphine will induce active behaviors like

grooming and quick, darting movements in rats. The effect of kainate

on this morphine action can be examined at the same time and provide

additional information about the relationship between morphine and


Methods and materials. Morphine sulfate 10 pmoles/kg (0.4 ml/kg)

i.p., was injected 30 min prior to icy injection of kainate. Nine

animals were injected with both systemic morphine and icy kainate;

eight received systemic saline (0.4 ml/kg) and kainate icv; and seven

received systemic morphine with saline icy. WDS, grooming behaviors,

and the presence of quick, darting movements and turns were recorded

for 1 hr. Grooming was counted every 15th sec. Since observations

were made for 1 hr, a maximum grooming score of 240 was possible.

Results. The results are summarized in Figure 4. Systemic injection

of morphine along with saline, icy, induced a mean grooming score of 69.

Quick, darting movements were observed in six of the seven animals treat-

ed in this way. WDS occurred infrequently (4 WDS per hr).

Icy kainate and systemic saline induced WDS, but little grooming

and no quick darting and turning. Grooming occurred at a mean rate of

10 per hr. Icy kainate and systemic morphine also induced WDS and a

small amount of grooming (mean = 9). Only one of these rats showed quick,

darting movements. This one rat produced only 1 wet-dog shake and, on

post-mortem examination, no evidence of ink was found in the ventricles.

This indicated that the intraventricular injection was not successful.

The animal acted as if it had only received systemic morphine.

The number of WDS produced following icy kainate injection did not

differ between the groups receiving morphine or saline systemically.

The group receiving kainate icy and systemic morphine had more WDS than

the one administered saline icy and systemic morphine (p < .01). The

group injected with kainate icy and systemic saline also had more WDS

than the group receiving intraventricular saline and systemic morphine

(p < .001).

Grooming scores for both kainate-injected groups did not differ, but

both groomed significantly less than the one administered saline with

systemic morphine (ps <.02). A smaller number of the rats injected with

kainate icy and systemic saline presented quick, darting movements than

those administered saline icy plus systemic morphine (p <.01). Fewer of

the rats receiving kainate icy with systemic morphine showed quick, dart-

ing movements than those receiving saline icy and systemic morphine (p <.05).

The two groups receiving kainate injections did not differ from each other.






Figure 4

Competition between Kainate and Morphine Behaviors

KS, kainate icy and saline ip; KM, kainate icy and
morphine ip; SM, saline icy and morphine ip.
Shaded area represents range of WDS in saline-
injected rats.


Discussion. Morphine, at this dose, does not prevent kainate-induced

WDS. This is not consistent with the hypothesis that kainate activates

the morphine abstinence syndrome. Thus, in regard to acute administra-

tion of morphine and naloxone, kainate-induced WDS do not respond as

morphine abstinence. In fact, kainate-induced WDS react in an opposite

manner. On the other hand, kainate did reduce morphine-induced behavior

suggesting an interaction between the two chemicals at some level.



Introduction. Bedard and Pycock (76 ) reported that systemic ad-

ministration of 5-HTP, following pretreatment with a peripherally act-

ing decarboxylase inhibitor, resulted in the occurrence of WDS. 5-HTP

is the metabolic precursor of the putative neurotransmitter, serotonin.

This response lasted for several hours. The maximum rate of WDS was

observed 2 hr after administration of 5-HTP.

The result of this pharmacological manipulation might be expected

to be due to increased amounts of serotonin in the brain. Consistent

with this expectation, Bedard and Pycock ( 76 ) reported massive in-

creases in cerebral serotonin content following systemic 5-HTP injection.

If kainate were inducing WDS by increasing serotonin receptor stimu-

lation, then a kainate-induced lesion should not affect 5-HTP-induced

WDS unless the kainate-sensitive neurons were the serotonergic ones.

In the latter case, the loss of serotonergic presynaptic endings could

result in a lowered rate of conversion of 5-HTP to serotonin and a

lowered rate of WDS. On the other hand, if 5-HTP application resulted

in WDS because of increased glutamate (kainate) receptor stimulation,

then a lesion which reduces kainate-induced WDS should equivalently re-

duce 5-HTP-induced WDS.

Methods and materials. General surgical procedures were the same

as used in previous studies. Holes at the coordinates for the lateral

ventricles (P2.5, L4.7) were drilled on both sides of the skull. A 30

gauge needle, attached to a microliter syringe by polyethylene tubing,

was lowered into each lateral ventricle--one at a time. A volume of

1.5 pl of a Inmol/pl solution of kainate, or the vehicle solution, was

injected over a 90 sec period into each lateral ventricle. The needle

was left in place an additional 45 sec and then slowly removed. Eight

rats received kainate solution and seven were administered the vehicle

solution. Following these injections, the scalp wound was closed and

the animals allowed to recover. The animals were tested 7 to 10 days


Carbidopa (25 mg/kg), a peripherally acting decarboxylase inhibitor,

was injected 30 min prior to 5-HTP (100 mg/kg, i.p.). Ninety min after

5-HTP was injected, the subjects were placed in the observation boxes

and wet-dog shakes were recorded for 1 hr. Bedard and Pycock (76

reported that the maximum rate of 5-HTP-induced wet-dog shakes occurred

during this period of time.

Following behavioral testing, the animals were given an overdose

of chloral hydrate, and perfused, intracardially, with saline followed

by 10% formalin. The brains were removed and later prepared for micro-

scopic analysis.

Results. Seven of the eight kainate-injected rats were found to

have extensive bilateral lesions of CA3 and CA4. The other subject

showed only a unilateral lesion which did not destroy all of CA3 or CA4.

Control injections produced no lesions beyond the cannula track.

Injection of 5-HTP in control animals produced a behavioral constella-

tion consisting of wet-dog shakes (Mn = 35.3; Mdn = 44), scratching,

stretching and yawning. These animals were also very active for the

entire observation period.

In animals pretreated with kainate the number of WDS induced by

5-HTP was less (Mn = 12; Mdn = 3.5) than in control animals. This

difference was significant (p <.03). The one kainate-treated rat

showing only a partial lesion accounted for 65% of the WDS seen in

the experimental group. If he is removed for the statistical analysis,

the difference between groups is significant at the p <.005 level.

Discussion. Lesions of the CA3-4 region of the hippocampus strongly

reduce 5-HTP-induced WDS. After removing the one rat showing only a

partial lesion, the total reduction was about 90%. This is about the

same as the amount of reduction in kainate-induced WDS following the

same kind of lesion. This data suggests that 5-HTP acts on neurons

afferent to kainate-sensitive neurons or that the kainate-sensitive

neurons are serotonergic.


Introduction. If kainate induces WDS by releasing serotonin, then

blockade of serotonin synthesis should reduce the effects of kainate

administration. PCPA (parachlorophenylalanine) has been reported to

competitively block tryptophan hydroxylase, an essential enzyme for

the synthesis of serotonin (100). Administration of 100 mg/kg daily for

three consecutive days has been shown to lead to a 90% decrease in the

total brain concentration of serotonin at 48 hr after the third injec-

tion (100).

Methods and materials. Thirteen male, Long-Evans rats were implant-

ed with a permanent 23 gauge cannula guide as described in Chapter VIII.

One week later, seven of these rats were injected with 100 mg/kg of

PCPA, given intraperitoneally, once a day for three consecutive days.

The six remaining subjects acted as controls and were injected with the

vehicle for PCPA (saline plus 0.01% HC1) only. Forty-eight hours after

the third injection all subjects received 0.2 nmol of kainate (2 p1)

injected into the lateral ventricle in the manner previously described.

They were immediately placed in observation boxes and wet-dog shakes

counted for one hour.

Results. Injection of kainate in vehicle-treated rats induced WDS

(mean = 120.4; median = 92). This same injection in the seven PCPA-

treated animals induced a mean of 119.6 WDS (median = 129). This differ-

ence was not significant by either nonparametric tests or the t-test.

Discussion. Administration of PCPA to rats did not significantly

affect the number of WDS induced by 2 pl of a 0.1 nmol/pl solution

of kainate. Since it is expected that the PCPA strongly reduced the

amount of serotonin in the brain, this result suggests that kainate is

not inducing WDS by releasing serotonin.

In relation to the preceding study, this result suggests that a

kainate-induced lesion does not reduce 5-HTP-induced WDS by destroying

serotonergic neurons.

A schema in which serotonergic neurons are afferent to kainate-

sensitive neurons would be consistent with both results. Since kainate

induces WDS via a neural mechanism including the CA3-4 region of the

hippocampus, this serotonergic influence might be terminals of raphe

nuclei on the CA3-4 pyramidal neurons.



The endorphins and enkephalins induce WDS following intraventricular,

but not intracisternal administration (77 ). This manifestation of

their administration is prevented by naloxone (77 ), but not morphine

(Bloom, personal communication). The number of WDS induced by kainate

are affected in a similar manner by these same manipulations. The

similarity of kainate- and endorphin-/enkephalin-induced WDS was further

examined by testing the ability of kainate-induced lesions to prevent

endogenous opiate-induced WDS. Such lesions do prevent kainate-induced


Methionine-enkephalin (met-enkephalin) was used because it was most

readily available. Urca, Frenk, Liebeskind and Taylor (101) reported

that 100 pg of met-enkephalin (in 10 pl) resulted in WDS in rats.

Methods and Materials

Seven rats were bilaterally injected with 1.5 pl of a 1 nmol/pl

solution of kainate one week before testing to induce lesions of the

CA3-4 regions. Nine more male, Long-Evans hooded rats received the

vehicle for kainate alone. Following injection of kainate or its vehicle,

a 23 gauge cannula guide was implanted as described in Chapter VIII.

On the test day, met-enkephalin was dissolved in saline shortly

before use. A solution containing 14 pg/pl (approximately 20 nmol/l)

was prepared. Five microliters of this solution was administered to

each rat. Immediately after the injection, the subjects were placed in

individual observation boxes and observed for one hour. Following be-

havioral testing, the subjects were sacrificed and their brains prepared

for microscopic examination.


Histological examination of the brains revealed that kainate pre-

treatment resulted in complete, but selective lesions of the CA3-4 area

of the hippocampus. The only lesion seen in vehicle-pretreated subjects

was that associated with the cannula itself.

In vehicle-pretreated rats, enkephalin induced a mean of 11.9 WDS

in an hour. They also spent considerable time engaged in maintenance

behaviors. Pretreatment with kainate reduced the number of WDS resulting

from administration of met-enkephalin by 88%. This was significant at

the p <.02 level (Mann-Whitney U test).


Pretreatment of rats with a neurotoxic dose of kainate reduced the

number of WDS induced by met-enkephalin. The reduction in terms of

percent of control behavior was similar to that seen for 5-HTP-

and kainate-induced WDS. This result further indicates a similarity

between kainate and endogenous opiate-induced WDS.



Wet-dog shaking in rats can be induced by endogenous opiates and

this action can be prevented by naloxone. In pharmacological terms

this indicates that WDS can be induced by stimulation of opiate recept-

ors. However, the dose of naloxone required to block this opiate be-

havior is higher than required to block most other opiate-induced be-

haviors. For example, B-endorphin will induce WDS followed by cata-

lepsy. A dose of 1 mg/kg of naloxone will reverse the catalepsy, but

the WDS reappears. A dose of 2 mg/kg of naloxone will prevent the

occurrence of WDS ( 77). This might suggest that a specific type of

opiate receptor is responsible for WDS.

Etorphine, a potent morphine-like agonist, has been reported not to

induce WDS (102) and morphine will only induce WDS at doses far beyond

those necessary to induce analgesia (79 ). These data could also be

taken to imply that endogenous opiate-induced WDS is mediated by a

second type of opiate receptor, one that is relatively insensitive to

morphine or naloxone.

The presence of such second opiate receptors has been suggested

(103,104) and ketocyclazocines have been suggested as selective agonists

(103). It has been found that intraventricular injection of small

amounts of ketocyclazocine itself ( 2pl of 1 nmol/pl solution) will

induce WDS. This response was prevented by 10 mg/kg of naloxone, but

not 1 mg/kg of naloxone (Lanthorn, Smith and Isaacson, submitted).

These results strongly suggest that endogenous opiate-induced WDS

is mediated by a second opiate receptor known as the kappa (K)-opiate

receptor. If this is true, then kainate-induced lesions should reduce

ketocyclazocine-induced WDS as they reduce kainate- and met-enkephalin-

induced shaking.

Methods and Materials

Surgical and injection procedures were identical to those of the

preceding experiment. Seven rats were treated with kainate, while eight

received vehicle only. Ketocyclazocine (Sterling-Winthrop) was dissolved

in saline containing 2 mM HC1. The ketocyclazocine solution was prepared

shortly before each use. Two pl of a 1 nmol/pl (0.29 pg/pl) solution

of ketocyclazocine was injected into each subject. Observation began

immediately after the injection.


All seven kainate treated rats showed extensive CA3 and CA4 lesions.

No lesions were evident in control animals except for the cannula track.

In saline-injected animals, ketocyclazocine produced a behavioral

constellation consisting primarily of intensive grooming and wet-dog

shakes. The number of WDS induced was reduced 86% in rats pretreated

with kainate. This change was significant at the p <.01 level.


The appearance of WDS as a consequence of administration of ketocy-

clazocine can be largely prevented by pretreatment with a neurotoxic

dose of kainate. This result strengthens the hypothesis that endoge-

nous opiate-induced WDS is mediated by the K-opiate receptor by show-

ing that endogenous opiates and a K agonist utilize the same neural

circuit. At the same time it strengthens the belief that opiates and

kainate act on the same neurons to induce WDS.

Sodium Valproate


It has been reported that intraperitoneal administration of sodium

valproate, or dipropylacetate, will induce WDS and other morphine with-

drawal-like behaviors in rats (81 ,84 ). Sodium valproate is used as

an antiepileptic agent and, at least at high doses, increases the con-

centration of GABA in the brain. This increase in GABA is probably a

result of competitive inhibition of succinic semialdehyde dehydrogenase

(SSA-DH), an enzyme in the catabolic pathway of GABA.

The induction of WDS by sodium valproate can be blocked by low

doses of morphine (1 mg/kg) and were released by naloxone (1 mg/kg).

This is similar to how morphine withdrawal responds to these agents,

but is opposite to how kainate- and endorphin-induced WDS respond to

these chemicals.

Methods and Materials

A neurotoxic amount of kainate solution or an equal volume of

vehicle solution was administered (as described in the section on 5-HTP)

to six and thirteen male, Long-Evans hooded rats, respectively. Valproic

acid (300 mg/kg) or sodium valproate (345 mg/kg) was injected intra-

peritoneally immediately prior to observation.


Valproic acid (300 mg/kg) was injected into six rats pretreated

with the vehicle for kainate. These rats became very inactive and

were ataxic for 30-45 min. Following this period a couple of the sub-

jects produced sporadic wet-dog shakes. Such severe motor depression

was not reported by deBoers et al. (81 ) following similar doses of

sodium valproate. Therefore, the valproic acid was converted to its

sodiLn salt and an equimolar dose (345 mg/kg) was injected.

Sodium valproate was injected into six rats pretreated with kainate's

vehicle. They became somewhat inactive, took on a hunchback posture,

urinated excessively and produced wet-dog shakes (mean = 18.8; median

= 15). In seven animals pretreated with neurotoxic doses of kainate,

sodium valproate induced increased locomotion and exploration, excess-

ive grooming, excessive urination and wet-dog shakes (mean = 21;

median = 21).

The difference in the number of wet-dog shakes produced by the two

groups was not statistically significant. All shaking in both groups

occurred within 25 min of sodium valproate administration.


The wet-dog shake response to sodium valproate does not appear to

be affected by kainate-induced lesions. This indicates that not all

factors which induce WDS do so via kainate-sensitive neurons. The data

does not suggest whether sodium valproate acts on neurons directly

downstream from kainate-sensitive neurons or on a parallel circuit.

It is intriguing that sodium valproate- and kainate-induced WDS

react differently to naloxone and morphine and also to kainate-induced




Rats immersed in cold water will produce WDS and diarrhea (105).

This response is most consistently seen in moderately anesthetized

rats in which most escape behaviors are inoperable. This response to

cold water can be antagonized by morphine (10 mg/kg) (105) and prevent-

ed by a lesion at the medial parts of the diencephalic-mesencephalic

juncture (105), a lesion which also prevents naloxone-induced WDS in

morphine-dependent rats (106). Thus this response to an environmental

alteration reacts to certain manipulations in the same manner as mor-

phine withdrawal.

It has been reported that normal rats, anesthetized with 40 mg/kg

of sodium pentobarbital, dunked in cold water (40C) and then held in

the water for 5 min with their heads raised out of the water, exhibited

an average of 10-15 shakes (105).

Methods and Materials

The subjects for this study were those which comprised Group 3 in

Chapter VII. Four days after receiving their last injection of test

solution, they were anesthetized with sodium pentobarbital (40 mg/kg).

Thirty minutes later they were briefly completely immersed in ice-

water (40 C). Then the heads were raised out of the water and the

animal held in that position for 5 min. The number of WDS occurring in

that time was recorded.


All the rats subjected to this regimen exhibited WDS. A range of

7-20 WDS was observed (median = 10).



A neurotoxic amount of kainate does not abolish the shaking response

of rats to immersion in cold water. In fact, the number of shakes seen

in these rats appeared to be the same as the number reported by others

in normal rats (105).



These results indicate that kainate, an analogue of glutamate,

is a behaviorally active substance of tremendous potency. As little

as 3 ng injected through the lateral ventricle will induce a statistical-

ly significant alteration in the ongoing behavior of a rat. Concentra-

tions of kainate of less than 0.1 nmol/pl were most consistently corre-

lated with increased numbers of wet-dog shakes (WDS) exhibited by the

rats and were not correlated with obvious neurotoxicity. Higher con-

centrations of kainate were correlated with convulsions and with neuronal

degeneration. A concentration of 1 nmol/pl induced selective destruc-

tion of the CA3 and CA4 fields of the hippocampus.

The WDS-inducing effect of kainate appears to be localized to fore-

brain structures since intraventricular, but not intracisternal, in-

jection of kainate resulted in WDS. This effect can be further localized

to some circuit including CA3 and CA4 because kainate-induced lesions of

CA3 and CA4 prevent the induction of WDS by subsequent administration of

kainate. Electrolytic lesions of this same area also reduce kainate's

ability to induce WDS. Lesions not involving CA3 and CA4, but involving

CA1 or the dentate gyrus did not appear to decrease the effect of kainate,

but statistical confirmation was not possible because the groups were too


WDS is a conspicuous component of morphine withdrawal in the rat.

The WDS induced by kainate were not affected by a moderate dose of


morphine and were blocked by a high dose of naloxone. This suggests that

an opiate receptor may be involved in kainate-induced WDS, but not in the

same way that morphine withdrawal is.

Substances which induce WDS in opiate-naive rats include 5-HTP (the

metabolic precursor to serotonin), met-enkephalin (an endogenous opiate),

ketocyclazocine (agonist of the kappa opiate receptor) and sodium val-

proate (an anticonvulsant with influence on GABA metabolism). Lesions,

of the type which prevent kainate-induced WDS, were used to test the

similarity of these agents to kainate in the manner in which they induce

WDS. Destruction of CA3 and CA4 prevented 5-HTP, met-enkephalin and

ketocyclazocine from inducing WDS. The WDS-inducing ability of sodium

valproate was not affected by this lesion. The wet-dog shake response

of rats to immersion in cold water is not affected by a lesion which

blocks kainate-induced WDS.

There are several points of interest which these studies can be used

to speak to. First is whether or not kainate-induced behavior is direct-

ly related to morphine withdrawal. The evidence presented here suggests

that kainate-induced behavior is not synonomous with morphine withdrawal.

In particular, it would be expected that morphine should counter the signs

of withdrawal while a morphine antagonist, such as naloxone, should enhance

these behavioral symptoms. In sharp contrast to expectation, morphine

had no effect on kainate-induced behavior while naloxone, albeit at a

high dose, reduced kainate-induced behavior. Thus, pharmacologically,

kainate-induced behavior is not identical to withdrawal as it is normally


Anatomically, also, morphine withdrawal and kainate-induced behavior

can be distinguished. Withdrawal can be elicited in morphine-dependent

animals by administration of naloxone or other antagonists (73). In-

jection of opiate antagonists into the brain itself is effective when

the antagonist is allowed to interact with structures lining the fourth

ventricle and cerebral aqueduct (107). Injections restricted to the

lateral and third ventricle or application directly into the hippocampus

are not effective (107,108). On the other hand, kainate appears to in-

duce WDS by its action on the hippocampus. It would not be surprising

if more than one part of the brain is involved, but this data does indi-

cate that kainate-induced behavior does not involve all of the neural

circuit employed by morphine withdrawal.

Coronally-oriented knife cuts of the medial part of the brain

abolish withdrawal induced by systemically injected naloxone only when

those cuts are made in the caudal brainstem (106), caudal to the peri-

aqueductal region. This suggests that the withdrawal effects arising

from this naloxone-sensitive region are caudally directed. Destruction

of the CA3 and CA4 fields of the hippocampus, preventing kainate-induced

behavior, does not affect ice-water induced shaking. However, ice-water

induced shaking is prevented by morphine and by lesions similar to those

which prevent withdrawal (105). These results suggest that if kainate-

induced and morphine withdrawal-induced behavior are wired in series,

the kainate influence must be upstream from morphine's site of action.

In such a scheme, kainate's action should be subject to modulation by

opiates in the same way they influence withdrawal. This, however, is not

the case, suggesting that kainate- and withdrawal-induced behaviors

are on parallel circuits at these levels.

However, kainate-induced behavior is sensitive to opiate antagonists.

This suggests that opiates act downstream from kainate's site of action,

in the expression of kainate-sensitive behavior. Furthermore, it has

been reported that endogenous opiates will induce these same behaviors,

thus confirming the involvement of opiates in these behaviors. The

site of this opiate-induced action appears to be a kappa opiate receptor

because it is relatively insensitive to morphine or naloxone, but quite

sensitive to ketocyclazocine (103,104). In the studies presented here,

evidence turns up which suggests a close relationship between kainate-

and endogenous opiate-induced behavior. Specifically, both occur after

intraventricular, but not intracisternal injection. Lesions of the CA3

and CA4 fields of the hippocampus prevent either from inducing WDS.

Finally, relatively high doses of naloxone are required to block kainate-

induced WDS just as relatively high doses are required to block endogenous

opiate- or ketocyclazocine-induced WDS. Thus, both anatomically and

pharmacologically, kainate and kappa opiate receptor agonists show

strikingly similar behavioral characteristics.

The evidence from these studies can be seen as suggesting two

different, but both opiate-sensitive, mechanisms for inducing WDS. One

of these mechanisms is also kainate-sensitive. This suggested scheme

appears to be consistent with more of the evidence. Some of the sub-

stances which induce WDS are sensitive to blockade of their behavioral

manifestations by naloxone. These include kainate, endogenous opiates

and ketocyclazocine. The behavioral responses to other WDS-inducing

agents are not blocked by naloxone, but are blocked by low doses of

morphine. These include TRH, theophylline, sodium valproate, AG-3-5 and

RX 33G-M. In the pharmacological studies presented here, the ability of

kainate, met-enkephalin or ketocyclazocine to induce WDS was prevented

by lesion of regio inferior. The ability of sodium valproate to induce

WDS was not affected by similar lesions. Even though this is a small

sample of the available substances, those substances sensitive to

the lesion are sensitive to naloxone antagonism, while the one agent

insensitive to the lesion is insensitive to naloxone antagonism, but

is sensitive to morphine antagonism. Such evidence indicates that the

available data may be characterized as being mediated by two pharmaco-

logically and anatomically distinct mechanisms. One mechanism, sensi-

tive to kainate, utilizes the hippocampus and kappa opiate receptors.

The other utilizes the periaqueductal region and the classical mu (mor-

phine) opiate receptor.

The relationship of the hippocampus to situations in which animals,

rats in particular, exhibit wet-dog shaking is unknown. This is largely

because the behavior has not been widely described outside of two situ-

ations--immersion in water and morphine withdrawal. That the hippocampus

is capable of inducing this behavior is strongly suggested by the present

set of studies and confirmed by MacLean (109) who reported that electric-

ally stimulated afterdischarge of the hippocampus of rats was accompanied

by wet-dog shaking.

In order to further study the role of kainate-sensitive elements

of the hippocampus in normal behavior it will be necessary to discover

situations in which rats exhibit WDS. By accident, I have found that

rats exhibit WDS when repeatedly rebuffed in their attempts to achieve

sexual intercourse. Thus, it appears that rats may exhibit WDS in more

situations than were previously studied and it is possible that examina-

tion of such situations may provide a test behavior for kainate and also

reveal more of the subtleties of hippocampal modulation of behavior.


1. D. R. Curtis, J. W. Phillis and J. C. Watkins (1960) J. Physiol.,
150, 656.

2. D. R. Curtis and J. C. Watkins (1960) J. Neurochem., 6, 117.

3. K. Krnjevic and J. W. Phillis (1963) J. Physiol., 165, 274.

4. J. M. Crawford (1970) Br. Res., 17, 287.

5. C. Yamamoto, H. Yamashita and T. Chujo (1977) Japan. J. Physiol.,
27, 225.

6. J. S. Coombs, J. C. Eccles and P. Fatt (1955) J. Physiol., 130, 326.

7. R. V on Baumgarten, R. Bloom, F. E. Oliver and G. C. Salmoiraghi
(1963) Arch. ges. Physiol., 277, 125.

8. H. McLennan, R. D. Huffman and K. C. Marshall (1968) Nature, 219,

9. W. J. Logan and S. H. Snyder (1971) Nature, 234, 297.

10. V. J. Balcar and G. A. R. Johnston (1972), J. Neurochem., 19, 2657.

11. V. J. Balcar and G. A. R. Johnston (1973) J. Neurochem., 20, 529.

12. R. W. James, G. G. Lunt, and J. F. Donnellan (1977) Biochem. Soc.
Trans., 5, 170.

13. E. K. Michaelis, M. L. Michaelis and L. L. Boyarsky (1974)
Biochem. Biophys. Acta, 367, 338.

14. E. K. Michaelis (1975) Biochem. Biophys. Res. Commun., 65, 1004.

15. P. J. Roberts (1974) Nature, 252, 399.

16. I. L. Crawford and J. D. Connor (1973) Nature, 244, 442.

17. J. Storm-Mathisen (1977) Br. Res., 120, 379.

18. P. Andersen, T. W. Blackstad and T. Lomo (1966) Exp. Br. Res., 1,

19. P. A. Schwartzkroin and P. Andersen (1975) Adv. Neurol., 12, 45.

20. H. J. Spencer (1976) Br. Res., 102, 91.

21. W. F. White, J. V. Nadler, A. Hamberger, C. W. Cotman and J. T.
Cummins (1977) Nature, 270, 357.

22. A. L. Lehninger, Biochemistry, Worth Publishers, Inc., 1970.

23. N. Lemkey-Johnston and W. A. Reynolds (1974) J. Neuropath. and
Expt'l. Neurol., 33, 74.

24. J. W. Olney (1976) Adv. Expt'l. Med. Biol., 69, 497.

25. H. McLennan (1976) Br. Res., 115, 139.

26. C. C. Mao, A. Guidotta and E. Costa (1974) Br. Res., 79, 510.

27. P. Wiechert and A. Herbst (1966) J. Neurochem., 13, 59.

28. P. Wiechert and G. Gollnitz (1968) J. Neurochem., 15, 1265.

29. H. N. Bhagavan, D. B. Coursin and C. N. Stewart (1971) Nature,
232, 275.

30. C. N. Stewart, D. B. Coursin and H. N. Bhagavan (1972) Toxicol.
Appi. Pharmacol., 23, 635.

31. M. Avoli, P.F. A. Barra, A. Brancati and M. Deodati (1976) Boll.
Soc. Ital. Biol. Sper., 52, 1531.

32. W. J. Freed and E. K. Michaelis (1976) Psychopharmacol., 50, 293.

33. J. W. Olney and 0. L. Ho (1970) Nature, 227, 609.

34. R. M. Burde, B. Schainker and J. Kayes (1971) Nature, 233, 58.

35. G. A. R. Johnston, D. R. curtis, J. Davies and R. M. McCulloch
(1974) Nature, 248, 804.

36. R. M. McCulloch, G. A. R. Johnston, C. J. A. Game and D. R. Curtis
(1974) Expt'l. Br. Res., 21, 515.

37. R. J. Walker (1976) Comp. Biochem. Physiol., 55, 61.

38. H. Shinozaki and S. Konishi (1970) Br. Res., 24, 368.

39. T. J. Biscoe, R. H. Evans, P. M. Headley, M. Martin and J. C.
Watkins (1975) Nature, 255, 166.

40. Stephanson and G. A. R. Johnston (unpublished observation), reported
in ref. 36.

41. V. K. Singh, E. G. McGeer and P. L. McGeer (1977) Soc. Neurosci.
Abs., 3, #140.

42. E. G. McGeer, P. L. McGeer and K. Singh (1978) Br. Res., 139, 381.

43. J. Lakshmanan and G. Padmanaban (1974) Biochem. Biophys. Res.
Commun., 58, 690.

44. J. W. Olney, V. Rhee and 0. L. Ho (1974) Br. Res., 77, 507.

45. P. Polc and W. Haefely (1977) N-S Arch. Pharmacol., 300, 199.

46. J. R. Simon, J. F. Contrera and M. J. Kuhar (1976) J. Neurochem.,
26, 141.

47. L. T. Graham, Jr., R. P. Shank, R. Werman and M. H. Aprison (1967)
J. Neurochem., 14, 465.

48. D. R. Curtis and R. W. Ryall (1966) Exp. Br. Res., 2, 81.

49. R. W. Ryall and M. F. Piercey (1971) J. Neurophysiol., 34, 242.

50. A. W. Duggan (1974) Exp. Br. Res., 19, 522.

51. G. B. Hutchinson, H. McLennan and H. V. Wheal (1978) Br. Res., 141,

52. J. T. Coyle and R. Schwarcz (1976) Nature, 263, 244.

53. R. Schwarcz and J. T. Coyle (1977) Life Sci., 20, 431.

54. R. Schwarcz and J. T. Coyle (1977) Br. Res., 127, 253.

55. E. G. McGeer and P. L. McGeer (1976) Nature, 263, 517.

56. T. Hattori and E. G. McGeer (1977) Br. Res., 129, 174.

57. J. W. Olney and T. DeGubareff (1978) Br. Res., 140, 340.

58. T. Hattori, E. G. McGeer and P. L. McGeer (1977) Soc. Neurosci.
Abs., 3, #112.

59. R. Schwarcz, J. P. Bennett, Jr. and J. T. Coyle (1977) J.
Neurochem., 28, 867.

60. I. Divac, F. Fonnum and J. Storm-Mathisen (1977) Nature, 266, 377.

61. P. L. McGeer, E. G. McGeer, U. Scherer and K. Singh (1977)
Br. Res., 128, 369.

62. J-S. Kim, R. Hassler, P. Haug and K-S. Paik (1977). Br. Res.,
132, 370.

63. V. K. Singh, E. G. McGeer and P. L. McGeer (1977) Soc. Neurosci.
Abs., 3, #140.

64. J. S. Hong, H-Y. Yang and E. Costa (1977) Neuropharmacol., 16, 451.

65. P. R. Sanberg, J. Lehmann and H. C. Fibiger (1977) Soc. Neurosci.
Abs., 3, #138.

66. P. L. McGeer and E. G. McGeer (1977) Soc. Neurosci. Abs., 3, #126.

67. G. DiChara, M. Olianas, l1. DelFiacco, p. F. Spano and A.
Tagliamonte (1977) Nature, 268, 743.

68. R. M. Herndon and J. T. Coyle (1977) Science, 198, 71.

69. J. V. Nadler, B. W. Perry and C. W. Cotman (1978) Nature, 271, 676.

70. T. DeGubareff and J. W. Olney (1977) Soc. Neurosci. Abs., 3, #1644.

71. L. J. Pelligreno and A. J. Cushman (1967) A Stereotaxic Atlas
of the Rat Brain, Appleton-Century-Crofts, New York, N. Y.

72. J. Donaldson, T. St-Pierre, J. Minnich and A. Barbeau (1971)
Can. J. Biochem., 49, 1217.

73. J. Blasig, A. Herz, K. Reinhold and S. Zieglgansberger (1973)
Psychopharmacol., 33, 19.

74. H. Kruse (1977) Arch. of Pharmacol., 297, 561.

75. E. Wei, S. Siegel, H. H. Loh and E. L. Way (1975) Nature, 253,

76. P. Bedard and C. J. Pycock (1977) Neuropharmacol., 16, 663.

77. F. Bloom, D. Segal, N. Ling and R. Guillemin (1976) Science,
194, 630.

78. L. Leybin, C. Pinsky, F. S. LaBella, V. Havlicek and M. Rezek
(1976) Nature, 264, 458.

79. H. Frenk, G. Urca and J. C. Liebeskind (1978) Br. Res., 147,

80. E. T. Wei (1976) J. Pharm. Pharmacol., 28, 722.

81. T. deBoer, H. J. Metselaar and J. Bruinvels (1977) Life Sci.,
20, 933.

82. U. Jahn and G. Mixich (1976) Psychopharmacol., 46, 191.

83. N. Rezek, V. Havlicek, L. Leybin, F. S. LaBella and H. Friesen
(1978) Can. J. Physiol. Pharmacol., 56, 227.


84. A. Cowan and T. Watson (1978) Psychopharmacol., 57, 43.

85. H. 0. J. Collier, D. L. Francis, G. Henderson and C. Schneider
(1974) Nature, 249, 471.

86. S. Haldeman and H. McLennan (1972) Br. Res., 45, 393.

87. T. W. Stone (1973) J. Physiol., 233, 211.

88. H. J. Spencer, V. K. Gribkoff, C. W. Cotman and G. S. Lynch
(1976) Br. Res., 105, 471.

89. M. Segal (1976) Brit. J. Pharmacol., 58, 341.

90. G. Clarke and D. W. Straughan (1977) Neuropharmacol., 16, 391.

91. N. H. Bass and P. Lundborg (1973) Br. Res., 52, 323.

92. S. M. Wuerthele, K. L. Lovell, M. Z. Jones and K. E. Moore (1978)
Br. Res., 149, 489.

93. R. Lorente de Nd (1934) J. fur Psych. Neurol., 46, 113.

94. T. W. Blackstad (1956) J. Comp. Neurol., 105, 417.

95. R. L. Isaacson (1974) The Limbic System, Plenum Press, New York,
N. Y.

96. R. B. Chronister and L. E. White, Jr. (1975) in The Hippocampus,
Vol. 1, R. L. Isaacson and K. H. Pribram, eds., Plenum Press,
New York, N. Y.
97. M. Segal and F. E. Bloom (1974) Br. Res., 72, 99.

98. R. Y. Moore and A. E. Halaris (1975) J. Comp. Neurol., 164, 171.

99. P. Andersen (1975) in The Hippocampus., Vol. 1, R. L. Isaacson
and K. H. Pribram, eds., Plenum Press, New York, N. Y.

100. B. K. Koe and A. Weissman (1966) J. Pharmacol. Exp. Ther., 154,

101. G. Urca, H. Frenk, J. C. Liebeskind and A. N. Taylor (1977)
Science, 197, 83.

102. E. T. Wei, L. F. Tseng, H. H. Loh and C.H. Li (1977) Life
Sci., 21, 321.

103. W. R. Martin, C. G. Eades, J. A. Thompson, R. E. Huppler and
P. E. Gilbert (1976) J. Pharmacol. Exp. Ther., 197, 517.

104. M. Hutchinson, H. W. Kosterlitz, F. M. Lselie, A. A. Waterfield
and L. Terenius (1975) Brit. J. Pharmacol., 55, 541.


105. E. T. Wei, H. H. Loh and E. L. Way (1973) Life Sci., 12, 489.

106. E. T. Wei (1973) Life Sci., 12, 35.

107. E. Laschka, Hj. Teschemacher, P. Mehraein and A. Herz (1976)
Psychopharmacol., 46, 141.

108. W. Wei, H. H. Loh and E. L. Way (1973) J. Pharmacol. Exp. Ther.,
185, 108.

109. P. D. MacLean (1957) AMA Arch. Neurol. Psychiat., 78, 128.


Thomas Lanthorn was born in St.Louis, Missouri, on November 4, 1950.

He received his elementary and secondary education in Grand Rapids,

Michigan,and was graduated from Ottawa Hills High School in June,

1968. He attended Lyman Briggs College at Michigan State University

and received the degree of Bachelor of Science in June, 1972. He

began work as a graduate student at the University of Florida in the

summer of 1973 under the direction of Professor Robert L. Isaacson.

He is presently a candidate for the degree of Doctor of Philosophy

at the University of Florida.

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Robert L. Isaacson, Chairman
Professor of Psychology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Carol Van Hartesveldt
Associate Professor of Psychology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Marc N. Branch
Associate Professor of Psychology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Adrian J. Dunn
Associate Professor of Neuroscience

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Luther J. Willmore
Affiliate Assistant Professor of

This dissertation was submitted to the Graduate Faculty of the Depart-
ment of Psychology in the College of Liberal Arts and Sciences and to
the Graduate Council, and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.

December, 1978

Dean, Graduate School

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs