Behavioral effects of hippocampal lesions after adrenergic depletion of the septal area


Material Information

Behavioral effects of hippocampal lesions after adrenergic depletion of the septal area
Physical Description:
x, 143 leaves : ill. ; 29 cm.
Baisden, Ronald Hubert, 1947-
Publication Date:


Subjects / Keywords:
Hippocampus -- drug effects   ( mesh )
Nervous System -- drug effects   ( mesh )
Neuroscience Thesis Ph.D   ( mesh )
Dissertations, Academic -- neuroscience -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1973.
Bibliography: leaves 98-113.
Statement of Responsibility:
by Robert Hubert Baisden.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000887078
oclc - 24624196
notis - AEJ5295
System ID:

This item is only available as the following downloads:

Full Text






This dissertation is dedicated with much love to my

parents, Mr. and Mrs. Russel H. Baisden, who provided me with

the inspiration, to my children, Bert and Tiffany, who

provided me with the motivation, and to my wife, Marilyn,

who stood beside me throughout the trying period of experi-

mentation and preparation of this work. I shall always be

thankful for their belief and faith in me.


The author takes this opportunity to thank

Dr. Robert L. Isaacson for his direction, support and

assistance during the preparation of this dissertation.

Dr. E. M. Johnson and Dr. J. J. Bernstein for

their help and guidance during the early part of his studies.

Dr. Carol Van Hartesveldt for her support and

editorial assistance in the writing of this manuscript.

To the members of his supervisory committee not

mentioned above (Dr. F. A. King, Dr. L. Larkin, D. C. J.

Vierck, and Dr. S. Zornetzer) for their help and moral sup-


Dr. Michael L. Woodruff, who has been his closest

associate and friend during his studies, for his invaluable

assistance both in preparing this manuscript and in direct-

ing his thinking.

John Gonzalez and Leah Gilbert for their time

and help in the performance of these experiments.

Mrs. Pauletta Sanders for her technological help

in the preparation of the histological materials and for her

constant moral support.

Mrs. Virginia Walker for her understanding and

patience, and her aid in the preparation and typing of this



To the students and associates of the Departments

of Psychology, Anatomy, and Neurosciences with whom he has

been fortunate enough to come into contact during the course

of his studies and who have provided him with both friendship

and assistance which will always be remembered and revered.

In particular: Ms. Barbara Schneiderman, Mr. Richard Kearley,

Dr. Douglas Poorman, Dr. John B. Gelderd, Mrs. Sandra Reynolds,

and Mrs. Barbara McGuire.

The author would like to take this opportunity to

give a note of special appreciation to Mr. Willie J. Sanders

for his unending support and assistance during those periods

of his graduate career when his spirits were lowest and his

future less than certain. Will has been a true friend and

an excellent teacher.






ABSTRACT . . . ix


METHODS. . . .. 44

RESULTS... Experiment I. . .. 53

RESULTS... Experiment II .. . 69


REFERENCES . . .. 98

APPENDIX A . . 114

APPENDIX B . . 132




1 Experiment I Number of subjects showing
increased (i); decreased (d); or no change
(x) in DRL-20 Performance. Experimental
(6-OH-dopamine Treated Group .

2 Experiment I Number of subjects showing
increased (i); decreased (d); or no change
(x) in DRL-20 Performance. Vehicle
Control Group . .

3 Experiment I Passive Avoidance Latency
Scores . . .

4 Experiment II Passive Avoidance Latency
Scores Before Hippocampal Lesions .

5 Experiment II Passive Avoidance Latency
Scores Following Hippocampal Lesions .


. .58

. .

A-I Statistical significance and direction
of change Number of Responses. .

A-2 Statistical significance and direction
of change Number of Correct Responses

A-3 Statistical significance and direction
of change Percent Correct Responses. .

A-4 Statistical significance and direction
of change experimental subjects
Number of Responses. . .

A-5 Statistical significance and direction
of change experimental subjects
Number of Correct Responses .

A-6 Statistical significance and direction
of change experimental subjects
Percent Correct Responses. .

A-7 Experiment I Water consumption .

* 115

* 116

* 117

* 118

* 119

* 120

. 121


* .

LIST OF TABLES (continued)



B-I Experiment II Open field activity -
subject mean (X), standard deviation (SD),
and standard error of the mean (SE) for
each treatment . . .

B-2 Experiment II Mean open field activity
distribution preoperativee scores) 7 days
X = one subject's mean score in the designated
range. . . .

B-3 Experiemnt II Mean open field activity
distribtuion prehippocampal septal treat-
ment and hippocampal and cortical control
lesions 25 days . .




B-4 Mean open field activity distribution -
following hippocampal lesion and controls -
10 days. . . .

B-5 Experiment II Statistical significance
and direction of change open field
activity preop. to septal treatment. .

B-6 Experiment II Statistical significance
and direction of change open field
activity septal treatment to hippocampal
lesion comparison and controls .

B-7 Experiment II Water consumption
Prehippocampal and Posthippocampal .

B-8 Experiemnt II Water consumption. .

. 138

. 139

. 140






1. Section through the Septal Area at the level
of the Anterior Commissure . 28

2. Diagrammatic representation of the differ-
ential origin and termination of Hippocampal-
Septal connections . 32

3. Diagrammatic representation of the differ-
ential origin and termination of Septo-
Hippocampal connections . 34

4. Photomicrograph of representative sections
through the Septal Area in subjects of
Experiment I . . 54

5. Reconstructions of maximal (lined) and minimal
(solid) extent of Hippocampal ablation in sub-
jects of Experiment I. . .. 57

6. DRL-20 Group Mean IRT Distribution 62

7. Photomicrograph of representative sections
through the Septal Area in subjects of
Experiment II receiving intraseptal injections 70

8. Photomicrograph of representative sections
through the Septal Area in subjects of
Experiment II receiving Septal electro-
lytic lesions . . 7

9. Reconstructions of maximal (lined) and minimal
(solid) extent of Hippocampal ablation in
subjects of Experiment II . 75

10. Reconstructions of maximal (lined) and minimal
(solid) extent of Neocortical ablation in
subjects of Experiment II . 77

Cl through C6. DRL-20 performance in vehicle control
subjects of Experiment I. 123

El through E6. DRL-20 performance in drug treated
subjects of Experiment I. 127

C7 and E7. DRL-20 performance in subjects not
surviving all operative procedures. 131

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



Ronald Hubert Baisden


Chairman: Robert L. Isaacson
Major Department: Neuroscience

The effects of bilateral injections of 6-OH-dopamine-

HBr into the septal area of adult rats prior to hippocampal

ablation were studied using several behavioral measures

known to be sensitive to hippocampal and/or septal damage.

Hippocampal lesions performed 12-14 days after injection of

the drug did not produce the usual increase in rate of re-

sponse on a DRL-20 operant schedule, although the number of

reinforcements obtained did decrease. When tested 15-20 days

after the ablation, these subjects did not show the usual

lesion-produced deficit in passive avoidance responding. In

a second experiment, hippocampal lesions were performed 30-50

days after the intraseptal injections of 6-OH-dopamine-HBr.

In these subjects, the usual deficit in passive avoidance re-

sponding was observed in the drug treated group when tested

15-20 days following hippocampal lesion. In this group,

treatment of the septal area with 6-OH-dopamine did not affect

water consumption, nor prevent hippocampal lesion-induced

increases in open field activity.

Injection of the septum with 6-OH-dopamine alone did

not induce changes in water consumption, open field activity,

passive avoidance responding, or DRL performance levels. DRL

behavior was affected, however, in that the performance of

the drug treated group (as opposed to the group receiving

vehicle control injection) did not improve their performance

over the period of testing.

These results are interpreted to indicate that the re-

organization of the adrenergic fibers of the septum can play

a role in producing some of the behavioral changes seen after

hippocampal lesions. The behavioral responses that are in-

fluenced by these changes in the septum appear to be partially

mediated via the hippocampal-septal system, and modulated by

the adrenergic innervation of the septal area.


This study is concerned with the behavioral manifesta-

tions of neural reorganization subsequent to brain damage.

In particular, these experiments are designed to examine

the effects of changes in the adrenergic component of the

septum following hippocampal lesions. In this section an

attempt is made to review the necessary information needed

to consider this particular problem and interpret the re-

sults obtained. The discussion consists of three major


1. Plastic phenomena within the central nervous sys-

tem brain lesion-induced effects.

2. Septo-hippocampal system.

3. Morphology of the noradrenergic innervation of the

septum and its possible underlying behavioral mechanisms.

Much of our current knowledge of brain-behavior rela-

tionships has come from studies involving brain-lesioned

animals. In this type of study, indications of structure-

function relationships are derived from measured differences

in behaviors both before and after damage to a specific brain

locus. This correlative approach assumes that the structure

sustaining the lesion plays a role in mediating the behaviors

being considered and that the specific changes observed are

the result of the destruction of the structure itself.

Although these assumptions are valid under many circum-

stances, recent work indicates that many behavioral changes

resulting from damage to the central nervous system may be a

consequence of the disruption of complex neural networks

rather than due to destruction of a specific anatomical struc-

ture (Rosenzweig and Leiman, 1968).

It has long been known that when nervous system elements

are damaged some of the remaining structures undergo compen-

satory changes in both their morphological and physiological

characteristics. Von Monakow (1914) noted that a local

lesion can have effects at points where fibers coming from

the injured area enter into the grey substance of primarily

intact nervous tissue. He referred to this phenomenon as

"diaschisis." In addition, Von Monakow states that "Any in-

jury suffered by the brain substance will lead to a

struggle for the preservation of the disrupted nervous func-

tion, and the central nervous system is always prepared

for such a struggle." It may be the case that the changes

that take place in the elements associated with a damaged

structure may contribute to the observed behavioral effects

of the lesion, especially if the changes take place in associ-

ated structures which are components of the neural network

mediating some aspect of the behavior being measured. The

behavioral alterations resulting from these secondary changes

could conceivably result in an augmentation of the effects of

the lesion itself or an attenuation of these effects. In the

latter case these changes may be responsible for inducing re-

covery phenomena in some instances.

The following discussion will consider some of the phe-

nomena occurring after brain lesions that may have a signifi-

cant effect on nervous system functions. The review will be

focused on the effects of neuronal denervation, regeneration

of severed axons and collateral sprouting of intact fibers in

response to local damage. Each topical discussion will be con-

cerned with a description of the phenomenon, a consideration of

the mechanism involved and some possible effects on subsequent

neural function. Each topic is also discussed in reference to

how the catecholamine fiber systems may demonstrate or other-

wise affect the manifestation of each of these neural proper-

ties. This is for two reasons: first, the catecholanine sys-

tem provides a good, if not the best, example of the topic be-

ing considered, and second the noradrenergic component of the

catecholamine system and its propensity for undergoing or in-

ducing plastic changes is an intergral part of the experiment.

Deafferentation of Central Nerve Cell Bodies

When the input to a neuron of the CNS is removed, a modi-

fication of various morphological and physiological properties

of that neuron is induced. It is thought that this may result

from the loss of trophicc" influences of the lost inner-

vation (Guth and Windle, 1970). These trophic interactions

are necessary for the normal growth and maintenance of the

structures innervated (Guth, 1968; Singer, 1952). From stud-

ies of peripheral trophic effects in the neuromuscular and

sensory systems, we know that all nerve fibers are capable of

providing a trophic influence on the structures upon which

they terminate. Furthermore, a minimal number of fibers are

necessary to exert this effect on the innervated structure

(Singer, 1952; Guth, 1968; Guth and Windle, 1970). Therefore

it is not unreasonable to postulate that a decrease in trophic

interactions due to partial deafferentation of central elements

could lead to observable morphological and physiological


Morphological Changes

Electron micrographs of denervated cells and their imme-

diate surroundings have shown that synapses of severed axons

undergo a variety of degenerative changes (for review of this

topic see Raisman and Matthews, 1972). Kjerulf et al. (1973)

report a sequence of changes in degenerating synapses of the

lateral cuneate nucleus which can be correlated with changes

in certain physiological properties described below. These

degenerating synapses can be seen to be displaced or engulfed

by invading glial elements. In some instances both pre- and

post-synaptic membranes are attacked by the glial processes.

(Collonier, 1964; Blinzinger and Kreutzberg, 1967; Wisniew-

ski et al., 1972).

Loss of innervation to a neuron can result in structural

changes at the cell body and along the shafts and branches of

the dendrites. In the cell body these changes take the form

of a generalized chromatolytic reaction. This includes an

increase in Nissl substance, a breakdown in the golgi apparatus,

hypertrophy of neurofibrils and eccentric displacement of the

nucleus (Cajal, 1928). These changes are very similar to the

alterations seen in cell bodies following axotomy (Eccles

et al., 1958).


Dendritic changes manifest themselves in a variety of

ways. A decrease in branching (Westrumetal-, 1965; Bern-

stein and Bernstein, 1972; 1973) results in a decrease in to-

taldendritic field (Jones and Thomas, 1962). The number of

spines decrease (Rutledge et al., 1972) and in many instances

the formation of beaded nodular varicosities occurs all

along the shaft (Scheibel and Scheibel, 1968; Bernstein and

Bernstein, 1972; 1973). Bernstein and Bernstein (1972) postu-

late that these dendrites have undergone a form of dediffer-

entiation to an early, more receptive developmental state.

This is evidenced by the fact that new synapse formation by

ingrowing sprouts takes place at the site of the dendritic


Physiological Changes

Denervation of excitable tissue results in changes in

electrical and chemical properties that can manifest them-

selves in at least three ways. The deafferented neurons

become spontaneously hyperexcitable, supersensitive to elec-

trical stimulation and/or supersensitive to chemical stimuli.

Loeser and Ward (1967) demonstrated that the spontaneous

activity of spinal cord neurons was increased to abnormally

high levels in regions denervated by dorsal rhizotomy or

hemisection. Likewise, cells of the lateral cuneate nucleus

display this hyperexcitability after deafferentation by cervi-

cal rhizotomy (Kjerulf and Loeser, 1973). These investigators

demonstrated a sequential development of this abnormal elec-

trical activity after the lesion. Immediately after surgery,

the cells were observed to undergo a period of relative si-

lence. Within ten days, however, the abnormal firing rates

are firmly established. Correlative electron microscopical

investigation showed that the initial silent period is associ-

ated with depletion of synaptic vesicles in the terminals of

the primary dorsal root afferents. The hyperactive phase

corresponded anatomically to the presence of vacated synaptic

sites and a reduction in dendritic spine density (Kjerulf

et al., 1973). Similar hyperactive responses have been seen

following denervation of CNS cells in the spinal trigeminal

nucleus (Ward, 1969; Westrum and Black, 1971) and in cerebral

and cerebellar cortex (Sharpless, 1969).

Both abnormal responses to electrical stimulation and

spontaneous hyperexcitability were seen in the cord prepara-

tions of Loeser and Ward (1967). A single stimulus elicited

prolonged bursts of activity (up to 100 msec ) in these units.

rTiSe phenomena had been previously described by Teasdale and

Stravraky (1953) in spinal preparations. They found that

chronic hind limb deafferentation resulted in a reduced thresh-

old and an increase in amplitude and response duration follow-

ing pyramidal tract stimulation. Eccles et al. (1962) and

Kostyuk (1963) demonstrated an increased monosynaptic reflex

following cord denervation. In addition, these investigators

could demonstrate no change in membrane properties of motor-

neurons (e.g., resting membrane potential, conductance and

resistance) as would be expected to occur in degenerating

membranes. These factors and the observations made on hyper-

active cells (above) all tend to indicate that these changes

are due to a loss of incoming stimulation rather than to the

chemical or morphological changes taking place in the cell.

Further evidence for this contention was provided by Rutledge

et al. (1967) who showed that low level, daily, subthreshold

stimulation of isolated cortical slabs prevented the develop-

ment of this abnormal electrical activity.

Supersensitivity to pharmacological stimuli after deaf-

ferentation has been shown to occur in CNS neurons (Drake and

Stravraky, 1948; Sharpless, 1964). Cerebral cortical tissue

is known to be sensitive to topically applied acetylcholine

in a rostral-caudal gradient with occipital cortex being the

most sensitive. Chronic isolation by undercutting renders

all areas more sensitive to this treatment, the least sensi-

tive area reaching values equivalent to normal occipital cor-

tex (Echelin and Battista, 1962; 1963). As with electrical

supersensitivity this could be prevented with daily low in-

tensity stimulation (Rutledge et al., 1967). This stimulation

also prevented the normal decrease in acetylcholinesterase

observed in these preparations. These observations further

illustrate the importance of maintaining adequate afferent

stimuli to cells and that loss of input may result in a vari-

ety of far-reaching effects.

An example of how abnormal electrical activity induced by

deafferentation can have a significant influence on the ner-

vous system is provided by Ward's (1969) model of the deaffer-

entated epileptic nerve cell. Neurons of the focal area show

the same types of morphological changes and electrical prop-

erties of denervated neurons described above (Sharpless,

1969). During a grand mal seizure abnormal activity derived

from this small group of cells propagates over and influences

the electrical activity of the entire nervous system.

The recent development of pharmacological agents which

either specifically destroy catecholamine nerve terminals

'e.g., 6-OH-dopamine (Ungerstadt, 1968) or selectively inhib-

it the synthesis of these transmitter substances (e.g.,

alpha-methyl-para-tyrosine (Spector et al., 1965) have al-

lowed for the investigation of deafferentation effects within

brain regions innervated by catecholamine containing fibers.

As yet there have been no direct demonstrations of super-

sensitivity after specific lesions made by local application

of these agents. However, there does exist a moderate amount

of correlative behavioral evidence to indicate that some

kinds of excitability changes do occur and can have signifi-

cant behavioral effects.
Preliminary observations of a possible catecholamine

supersensitivity effect were made by Anden et al. (1966) who

observed that animals with unilateral lesions of the corpus

striatum or of the dopaminergic nigrostriatal pathway turned

toward the side of the lesion, that is toward the side of

less catecholamine transmitter. When these animals were giv-

en L-Dopa the animals rotated in the contralateral (opposite)

direction. This indicates that the drug was more active on

the lesioned side, i.e., the cells on the lesioned side were

more sensitive to the pharmacological agents. This was

replicated by Ungerstadt (1971b) who showed the contralateral

rotational response in animals treated with L-Dopa or apo-

morp)hine after unilateral 6-Oii-dopar;ine lesions of the

nigra-striatal system.

The hypothalamus receives a substantial amount of cate-

cholamine input from the mesencephalon via the medial fore-

brain bundle (MFB). This complex fiber bundle originates in

the brain stem, traverses the base of the brain and passes

bilaterally through the lateral hypothalamus. Lesions in

this hypothalamic area produce adipsia and aphagia. Both of

these abnormalities recover over a period of time through a

series of well-defined stages (Teitelbaum and Epstein, 1962).

The anorexic component can be induced by electrolytic or

pharmacologic lesion to the midbrain neurons of the adrener-

gic component of the MFB (Ahlskog and Hoebel, 1973) and has

been shown to be reversible by intraventricular administra-

tions of exogenous norepinephrine (Berger et al., 1971).

These factors have been interpreted as indicating that the

adrenergic terminals in the lateral hypothalamus may play

a role in mediating some types of feeding behavior. It is

suggested that normal recovery after MFB lesions is due to

either the internal resetting of optimal weight levels or to

the gradual development of supersensitivity in order to respond

to the reduced levels of norepinephrine induced by the lesion.

Evidence for the latter contention was presented by Glick

et al. (1972) who showed that treating rats with alpha-

methyl-para-tyrosine three days prior to inflicting lateral

hypothalamic damage allowed the animal to be able to eat,

drink, and gain weight immediately after surgery. In some

cases some of the subjects did show a mild form of the syn-

drome which underwent a very rapid recovery. It has been

suggested that this pretreatment may induce the development

of supersensitivity which is adaptive. This factor would al-

low for the reduction of the recovery period seen after this

type of brain damage. This explanation was corroborated by

Balagura et al. (1973) who demonstrated that pretreatment

with 2 hormones regulating food intake, one of which (Insulin)

induced alterations in norepinephrine synthesis, had differ-

ential effects on recovery after this same lesion. In the

insulin treated subjects a reduced recovery period was ob-

served in comparison to saline treated control. Treatment

with the other hormone (Glucagon) greatly augmented the re-

covery period to levels much longer than is seen in the con-

trol group.

Stein et al. (1972) have shown that recovery of self-

stimulation in the lateral hypothalamic portion of the median

forebrain bundle after 6-OH-dopamine lesions can similarly

indicate an increase in receptor sensitivity. After the

lesions, animals increase their self-stimulation rates to

values much larger than normal when given injections of nor-

epinephrine. The development of supersensitivity in the cells

deprived of their adrenergic input by this lesioning procedure

would explain this observation. These results have been

corroborated by Mandell et al. (1972) who demonstrated that

subjects with lowered exploratory activity levels induced by

depletion of norepinephrine by intraventricular injection of

6-OH-dopamine showed an abnormal increase on this measure when

treated subsequently with norepinephrine.

These studies suggest that neurons, after being deprived

of a portion of their input, tend to become more excitable.

Although these excitability changes can be shown to have sig-

nificant effects, many questions must still be answered be-

fore the causes of these changes are understood. Could the

loss of trophic factors induce a cellular dedifferentiationn"

of the type proposed by Bernstein and Bernstein (1972; 1973),

and account for these effects? This is unlikely, since most

immature nerve cells do not appear to show these abnormal

electrical properties under normal conditions (see Purpura,

1969). Are these changes brought about through independent

mechanisms or are they interrelated effects? The latter

possibility seems most likely at present since the causative

factor in all cases denervationn) is similar. Or finally, are

these induced changes indicative of a generalized homeostatic

response of neurons which is manifested under abnormal condi-

tions as is proposed for the collateral sprouting response?

(see belcw)


It has long been known that many of the lower vertebrates

retain the capacity to regenerate and reconstitute long fiber

tracts throughout their life-span. Mammals, birds and a few

of the lower species (e.g., anurans) apparently loose this

ability after the completion of embryonic development. Al-

though a few early reports provided some evidence for struc-

tural and functional restitution after CNS lesions in the

higher phyla (Sugar and Gerard, 1940; Hamberger, 1955), most

investigators thought that these forms were unable to re-

organize themselves to form any type of functional contact

over even short distances. These early observers reported

only abortive or abnormal growth processes around the lesion

site (Cajal, 1928; Clark, 1942; Windle and Chambers, 1950;

Windle et al., 1952; Windle, 1955). More recent investigations

have revealed that these processes are newly formed sprouts

from local severed and intact axons (Rose et al., 1960;

Estable-Puig et al., 1964; Estable-Puig and de Estable-Puig,

1972; Bernstein, 1964; Bernstein and Bernstein, 1967; 1968;

1972; 1973).

The formation of nerve sprouts by severed or intact axons

was extensively described by Cajal (1928). Using silver

techniques, Cajal observed the outgrowth of sprouts and de-

scribed the process as a random ameboid outgrowth of "growth

cones." Subsequent studies indicated that the progress of

the growing tip is determined by the direction and number of

sprouts, reactions within the cell, and resistance of the en-

vironment to growth of the tip of the axon (Weiss, 1955).

This signifies that regenerative phenomena are dependent upon

many factors, including the nature of the lesion (e.g., type,

location, and extent) and the nature of the outgrowing cell

(e.g., age and size). These factors pose limitations on the

potential for regeneration of any given system. Only recently

has there been any indication that the factors promoting re-

generation of CNS structures can exist in the higher vertebrate


Terminal Axon Regeneration

Many aspects of the regeneration of severed axons are

presently under extensive investigation. Reviews exist to

which the reader is referred for information regarding the

present status of these various approaches (e.g., Schneider,

1972; Guth and Windle, 1970; 1973; Windle, 1955; 1956;

Raisman and Matthews, 1972; Bernstein and Bernstein, in press;

Clemente, 1964). This discussion will be concerned only with

those mammalian CNS systems in which regeneration has been

shown to occur.

Central fiber tracts composed of monoamine containing

axons have been shown to possess a high potential for termi-

nal regeneration of cut axons. This process has been described

as taking place in a series of well-defined stages. When an

axon of a central monoamine pathway is severed, accumulations

of transmitter substances extend 1 to 2 mm up the proximal

stump. Initially these accumulations are intensely fluor-

escent but gradually decrease in intensity over the post-

operative period. Within two to three weeks only a few diffuse

accumulations remain (Dahlstrom and Fuxe, 1965; Ungerstadt,

1971a). Within a week following electrolytic lesions which

severed the catecholamine fiber tracts in the spinal cord

Bjorklund et al., 1971) and brain stem (Katzman et al., 1970)

the intensely fluorescent accumulations of neurotransmitters

were accompanied by many fine varicose fibers which seemed to

be concentrated around the proximal stumps. Over the next

few weeks these small fibers increased to form densely packed

bundles which extended along the border of the lesions. Some

of the fibers were observed to enter into the lesion itself.

These structures were still present and numerous seven weeks

post operatively. This was taken to indicate that they were

a permanent component in a reorganized system.

Although the new growth processes appeared to be derived

from the proximal stump of the sevemd catecholamine axons,

the fact that peripheral elements are known to have the ability

to grow into central tissue (see Glees, 1955)led to the specu-

lation that these fluorescent sprouts could be derived from

the autonomic elements of local circulatory structures. A

cervical sympathectomy was performed and shown to have no

effect on the development, growth or maintenance of the

fluorescing fibers. This implied that these centrally lo-

cated fibers were formed from sprouts originating within the

central nervous system, presumably from the cut axons them-


Regenerated fibers of both the brainstem and cord pre-

parations (above),however, were confined to the locus of the

lesion and (as is traditionally accepted) demonstrated no

tendency to grow along the pathway of the severed axons. Thus,

since no proof for the formation of any type of functional

connection existed, this phenomenon appeared similar to those

abortive regeneration attempts described earlier. By using

homographs transplanted into the area of the regenerating

fibers, some proof of functional regeneration was obtained.

Bjorklund and Stenevi (1971) and Bjorklund et al. (1971) trans-

planted various peripheral tissues into severed catecholamine

pathways of the CNS. Tissues which normally received graded


amounts of catecholamine (norepinephrine) innervation were

used. In all cases the transplanted tissue appeared healthy

and the normal pattern of innervation and catecholamine con-

tent was reinstated. In transplants of striated muscle,

which do not normally receive noradrenergic fibers, there

was no evidence of fluorescent fiber growth into the trans-

planted tissue. It thus appears that there exists in these

neurons a capacity to regenerate over short distances and to

form functional connections provided an appropriate structure

is present.

This formation of new connections appears to be deter-

mined by the characteristics of both the regenerating fiber

and the structures to be innervated. This may imply that some

form of specificity is operating in the regenerating system.

Several other observations made on these preparations provide

further evidence supporting this assumption:

1. The initial outgrowth of sprouts appears to be ran-

dom and multi-directional. With time, however, these fibers

coalesce into faciscles, a portion of which consistently inner-

vate local blood vessels even in sympathectomized animals

(Bjorklund et al., 1971).

2. When the fibers which reinnervate the transplants

were analyzed as to monoamine fiber types, it was shown that

only norepinephrine fibers completely invaded the peripheral

tissues. Very few dopamine fibers were seen to be within the

transplant and serotonin (5-Ht) endings were only demonstrable

on the external borders. The tissue seemed to display a pre-

ferential affinity for the former fiber (transmitter) type


(Bjorklund and Stenevi, 1971; Bjorklund et al., 1971).

Another system of the mammalian brain has been found to

possess a similar capacity for terminal regeneration of cut

nerve fibers. After sectioning the infundibular stalk of the

ferret, Adams et al. (1968; 1971) observed regeneration of

neuro-secretory fibers back into the neuro-hypophysis, Within

a year, complete reinnervation and functional restitution

could be demonstrated. These observations were replicated

and extended by John Kiernan at the University of Cambridge

(cited in Schneider, 1972). Following transaction of the

stalk, the neuro-secretory fibers originating in the supra-

optic and paraventricular nuclei of the hypothalamus re-

generated across a glial scar and reinnervated the neuro-

hypophysis. Kiernan records slight modifications in the size

and position of the newly innervated structure. In subsequent

studies he observed that following hypophysectomy or removal

of only the neuro-hypophysis, complete regeneration of a

slightly modified, but totally functional pituitary occurred.

The new neuro-secretory fibers were seen to selectively inner-

vate specific elements (Pituicytes) as in the normal condition.

Kiernan further showed that when the vagus nerve was

sectioned and the proximal stump inserted into the neuro-

hypophysis, regeneration of the vagus nerve was observed.

The regenerating fibers however, did not enter into the

pituitary or reinnervate deafferented pituicytes. In a

similar manner, transplants of neuro-hypophysis to the cor-

tex did not result in innervated glandular tissue. Only the

neuro-secretory fibers were found to be capable of reinner-


vating these cells. No neuro-secretory fibers were seen to

penetrate the implanted vagus nerve. The regenerating fi-

bers displayed what appeared to be an affinity for those

structures they were connected with prior to being severed.

Likewise the pituicytes seemed unable to either accept or

"call" other types of axons.

It may be that some, perhaps specialized, mammalian

central axons can regenerate to some degree. When it occurs,

there appears to be at least a potential for functional res-

titution. This is aided by what appears to be specificity

factors inherent in both the regenerating and the denervated

elements. These same tendencies can be seen under some cir-

cumstances when intact axons sprout to fill vacant synaptic


Collateral Sprouting

Collateral sprouting of intact axons is a process where-

by an axon, in response to some stimulus in the area, gives

rise to a new terminal in order to fill a vacated postsynap-

tic site. Although this phenomenon has long been known to

occur in peripheral nerves, both in cutaneous sensory and

neuro-muscular systems (see review by Edds, 1953), early in-

vestigators of regenerative phenomena in central nervous sys-

tems failed to recognize its potential significance within

the brain. The first conclusive evidence of an organized

collateral response by central axons was presented by Liu and

Chambers (1958). These investigators unilaterally destroyed

several dorsal roots or lesioned the cortico-spinal tract and

observed that intact dorsal root fibers had sprouted to

fill the areas denervated by the lesion. Subsequent investi-

gators extended these observations to other regions of the

brain and have attempted to discover the laws and mechanisms

which govern this process.

The first theoretical explanation of this phenomenon

was presented by Edds (1953), who offered a competitive sprout-

ing hypothesis to account for both the histological and physio-

logical results obtained after peripheral deafferentation of

muscles. This theory states that after deafferentation, some

stimulus occurs, presumably a chemical given off by the dener-

vated cell or site, which induces an initial overabundance

of sprout growth in local intact axons. Of these, only a few

will make contact and maintain structural and, presumably,

functional integrity. The remaining nonconnected sprouts

will be subsequently resorbed. This hypothesis may account

for some of the observations made on sprouting of central

elements. For example, after hemi-section of a mammalian

spinal cord, the number of synapses rostral to the lesion site

are first observed to decrease due to orthograde degeneration

of the descending long tract fibers. This is accompanied by a

retraction or deafferentation of dendritic spines on

those cells which have lost these contacts. Subsequent exami-

nations of these preparations at different postoperative periods

show a progressive increase of synaptic boutons and rediffer-

entiation of the dendritic processes. During this time

interval, many axonal sprouts from local fibers can be

seen growing into the area forming fascicles. These newly

formed collaterals make contact with the denervated cells.

After the number of synapses stabilize at a maximal level,

no new axonal fascicles can be observed. Those that are un-

able to make contact gradually disappear (Bernstein and Bern-

stein, 1972; 1973).

Evidence that the new boutons are of local origin is

obtained when the cord is transected again above the original

site of the lesion. After this procedure the number of bou-

tons does not decrease as with the first lesion, indicating

that the new endings are not derived from the original long

tract fibers but that the deafferented postsynaptic sites

have accepted synaptic terminals from cells which do not nor-

mally innervate these regions. This also indicates that the

competitive sprouting process in this case is nonspecific in

that the new terminals can be derived from various types of

axons growing through the area. Bernstein suggests that the

sprouts are responding to a nonspecific stimulus from the

dedifferentiatedd" dendrites and randomly reinnervate these

sites (Guth and Windle, 1970; Bernstein, personal communica-

tion). It may be the case, however, that only specific types

of axons initially respond to the sprouting "stimulus," since

in some instances specific contacts are formed. Murray (1973)

reports that after a partial hemi-section sparing the dorsal

columns, a consistent pattern of sprouting into specific spi-

nal laminae by dorsal root fibers can be observed. The re-

innervation pattern within the laminae can be correlated with


electro-physiological and behavioral observations on abnormal

cutaneous and muscular reflex development.

In addition to these spinal cord studies, further evidence

for specific interactions in the formation of new connections

by collateral sprouts has been derived from studies in other

brain areas. In the study mentioned above by Liu and Chambers

(1958) sprouting was seen not only in the spinal cord pro-

jections of the dorsal roots but in the sensory relay nuclei

as well, where it was limited to specific localized areas.

This can be interpreted as an indication that some type of

limitation was placed upon the sprouting dorsal root fibers

within the sensory relay nuclei. Goodman and Horel (1966)

also demonstrated a limited sprouting phenomenon in central

nervous system. Sprouting of optic tract fibers within the

optic tectum was seen only at some of the loci deafferented

by visual cortex removal. This suggests that at some of the

denervated loci either no fibers filled the vacated sites

or fibers from some other neural system, not shown by their

degenerative procedures, filled these sites. If the latter

is the case, then there must be some specificity factor oper-

ating at these regions which would select for the sprouted

fibers of other systems. These studies were replicated in

young animals by Schneider (1970) and Mathers and Chow (1973)

and in adults by Cunningham (1972). In a subsequent study

by Goodman, Bogdassarian, and Horel (in press) a generalized

sprouting of optic fibers at several different loci within

the optic pathway was observed. Following unilateral enu-

cleation in rats, sprouting of the contralateral optic tract

took place only at several of the regions where the projec-

tions from both eyes normally overlap.

From these and other studies these investigators postu-

late that sprouting occurs as a preferential process in which

both specificity and nonspecific competition play a role in

establishing the new connections. That is, for each dener-

vated synaptic site there exists some preferred type of axon-

al process that will be chosen over other sprouted fibers.

When the preferred ending is not available the postsynaptic

membrane can accept other types of sprouted presynaptic fi-

bers on a competitive basis. This is illustrated in the fol-

lowing example: In a series of studies with the electron

microscope, Raisman (1969) and Raisman and Field (1973) demon-

strated sprouting in the septal area. Terminal inputs either

from the hippocampus by way of the fornix or from the hypo-

thalamus and brain stem via the median forebrain bundle (I.FB)

synapse at different loci of the same septal cells and have

different ultrastructural characteristics. Lesions of either

fiber systems induced sprouting of the terminals of the other

system, primarily of those terminals on the same or adjacent

cell. Because the MFB fibers are known to have a considerable

adrenergic component in the septal area (Anden et al., 1966;

Ungerstadt, 1971a) Moore et al. (1971) were able to demon-

strate sprouting of this system with the fluorescent micro-

scope. Sprouting occurred only when the fornix section was

complete. The fluorescent procedure revealed no sprouting if

the lesions left any fibers of the fornix intact. This could

indicate that the remaining hippocampal axons sprouted pre-

ferentially over the adrenergic fibers and filled those sites

deafferented after the incomplete lesion. If this is the

case there seems to be both specificity and a nonspecific

competition that interact in this system, with the relative

influence of either process being determined by the conditions

of the lesion.

The presence of a specificity factor that can be modu-

lated to fit the existing conditions provides a basis for a

variety of possible functional consequences arising as a re-

sult of the sprouting process. In the above system the nature

of the lesion determined not only which fibers would sprout,

but the transmitter system that would reinnervate the area.

Few studies exist that have correlated sprouting with func-

tional effects, but as knowledge increases these problems are

being examined. For example, development of spasticity in

cats and the maintenance and strengthening of undulatory

movements in sharks after spinal cord section has been corre-

lated with sprouting (McCouch et al., 1958; Gelderd, 1972a;

1972b). After hemi-section of the spinal cord, sparing the

dorsal columns, Murray (1973) showed that collateral sprout-

ing of dorsal root fibers into specific spinal laminae could

correlate with sensory motor functions after the lesions.

Lashley (1935), Horel et al. (1966) and LeVere and Mor-

lock (1973) showed that after visual cortex removal, a pre-

viously learned brightness discrimination could be relearned.

This ability was not subsequently lost following total corti-

cal removal which indicated that some subcortical reorgainza-

tion took place. Goodman and Horel (1966) found that sprout-

ing of optic tract fibers in the superior colliculus could

provide a neural basis for these observations. Schneider

(1970) suggests that sprouting of visual pathways after visual

cortex or tectal lesions in neonatal hamsters is the basis for

the sparing of pattern vision observed when these animals are

measured in adulthood. It should be noted, however, that

Mathers and Chow (1973) could find no electrophysiological

evidence that functional contacts were formed from sprouting

optic tract fibers in the tectum following eneucleation in

infant rabbits.

Some electrophysiological evidence, however, does exist

to prove that sprouting does make functional contact and can

provide some functional reorganization. When entorhinal

lesions are performed in neonatal subjects, Fink-Heimer prep-

arations show the distribution of the commissural afferents

to the inner-molecular layer of the dentate gyrus increases

to include the outer-molecular layer in its zone of termina-

tion (Lynch et al., 1973a; 1973b). Laminar analysis of field

potentials in the molecular layer of these subjects shows that

connections which are capable of illiciting responses now

extend over the abnormal terminal area (Lynch et al., 1973b).

Similarly, Wall and Egger (1971) showed that in the ventral

lateral thalamic nuclei units that responded in the intact

organism to only forelimb stimulation became responsive to

both fore and hindlimb stimulation after lesions of nucleus

gracilis. This effect developed within 3 to 21 days after the


lesion,with some units reaching full response in 7 days.

This time course is the same as that seen in many studies of

both recovery of function and the sprouting process as meas-

ured by electron microscopy (Raisman, 1969; Lund and Lund,

1971), fluorescence microscopy (Katzman et al., 1970; Bjork-

lund et al., 1971; Moore et al., 1971), and light microscopy

using the Rassmussen silver technique (Gelderd, 1972a; 1972b).

These considerations indicate that axonal sprouting

following CNS lesions can give rise to many lesion effects

and,in some cases, may account for recovery of function (Smith

et al., 1973). Since it appears that the mechanisms for the

formation of new connections may differ with different types

of synapses, cells or brain regions (e.g., specificity in the

cord as opposed to the tectum), sprouting's contribution to

neural functioning will depend upon the conditions of the

system under consideration.

In the above discussion, collateral sprouting has been

characterized as a response of axons to local injury or dener-

vation. A recent study, however, indicates that axonal sprout-

ing may not be just an abnormal lesion induced growth process.

With the electron microscope, Sotelo and Palay (1971) were

able to show that under normal conditions, continuous remodel-

ing of synaptic connection by collateral axonal sprout forma-

tion can take place. These authors demonstrated that in a

normal animal all of the changes characteristic of both de-

generating and regenerating synapses and the ultrastructural

elements seen in growth cones of collateral sprouts were pres-


ent in the vestibular nucleus. These same types of changes

have been seen in other normal animals (Hashimoto and Palay,

1965). This observation allows for the speculation that

axonal sprouting occurs normally as a homeostatic process of

system maintenance and repair in the intact nervous system.

In a lesioned system this ongoing process could attempt to

serve the same function by filling in deafferented areas and

thus compensating for the loss of terminals to any given cell

of the CNS.

It thus appears that regeneration both of severed and

intact axons can and does take place in the mammalian central

nervous system. These processes, however, are limited by

many factors such as the location of the lesion, the environ-

mental conditions within the tissue, the age of the organism,

the type of fibers involved and the relative absence or pres-

ence of specificity interactions. When all of these (and other)

factors that may be involved permit, the structural and func-

tional integrity of the damaged neuronal elements can be re-


The evidence indicates that secondary changes following

lesions to central nervous tissues can give rise to a variety

of functional effects. Moreover, these factors should be

considered in determining the effects any lesion has upon the

organism and its behavior. At the present, however, this

approach is difficult and beset by numerous complications.

Most conclusive functional or behavioral effects have been ob-

served in simple input-output systems in which definitive

measures of changes can be clearly shown. In studying higher

brain functions there are complex interactions involved which

make this approach confusing. For example, research concern-

ing the limbic system indicates that some influences of this

system upon behavior are mediated through a hippocampal-sep-

tal system. Although damage to either structure or their

interconnections produces many similar effects on certain be-

havioral measures (Fried, 1972; Altman et al., 1973), the be-

havioral changes resulting from hippocampal lesions have, in

the main, been ascribed to the lack of hippocampal influence

(Douglas, 1967). As has been discussed previously, however,

sprouting of noradrenergic median forebrain bundle inputs to

the septum results from damage to hippocampal-septal projec-

tions (Raisman, 1969; Moore et al., 1971). The significance

of these changes in terms of the resulting septal influences

on behavior has not been examined.

As has been noted, the central catecholamine containing

axons have been shown to undergo plastic changes and contri-

bute to functional alterations resulting from these phenomena.

In addition, it is known that some behavioral responses can be

significantly altered by variations in the amount of catecho-

lamine innervation (Stein et al., 1972). These factors fur-

ther indicate that the type of reorganization that takes place

in the septum following hippocampal damage may contribute to

some of the behavioral changes observed. The remainder of

this discussion will be concerned with the nature of this sys-

tem and how these elements may function in modulating behavior.

This will be effected by considering:

1) Hippocampal-scptal interconnections and interactions,

and 2) the morphology of septal adrenergic inputs and the

effects of catecholamines on certain aspects of limbic mediat-

ed behavior.

Hippocampal-Septal Relations

The hippocampus interconnects with the septum through

a large fiber bundle known as the fornix. This bundle is com-

prised of the major efferent and afferent connections of the

hippocampus. The hippocampofugal fibers, originating in the

pyramidal layer of the hippocampus proper and the dentate

polymorph layer, first project into the alveus (a layer of

white matter overlying the ventricular surface of the hippo-

campus). These fibers subsequently coalesce to form a more

discrete bundle, the fimbria. The fimbria extends over the

dorsum of the hippocampus and divides into three distinct

groups. One of these proceeds through the hippocampal commi-

sure into the contralateral limbic cortex. The largest group

extends anteriorly, ventral to the callosum, forming the main

body of the fornix. At the level of the anterior commissure,

the fornix divides into pre and postcommissural columns

which descend into and through the diencephalic grey matter

to innervate neurons of the thalamus, hypothalamus, and mid-

brain. The majority of fibers in the posterior descending

column terminate in (or pass through) the mammillary bodies.

Those of the precommissural fornix project to the septal area

(Figure 1) and surrounding structures (n. accumbens, and the

olfactory tubercule).

The final branch of the fimbria, accompanied by a strip


..: ~
Iv ~

1V tz

Figure 1. Section through the Septal Area at the level of the Anterior
Commissure. Tungstate modification of Golgi-Cox stain (xlO).

of grey matter continuous with the dentate gyrus (the fascio-

la cinirea), extends over the splenium of the corpus callosum

to form the grey and white stria of Lancisi, or induseum

grisium and dorsal fornix. After traversing the superior as-

pect of the callosum the dorsal fornix dips into and passes

obliquely through the fibers of the callosum to rejoin the

descending fornix columns. Many of these fibers terminate in

the septal region (Papez, 1937; Sprague and Meyer, 1950;

McLardy, 1955a; 1955b; Valenstein and Nauta,, 1959; Crosby,

Humphery and Lauer, 1962; Raisman, Cowan and Powell, 1966;

Siegel and Tassoni, 1971a).

Fibers originating within the septum form a major compo-

nent of the hippocampal afferents of the fornix and the dor-

sal fornix. These fibers pass through the same fiber systems

and structures as the hippocampal efferents and terminate at

various loci within the hippocampus, dentate gyrus and peri-

rhinal cortex (Cragg, 1965; Siegel and Tassoni, 1971b;

Mellgren and Srebro, 1973).

Several studies exist which have dealt with the patterns

of termination of the fibers between the septum and hippo-

campus. In the rat, Raisman (1966), Raisman et al. (1966),

and Raisman (1969) demonstrated with Nauta -Gygax stained

material and electron microscopy that fibers from hippocampal

field CA1 project bilaterally to the medial portion of the

septum. Their terminal field includes the entire medial sep-

tal nucleus and the medial portion of the lateral septal nu-

cleus, as well as the vertical limb of the nucleus of the

diagonal band. Fibers from the cells of these areas were

seen to form a unilateral projection back into hippocampal

fields CA3 and CA4 (as well as the dentate area) which in turn

project bilaterally to the remaining major portion of the

lateral septal nucleus and the horizontal limb of the diago-

nal band nucleus. No projections to the hippocampus from

the lateral septal nucleus could be observed in these pre-


This does not seem to be the case in the cat, however.

Using Fink-Heimer techniques, Siegel and Tassoni (1971a;

1971b) demonstrated that although a reciprocal relationship

between the hippocampus and septum exists, no differential

patterns to the CA fields within the hippocampus could be ob-

served. Rather, there seemed to be a reciprocity of connec-

tions between the dorsal hippocampus and the medial septal

nucleus, and between the ventral hippocampus and the lateral

septal nucleus. (Fibers from the lateral nucleus, however,

were seen to project to hippocampus). The septo-hippocampal

fibers terminated on the apical dendrites of dentate granule

cells and on both the apical and basiler dendrites of hippo-

campal pyramidals throughout the CA subregions (Ibata et al.,

1971). These differences in projections are illustrated in

Figures 2 and 3.

Many techniques have been applied in an effort to study

the nature and determine the functional significance of these

connections. For example, it has been shown that some of the

septo-hippocampal fibers are cholinergic. Lesions involving

these fibers (found both in the main fornix body and dorsal

fornix) induce a decrease of acetycholinesterase (ACHE) activi-

Figure 2.

Diagrammatic representation of the differ-
ential origin and termination of Hippocam-
pal-Septal connections. Left proposed
pattern for the Cat from Siegal and Tassoni
(1971a). Right proposed pattern for the
Rat from Raisman (1966). Note bilateral
projection from ventral hippocampus (Cat)
and area CA1 (Rat) to lateral septal nucleus.
Abbreviations: (cc) corpus callosum, (den.)
dentate gyrus, (DB) nucleus of the diagonal
band, (DH) dorsal hippocampus,(Fx) fornix,
(LSN) lateral septal nucleus, (MSN) medial
septal nucleus, (sub.) subiculum, (VH)
ventral hippocampus.




I- ---VH to LSN
4-DH to MSN

9J --CA1 to MSN-DB
E t----CA3&4 to L SN-DB


Figure 3. Diagrammatic representation of the differ-
ential origin and termination of Sapto-
nippocampal connections. Left proposed
pattern for the Cat from Siegal and Tassoni
(1971b). Right proposed pattern for the
Rat from Raisman (1966). Abbreviations:
(cc) corpus callosum, (den.) dentate gyrus,
(DB) nucleus of the diagonal band, (DH) dor-
sal hippocampus, (Fx) fornix, (LSN) lateral
septal nucleus, (MSN) medial septal nucleus,
(sub.) subiculum, (VH) ventral hippocampus.


dowbftC C an


-m LSN to VH
|^-------MSN to DH
.UH.I.-. MSN to CA.34
I A---LSN to MSN

CA384 to CAI



ty in the hippocampus with a concomitant buildup of this

activity in cells of the medial septal nucleus and the diago-

nal band (Lewis and Shute, 1967; Mellgren and Srebro, 1973;

Pepeu et al., 1973). Cholinergic stimulation of cells in the

septum produce many measurable behavioral changes. Direct

application of ACH induces increased drinking and decreased

eating in deprived rats as well as impairing the performance

of a previously learned avoidance task (Grossman, 1964).

Blockade of this system by direct local application of atro-

pine reduces water intake, improves performance on a learned

operant avoidance response (Grossman, 1964), induces a pass-

ive avoidance deficit (Hamilton et al., 1968), facilitates

acquisition of a conditioned avoidance response (Kelsey and

Grossman, 1969), and increases locomotor activity in a run-

way (Leaton and Rech, 1972). Such effects of chemical block-

ade of neurons in the septum are similar to many, but not all,

behavioral changes seen after hippocampal or septal lesion.

Hamilton et al. (1968) found no impairment of position re-

versals or one-way active avoidance effects with this treat-

ment. These kinds of observations and those involving system-

ic injections of agents affecting the cholinergic system

(e.g., Carlton, 1962; Douglas and Isaacson, 1966; Baisden

et al., 1972) led to the speculation that this system is inti-

mately involved in a response suppression mechanism mediated

by the hippocampus and septum (Douglas, 1967; Kimble, 1968;

Fried, 1972). This mechanism appears to be a component of a

more general acetycholine fiber system involved with the

inhibition of behavior (Carlton, 1969).


Hippocampal theta rhythm, which is thought to be an elec-

trical manifestation of hippocampal influences in attention

and orienting (Bennet, 1971), has been demonstrated to depend

on the structural integrity of septo-hippocampal connections

(Green and Arduini, 1954), especially those originating in

the medial septal nucleus (Gray, 1971). Unit discharges in

the medial septal nucleus, the diagonal band nucleus and the

hippocampus have been found to have a high degree of correla-

tion with each other (Petsche et al., 1962; Morales et al.,

1970). These synchronous bursts were furthermore enhanced

by the application of exogenous physostigmine, indicating a

possible cholinergic mediation of these responses (Macador

at Al., 1970).

Evoked potentials have been observed to occur in the

hippocampus and septum after electrical stimulation to either

structure. CAl stimulation resulted in both excitatory and

inhibitory potentials bilaterally in several regions of the

septum. The characteristics of the responses, however,

differed within the septal region. Analysis of the waves and

latencies of these evoked responses indicated that direct

orthograde transmission of the hippocampal fiber terminals

produced the excitatory component of the observed wave form.

The inhibitory responses were the result of interneuron ac-

tivation and retrograde stimulation of septal efferents to

the hippocampus (DeFrance et al., 1971; 1972a; 1972b; 1973a;

1973b; 1973c; 1973d). Stimulation of medial septal areas

elicited responses in the dentate gyrus and field CA4 as well


as on the soma and apical dendrites of pyramidal cells of

CAl (Anderson et al., 1961a; 1961b).

These findings all imply a high degree of interaction be-

tween various regions of the hippocampus and the septum. The

medial septal nucleus seems to be the major septal subdivision

linking this system together. The exact significance of

these interactions, however, still remains to be determined,

although some light has been shed on this problem by be-

havioral observations following extirpation of these structures.

The behavioral alterations resulting from hippocampal

or septal lesions have been extensively studied on many be-

havioral measures. These studies have been compiled into

numerous reviews to which the reader is referred for in-depth

consideration of the functional-behavioral properties of

these brain regions (McCleary, 1966; Douglas, 1967; Kimble,

1968; Fried, 1972; Isaacson, 1972; Isaacson and Kimble, 1972;

Altman et al., 1973; Jarrard, 1973, and Lubar and Numan,


The similarities in performance on particular behavior-

al measures following lesions to either structure indicate

that the septal area and hippocampus are both involved in a

generalized behavioral inhibitory function (see reviews

above).This is manifested in deficits in the ability to re-

duce certain kinds of operant response rates (Ellen and

Powell, 1962; Ellen et al., 1964; Clark and Isaacson, 1965;

Jarrard, 1965; Schmaltz and Isaacson, 1966; Van Hoesen et al.,

1971; 1972; Rickert et al., 1973), passive avoidance deficits

(McCleary, 1961; Isaacson and Wickelgren, 1962; Schwartz-


baum and Spleth, 1964; Papsdorf and Woodruff, 1970), facili-

tated 2-way shuttle-box avoidance (King, 1958; Isaacson, Doug-

las and Moore, 1961; Van Hoesen et al., 1969; Papsdorf and

Woodruff, 1970; Woodruff and Isaacson, 1972), alterations in

exploratory activity (Douglas and Isaacson, 1964; Douglas

and Raphaelson, 1966b; Corman et al., 1967), and persevera-

tive responding in choice situation, e.g., T-maze (Douglas

and Isaacson, 1964; Douglas and Raphaelson, 1966a; Means et

al., 1971) after lesions to either structure. Following a

review of lesions of the nuclei within the septum and

their subsequent behavioral manifestations, Fried (1972) pro-

poses that those aspects of limbic disfunction indicative of

an inability to alter response patterns is mediated through

the medial septal nucleus. Other types of behavioral

changes, suggesting a disrupted response suppression system,

are mediated primarily through hippocampal-septal connections.

The exact substructures involved, however, are unknown at

this time.

This account of septo-hippocampal lesion effects and

possible interrelations indicates that a high degree of mor-

phological and functional interaction between these structures

exists and is important in the manifestation of either struc-

ture's effects on certain behaviors. This study is designed

in part to further examine this interrelationship and to at-

tempt to provide a better understanding of the behavioral in-

fluences of these two structures. This will be done by taking

advantage of what is already known about the adrenergic sys-

tem after a hippocampal lesion, and its known behavioral sig-


Morphological and Behavioral Influences
of Septal Catecholamines

Catecholamine fibers of the forebrain are derived pri-

marily from neurons found in nuclear groups of the midbrain

and brain stem reticular formation (Anden et al., 1966).

Fluorescent histochemical analysis has revealed that their

axons ascend in the median forebrain bundle system and dis-

tribute to various components of the ventromedial forebrain.

Adrenergic terminals innervating the septum are most probab-

ly derived from cells within the reticular system (Anden et

al., 1966; Fuxe et al., 1970; Ungerstadt, 1971a). The fibers

enter the septal area from lateral hypothalamic areas via the

diagonal band and have been demonstrated to terminate, with

regional differences, in all of the septal nucleus areas. Al-

though an extremely dense distribution of adrenergic terminals

is seen in the nucleus accumbens, and preoptic areas, the

heaviest projection within the septum proper is to the later-

al septal nucleus. Here the fibers terminate in irregularly

arranged plexi on the cell bodies and proximal dendrites of

the septal cells. The rostral portion of this nucleus re-

ceives larger fibers which progressively decrease in thick-

ness toward the more caudal areas. The terminals of the me-

dial nucleus and the nucleus of the diagonal band receive

somewhat fewer but larger, more coarse terminals. These are

arranged in short vertical rows. No evidence of any dopa-

mine input to this structure has been demonstrated (Moore

et al., 1971).

l-though the behavioral significance of discrete nor-

adrenergic innervation to a structure is not presently known,

studies of adrenergic influence on behaviors have shown that

this system affects at least some kinds of consummatory be-

haviors, reward mechanisms and avoidance situations. For

example, destruction of noradrenaline terminals by intra-

cisternal injections of 6-OH-dopamine induces a suppression

of food and water intake (Zigmond and Stricker, 1972;

Hansen and Whishaw, 1972), grooming behavior (Hansen and

Whishaw, 1972), self-stimulation rates (Hansen and Whishaw,

1972), and acquisition of shuttle-box avoidance (Cooper et

al., 1972). In addition, these animals seem to display an

increased fearfulness in open field measures (Bresler and

Ellison, 1972), and are hypersensitive to handling (Hansen

and Khishaw, 1972). Some of these changes have been shown to

result from local intracerebral injections of 6-OH-dopamine.

Injections into lateral hypothalamus induce a decrease in

self-stimulation (Stein et al., 1972) as well as an adipsic

and aphagic syndrome similar to that seen following electro-

lytic destruction of this area (Smith et al., 1972).

Various manipulations of animals with this type of le-

sion have provided further evidence for noradrenergic involve-

ment in these behavioral situations. Interventricular in-

jection of norepinephrine has been shown to reverse the

lateral hypothalamic-anorexic syndrome (Berger et al., 1971)

and restore self-stimulation rates suppressed by 6-OH-dopamine

(Stein et al., 1972). Similarly 6-OH-dopamine induced sup-

pression of self-stimulation can be prevented initially by


pre-treatment with chlorpromazine (which prevents 6-OHI-dopa-

mine depletion of norepinephrine).

Direct stimulation of various structures by norepineph-

rine application has also provided evidence for an adrener-

gic effect on certain behaviors. Lateral hypothalamic as

well as amygdaloid and septal application of this agent re-

sults in increased food intake (Grossman, 1960; 1964; Booth,

1968). A deficit in passive avoidance after administration

of norepinephrine to the amygdala (Margules, 1968), and an

improvement in a learned bar press escape avoidance response

following application to the septum (Grossman, 1964) have

also been observed.

Stein et al. (1972) proposes that a reward system is

mediated through the adrenergic component of the MFB. This

is evidenced by the increase in consummatory responses (e.g.,

feeding and copulation) and the concommitant increase in

norepinephrine following rewarding or self-stimulation of the

MFB. Norepinephrine in this system would mediate its facili-

tatory effects by inducing a suppression in tonically active

elements. For example, the passive avoidance deficit follow-

ing adrenergic stimulation of the amygdala (Margules, 1968)

could result from the over suppression of the normal behavior-

suppressive effect of punishment. Normally, however, these

mechanisms would be activated in behavioral situations pro-

ducing pleasure or avoiding pain. This could be illustrated

in the case where septal adrenergic stimulation increased

avoidance responses over escape responses in a bar-press es-


cape-avoidance task (Grossman, 1964).

On the basis of these considerations it is reasonable

to speculate that an increase in adrenergic innervation to

a structure may modify its function in such a way as to pro-

duce behavioral manifestations. It may be the case that

these changes only influence certain types of behavior as

indicated above, or, when adrenergic terminals innervate

synaptic sites not normally receiving this type of transmitt-

er, as is the case following MFB sprouting in the septum,

they may demonstrate an influence on different aspects of


The following experiment is designed to examine the con-

tribution of septal adrenergic input on the behavioral syn-

drome induced by hippocampal lesions. This is accomplished

by comparing the behaviors of hippocampal animals with those

that have undergone selective destruction of the adrenergic

component of the septal area by local injection of 6-OH-

dopamine prior to hippocampal lesions.

It is predicted that some of the behavioral deficits

that result from hippocampal lesions (i.e., in those be-

haviors mediated through the hippocampal-septal system and

influenced by the information normally provided by the ad-

renergic fibers which are destroyed by septal 6-OH-dopamine

injections) will not be altered in the same manner usually

seen when the lesion is proceeded by septal 6-OH-dopamine

treatment. It is also predicted that any behavioral alter-

ations that do occur in this set of experiments will be sub-


ject to the functional and structural reorganization taking

place in the septum as a result of the combined lesioning

procedures employed.



Seventy-four male long Evans hooded rats were employed.

They were housed individually under a 12-hour on 12-hour

off light-dark cycle. The subjects were divided into groups

according to the procedures described in the following

sections. The surgical procedures will be outlined before

describing the behavioral methods employed.


General Surgical Procedures

Surgical anesthesia was achieved by injection of nembu-

tal (sodium pentobarbital, 50 mg/kg, i.p.). The subject

was placed into a Kopf stereotaxic apparatus, the scalp was

shaved and a midline incision made exposing the dorsum of

the skull. After removal of the scalp muscles and fascia

by blunt dissection, the animal was subjected to one of the

various surgical procedures described below. Stereotaxic

coordinates for electrolytic lesions and drug injection

were obtained from the atlas of Pelligrino and Cushman

(1967). System B of this atlas was used.

Following surgery, hemostasis was obtained and the

scalp was closed either by suturing or with 9 mm Clay

Adams wound clips. Bicillin (25,000 units, i.m.) was then

given and after the rat regained upright posture, it was

returned to the home cage.

Aspirative Lesions

After exposing the skulls as described above, the

temporalis muscle was removed by blunt dissection from its

insertion along the lateral edge. Holes were made with a

5 mm burr on the dorsolateral surface of the skull, approxi-

mately 2 mm behind bregma. Care was taken during the drill-

ing procedure to prevent subdural hematoma due to overheat-

ing by frequent applications of saline to the skull. Ron-

geurs were used to expand the opening in the skull both

posteriorly and down the lateral surface of the parietal


Aspirative lesions of the hippocampus and/or overlying

neocortex were produced using a 22 gauge aspirator needle.

This procedure has been described previously for the rat by

Isaacson, Douglas, and Moore (1961). Care was taken to re-

move the hippocampus as completely as possible in order to

destroy all output through the fornix.

Electrolytic Lesions

After exposing the skull, two small holes were drilled

with a 1 mm round dental burr, bilaterally, just anterior

to bregma. Care was taken (as above) to prevent overheating

and subdural hematoma. Using small iris rongeurs the burr

holes were expanded and carefully connected across the mid-

line. With the midsagittal sinus exposed,electrodes made

from .00 gauge stainless steel insect pins, coated with

epoxylite resin were stereotaxically placed at coordinates:

2 mm anterior to bregma and 5 mm deep. Lateral placements

vary with the course of the sinus as the electrodes were

placed bilaterally as close to the midline as possible.

Anodal lesions were produced with a C. H. Stolting lesion

maker (model 58090) by passing 3 mA for 10 seconds. Follow-

ing removal of the lesioning electrodes, bone wax was used

to seal the skull opening and the wound closed as above.

Intracranial Injections

Stereotaxic and surgical procedures were as described for

electrolytic lesions. Injections into the septal area were

made through a double cannula. The outer cannula consisted

of blunt 22 gauge hypodermic tubing. In order to prevent

clogging of the cannula, a stylette (made from a 27 gauge

needle) was inserted into the outer sleeve. After position-

ing the cannula, the stylette was replaced by a 27 gauge

inner cannula. This was connected to a microliter syringe

by P. E. 50 intramedic polyethylene tubing. Injections were

made with a Sage syringe pump (model 341). The cannula and

stylette were designed to extend approximately 1 mm beyond

the outer sleeve. This helped prevent infusion of the

chemical back up the cannula between the two tubes.

Four ul of solution were injected bilaterally into the

septum at a rate of .2 ul/min. Vehicle control injections

consisted of a carrier composed of 1 mg/ml ascorbic acid in

a 0.85 percent aqueous sodium chloride solution (Hedreen and

Chalners, 1972). A solution of 2 ug, 6-hydroxydopamine HBr

(Sigma Chemical) per ul vehicle was prepared just prior to

use and stored under refrigeration until used. The portion

remaining after surgery was discarded. If, during use, the

solution became discolored it was discarded and a fresh

sample made up.

Behavioral Testing: Experiment I

Preoperative DRL Training

Fourteen rats were deprived of food and maintained on a

regimen which kept them at 85 percent of their ad libitum

weight. They had full access to water in their home cages

throughout DRL training. Barpress acquisition, continuous

reinforcement and differential reinforcement of low rates of

responding (DRL) were all performed in a standard Lehigh

Valley 2-bar operant chamber (Model #1417). The subjects

were shaped to press only one of the bars. The other bar

was always retracted. Electro-mechanical relays, counters,

and timers controlled the recording of responses and pre-

sentation of reinforcement. Each rat was run for 30 minutes

five days a week. The rats were shaped to press the bar on

a CRF schedule and remained on this schedule until stable

barpress rates were obtained (8-14 days). They were then

placed on a DRL-20 schedule. They remained on the DRL

schedule until they were responding at a steady rate over

days and had achieved a level of over 25 percent reinforced

barpresses. In addition to total presses in 30 minutes

iTwo animals failed to reach the 25 percent correct re-
sponse criterion level on the DRL task prior to operation
after normal testing for 60 days. Due to time demands these
subjects were operated on at this time. Differences in the
scores obtained on these rats are noted in the Results

and percent correct, the interresponse times were tabulated

for each rat during each session.

Postoperative DRL Training

After each rat had developed a stable pattern of respond-

ing (10-20 days of DRL see footnote previous page) the

first operation (vehicle injection) was performed. Follow-

ing the DRL session the day before surgery, the subjects

were allowed free access to food overnight. The next day,

the animals received an intracerebral injection of the vehicle

carrier solution following the procedure described above.

After a 2-day recovery period, the animals were returned to

the DRL schedule and tested for 10 days. At this time one-

half of the subjects received another vehicle injection and

were designated to the vehicle control group (C). The re-

maining one-half were given the vehicle and 6-OH-dopamine

(0.2 ug/ml) and were placed in the experimental group (E).

After a 2-day recovery period, the subjects were again tested

on the DRL schedule for 10 days. Bilateral hippocampal ab-

lations were then performed on all subjects. Animals were

allowed 5 days for recovery before DRL testing was resumed

for 10 more days. Time between the second injection proce-

dure and hippocampal lesion was thus 14 days. After comple-

tion of DRL testing the rats were given food ad libitum and

placed on a 23-hour water deprivation schedule. Then they

were trained to obtain water in the passive avoidance chamber.

Passive Avoidance Training

Passive avoidance measures were obtained in a black

Plexiglas box 24" long and 10" wide and 12" deep. The box

was divided into two compartments 12" x 10" by a guillotine-

like black Plexiglas door. The compartment designated "shock

chamber" was fitted with a standard laboratory water bottle

of the type provided in the home cage. The "start chamber"

was bare. Each compartment was covered by a clear Plexiglas

door. A scrambled 5 mA footshock of 10 seconds duration was

administered to the grid floor of the "shock chamber" by a

shock generator (Grayson Stadler, model #E-106-GS). Presen-

tation and termination of shock was controlled by standard

electro-mechanical relay equipment.

Each rat was placed by hand into the start chamber and the

connecting door to the shock chamber was raised. Latency of

entrance into the shock chamber was measured. The subject

was allowed access to the water bottle for 10 seconds before

being replaced into the start chamber. On the sixth consecu-

tive entrance into the shock chamber the animal was subjected

to a footshock of the above parameters. After a 30-second

interval,latency of entrance back into the shock chamber was


Water Consumption

At the conclusion of passive avoidance testing the rats

were given full access to water for three days. Following

this period the subjects' water consumption was measured over

a 24-hour period. The number of grams consumed in each

period was recorded. At the conclusion of the water con-

sumption measure, Experiment I was concluded.

Behavioral Testing: Experiment II

Open Field Locomotor Activity

Rats were randomly divided into 6 groups of 10 subjects

each. These groups consisted of: (1) Group (D), animals

with 6-hydroxydopamine injected into the septum followed 28

days later by a bilateral hippocampal aspiration lesion; (2)

Group(V), animals with a septal injection of the vehicle

carrier solution followed by bilateral hippocampal ablation

28 days later; (3) Group (S), animals with septal electrolyt-

ic destruction followed 28 days later by bilateral hippo-

campectomy; (4) Group (N), normal controls; (5) Group (H),

animals with bilateral hippocampectomy; and (6) Group (C),

animals with bilateral neococtical destruction.

All subjects were tested in an open field activity cham-

ber for 7 days preoperatively and 25 days postoperatively (or

following the first surgical procedure in groups 1-3 above).

Those animals receiving bilateral lesions 28 days after the

first surgical procedure were tested in the open field for

10 days after 5 days'recovery from the second operation.

Control groups (4-6) were anesthetized only, allowed 5 days'

recovery, and tested 10 more days in the open field.

Two open field activity chambers of dimensions 40" x

40" x 18" deep were used. Each apparatus was equipped with

six sets of photocells and light columnators dividing the

floor of the chamber into sixteen equal squares. Counts were

monitored on remotely located standard electro-mechanical re-

lay equipment. The measure of activity consisted of the

number of times the subject broke a light beam during the

testing session. All subjects were run in daily 5-minute

sessions, throughout the testing period. At the beginning of

each session the subject was placed in a specific corner.

The same corner of the same activity chamber was used consis-

tently with each subject. Illumination during testing was

provided only by the columnated light beams within the appa-

ratus. A constant background masking noise was provided by

a ventilator fan in the operant chambers used for DRL test-


Passive Avoidance

At the conclusion of activity measurement the subjects

were deprived of water for 23 hours and passive avoidance

training was given as in Experiment I.

Water Consumption

Following passive avoidance training, the subjects

were allowed free access to water for 3 days. Water con-

sumption was then measured as in Experiment I.


At the conclusion of behavioral testing all operated

subjects were sacrificed with an overdose of Nembutal. They

were intracardially perfused with 0.9 percent saline followed

by a 10 percent formalin solution. The brains were removed

and embedded in celloidin. Sections were taken at 30 u.

Every section through the septal area and every fifth section

through the hippocampus was retained, slide mounted, and

stained with thionin for Nissl substance. Reconstructions


were made of the hippocampal and neocortical lesions.

Photomicrographs were 'made of the appearance of the septum

of animals in which 6-OH-dopamine injections, vehicle carrier

injections and electrolytic lesions were performed.


Histological Observations

Examination of the histological preparation of the

septum revealed that successful bilateral injections were

performed in all control animals and in all but one of the

experimental animals. In Subject E-6, a cannula tract could

be located only on one side of the septum. The other

cannula may have extended into the lateral ventricle, since

there was no evidence of its penetration into any subcortical

structure. This subject's scores were not included in the

group analysis for any behavioral measure.

In general, the cannula appeared to pierce the lateral

septal nucleus and extended ventrally to enter the medio-

lateral aspects of the medial septal nucleus. Photomicro-

graphs of representative sections of the septum following

the double injection procedure from subjects in both control

(Group C) and experimental (Group E) groups are presented in

Figure 4.

The amount of tissue destruction resulting from the drug

injections appeared equal to that of the subjects receiving

only the vehicle carrier solution. Contrary to the report of

Poirier et al. (1972), there was no sign of extensive tissue

damage resulting from local injections of 6-OH-dopamine.

Gliosis and scarring in the septum was observed only around

the cannula tract.



Photomicrograph of representative sections through the
Septal Area in subjects of Experiment I. Top row (sec-
tions A and B) from two subjects receiving a double in-
jection of vehicle carrier solution. Bottom row (sections
B and C) from two subjects receiving intraseptal injection
of vehicle followed by injection of 6-OH-dopamine.

Figure 4.

Brain reconstructions representing the extent of the

hippocampal lesions to the subjects in Experiment I are

presented in Figure 5. Hippocampal damage was extensive in

all cases. In most subjects only a small portion of the

ventral hippocampal formation remained. Extra-hippocampal

damage was confined to neocortex and corpus callosum. Only

the expected amount of gliosis resulting from neocortical re-

moval (Woodruff et al., 1973) was seen in the thalamus. This

was primarily restricted to the LGN (lateral geniculate

nucleus) and the MGN (medial geniculate nucleus).

Behavioral Analysis


Statistical analysis was done using a standard analysis

of variance and t-test for related samples (Furguson, 1966).

Both the grouped scores and those obtained for each subject

in terms of the rate of response, number of correct responses

and percent correct responses were compared both to normal

baseline levels (obtained during preoperative testing) and to

levels of performance attained following all other treatments.

Summary tables and graphical representations and charts of

the data and statistical finding are presented in Tables 1

and 2 (text) and in Tables 1 6 and Figures C-1 through C-7

and E-1 through E-7 in Appendix A.

Preoperative performance. All but two of the 14 original

subjects learned the DRL-20 task to an average performance

level of 20 to 25 percent correct responses over the

daily sessions for ten consecutive days (see footnote

Method section). One of the two subjects which performed







X -H


O d
4 -H



::I (


U- C *~ #4 0 .4 C 9 CU
4. 4. 4. 4. 4. 4. 4- 9 I I I



Experiment I Number of subjects showing increased (i);
decreased (d); or no change (x) in DRL-20 performance
Experimental (6-OH-dopamine Treated Group)

of Responses

(i) (x) (d)

3 3 1

4 2 1

3 3 -

2 3 1

3 1 2

3 2 1

Number Number
Correct Responses Correct Responses

(i) (x) (d) (i) (x) (d)

1 3 3 1 3 3

2 5 3 4

6 1 5

2 4 3 3

1 5 1 1 4

1 5 3 3

= Normal (pretreatment) testing period

= Vehicle treated testing period

= 6-OH-dopamien treated testing period

= post hippocampal testing period

to V

to D

to H

to D

to H

to H



Experiment I Number of subjects showing increased (1);
decreased (d); or no change (x) in DRL-20 Performance

Vehicle Control Group

of Responses

(i) (x) (d)

3 4 -

2 2 2

5 1 -

1 1 4

6 -

6 -

Number Percent
Correct Responses Correct Responses

(i) (x) (d) (i) (x) (d)

5 2 4 3

2 2 2 3 1 2

1 1 4 1 1 4

4 2 3 2 1

6 6

6 6

N = Normal (pretreatment) testing period

Vl= testing period following 1st vehicle injection

V2= testing period following 2nd vehicle injection

H = post hippocampal testing period

N to V1

N to V2

N to H

V1 to V2

V1to H

V2to H

poorly was placed into each group. The subject placed in the

control group (C-7) died during the second surgical procedure.

The other subject (E-i) survived all operative procedures.

Its scores, however, were quite different from those of other

subjects in the experimental group. Since there was no histo-

logical basis for the elimination of this subject from the

group comparisons, its scores are included in the statistical

analysis of the group performance. One experimental subject

(E-7) failed to recover after the hippocampal lesion. Its

scores were eliminated from the analysis of the grouped data.

The performance of each subject is shown in the tables and

figures of Appendix A.

Performance after the first vehicle injection. All sub-

jects in the study received a control injection of the vehicle

carrier solution. This resulted in an average increase in the

total number of responses (p <.01) and a decrease in the num-

ber of correct responses (p <.05) relative to preinjection

levels. There was no significant change in the percent of re-

inforced responses. The changes are reflected in the grouped

IRT distribution (Figure 6) by a slight increase in the num-

ber of responses made throughout the 20 second-inter-response

interval (as measured in ten 2-second bins) for both the ex-

perimental and vehicle control groups.

These changes in average performance levels, however,

were due to changes in only six of the subjects. Five of the

subjects maintained preinjection levels and one subject (E-5)

decreased his response rate and increased the number of cor-

rect responses. Charts on each subject's performance are pre-

Figure 6. DRL-20 Group Meen IRT Distribution. Top -
Vehicle Control group scores. Bottom -
Experimental (drug treated) group scores.
Abbreviations: (veh) vehicle injection,
(6-OH-D) drug injection, (hpx) hippocampal

Control Mean IRT

2 t -1 m I

Experimental Mooa IRT

* 4.

I g
I a
I a
II 0





d *




r '
1 V \!\
\1 g \

2d I 2O I

Seconds on-p

21 1 29








* *





sented in Tables 1 and 2 (text) and Tables 1-6 .:-.pendix A.

Performance after the second vehicle (control) injection.

All subjects in the control group (i.e., the group not receiv-

ing 6-OH-dopamine injection into the septum) received a second

injection of the vehicle carrier solution. After this proce-

dure the group showed an average decrease in total responses

(p < .005) and an increase in the number of correct responses

(p< .005) when compared to the levels attained after the first

vehicle injection. This resulted in a return of the total

number of responses made during the daily testing sessions to

levels similar to those seen prior to the first injection.

The number of correct responses, however, was significantly

greater than during the pretreatment period (p <.05). This

result is reflected in a mean decrease in the number of unre-

inforced responses made during the inter-reinforcement (IRT)

period (see Figure 6 text).

This type of behavioral change was seen in three of the

six subjects of this group. One of the subjects (C-l)

showed a change in the opposite direction on both measures.

The remaining subjects either did not alter their performance

levels relative to those attained prior to the second injec-

tion (C-2) or decreased both number of responses and number

of correct responses (C-6). These changes are illustrated in

Appendix A; Tables 1-3 and Figures C-1 to C-6.

Performance after 6-OH-dopamine (experimental) injections.

No changes in the performance levels relative to performance

after the vehicle injection followed the injection of


6-OH-dopamine. The changes in performance after the first

vehicle injection were maintained in this group and were mani-

fested by an increased number of responses (p <.05) and a de-

crease in the number of correct responses (p <.005) relative

to preinjection (normal) scores. This is reflected in the

mean IRT by an increase in the number of responses in the

two second bins in the latter part of the 20 second inter-

response interval (see Figure 6, text). These finding are

illustrated in Table 2 of the text.

In two subjects of this group (E-2, p <.025; E-5 p <.025),

a statistically significant increase in the rate of respond-

ing occurred during the testing sessions following 6-OH-

dopamine injection. Only one subject, however, showed a cor-

responding decrease in the number of correct responses. An-

other subject decreased his response rate, but there was no

change in the number of correct responses obtained following

the previous treatment period. These observations are pre-

sented in Appendix A; Tables 4 6, and Figures E-1 through


In the subject that received only a unilateral septal

injection of 6-OH-dopamine, no statistically significant

changes occurred in either total number of responses or the

number of correct responses. A slight change in this sub-

ject's performance, however, was indicated by a significant

decrease in the percent of correct responses achieved (p<.05)

(see Figure E-6, Appendix A).

Performance after hippocampal lesions. Following hippo-

campal lesions, all subjects in the vehicle treated control


grouo showed a mean increase in the number of responses

(p< .005) and a decrease in the number of correct responses

(p< .005) relative to any previous performance levels (for

individual subject confidence intervals, see Tables 1 3,

Appendix A).

No increase in the average number of responses was seen

in the experimental group following hippocampal lesions. An

overall decrease in the mean number of correct responses

(p <.005) and in the total percent correct responses (p< .01)

was seen. An increase in the number of unreinforced re-

sponses (as demonstrated by the mean IRT distribution Fig-

ure 6) was observed in the bins of the latter half of the

interreinforcement period.

All but one of the experimental subjects receiving the

combined septal 6-OH-dopamine-hippocampal lesion treatment

showed the decrease in the number of correct responses in-

dicated by the grouped scores. Subject E-3 did not change

its previously acquired level of correct response performance

although its response rate was seen to decrease (p <.025).

Two subjects of this group increased their rate of responding

after the hippocampal lesion. One of these was the subject

which failed to adequately learn the task during the preopera-

tive testing period (E-l). Subject E-4 also altered its per-

formance in this direction. These results are illustrated in

Table 2 (text) and Tables 4 6 and Figures E-1 through E-6

(Appendix A).

The subject which received a unilateral septal injection

of the drug, showed the expected hippocampal lesion effects,

namely, an increase in the number of responses (p <.005) and

a decrease in both the number of correct responses (p <.005)

and percent correct (p <.005) (see Figure E-6, Appendix A).

Passive Avoidance

At the end of the DRL testing period, all surviving

subjects were tested on the passive avoidance task. All

but one subject of the control (vehicle injected-hippocampal

lesion) group left the start chamber of the passive avoidance

apparatus before the 5 minute time limit during the test

trial. These five subjects entered the shock chamber within

11 seconds on the first post-shock trial (see Table 3, text).

In the experimental group, two of the five surviving sub-

jects which had received bilateral injections entered before

the maximal time period allowed. The latency scores of the

two subjects, however, was considerably higher than the scores

in the control group (above). For example, subject E-1

waited for 270 seconds and subject E-5 waited 50 seconds be-

fore entering the shock chamber (see Table 3, text).

A Mann-Whitney "U" test (Siegel, 1956) indicated that

this difference in performance between the two groups was

statistically significant (p <.05).

In the subject receiving only a unilateral injection of

the drug, test latency was 2 seconds before crossing into

the shock chamber. This was comparable to the scores seen in

the vehicle treated control group (see Table 3, text).

Water Consumption

No significant differences in amount of water consumed



Experiment I Passive Avoidance Latency Scores

Control Group

Subject Number Latency

C-I 5 sec.

C-2 2 sec.

C-3 9 sec.

C-4 300 sec.

C-5 11 sec.

C-6 1 sec.

Experimental Group

E-1 270 sec.

E-2 300 sec.

E-3 300 sec.

E-4 300 sec.

E-5 50 sec.

E-6 2 sec.


was observed between the control and experimental groups

based on a t-test for independent samples (Hays, 1963).

Similarly, a t-test for related samples (Furguson, 1966)

revealed no differences in amount of water consumed by in-

dividual subjects over the two consecutive days tested (see

Table 7, Appendix A).


Histological Observations

Successful bilateral cannula placements into the septum

were observed in all subjects receiving 6-OH-dopamine injec-

tion and all but one of the animals receiving the vehicle in-

jection. Tissue damage was similar to that found after septal

injections in Experiment I. Photomicrographs of the septal

area following this single injection procedure in both control

and drug treated groups are presented in Figure 7. In one

control subject the brain was grossly distorted and there was

no evidence of a septal injection. This subject (V-7) was

eliminated from the study.

All electrolytic lesions of the septal area attempted

were verified, but in only one subject was the entire septum

involved. In most instances small lesions involving the

dorsal aspects of the medial septal nucleus and portions of

the lateral septal nucleus were observed. Photomicrographs

of representative sections through brains with small, medium

and large septal lesions are presented in Figure 8.

As in Experiment I, large hippocampal lesions were made.

Two vehicle-treated subjects and one subject with an electro-

lytic lesion of the septal area died following hippocampal

surgery. In addition, one subject from each septal treated

group (Vehicle, 6-OH-dopamine, and septal) were eliminated

from the study because of substantial thalamic damage. One


Photomicrograph of representative sections through the Septal
Area in subjects of Experiment II receiving intraseptal in-
jections. Top row (sections A and B from subjects receiving
single injection of 6-OH-dopamine. Bottom row (sections B
and C) subjects receiving single vehicle injection.

Figure 7.

Figure 8. Photomicrograph of representative sections
through the Septal Area in subjects of
Experiment II receiving Septal electrolytic
lesions. Top smallest amount of tissue
destruction; Middle average amount of
tissue destruction; Bottom largest amount
of tissue destruction.


.. .

subject from the hippocampal lesion control group and the

cortical control group were eliminated due to the extra-hippo-

campal and extra-cortical damage incurred during the respec-

tive surgery. Secondary tissue destruction (e.g., thalamic

scarring and gliosis) was similar to that reported in Experi-

ment I. Reconstructions illustrating the extent of the maxi-

mum and minimum hippocampal and cortical lesions used in this

experiment are presented in Figures 9 and 10.

Behavioral Analysis

Open Field Activity

A two-way analysis of variance demonstrated no effect on

open field activity between normal control levels and any

group of animals receiving septal treatment. Similarly, no

effect of the various septal treatments was seen when post-

operative scores were compared to pre-operative scores. In

addition, activity levels attained by those subjects receiv-

ing hippocampal lesions along with septal treatment did not

differ from the levels demonstrated by the group receiving

hippocampal lesions without prior septal treatment. Cortical

control subjects did not significantly alter their activity

scores following the lesion.

Individually, all subjects receiving hippocampal lesions

alone or following vehicle injections showed an increase in

activity levels. Six of the nine subjects which received 6-OH-

dopamine prior to hippocampal lesions showed an increase af-

ter hippocampal destruction. In the septal electrolytic

group, however, only four of the eight surviving subjects

demonstrated an increase in activity following hippocampal


*H H
r- H

r:1 .



0 0

H ,Q
4-) -r

r, r-i



f4 0

0 4

,- C C C Es -
4 4 4. 4. 4. 4. 4. S


HUMMU -Mi! la"E -RIA


BMW tgr ja:-R



F 0

r-H i

rd *
) 0

*d -H


-. U
4J -H4

C) 0

-I a)






a* *owueNo o wawe
$@gaeous op",e


lesions. It should be noted that eight of the ten normal

subjects also increased their activity levels over the time

period of the study. Tables 2 7 Appendix B, contain the

results of the statistical analysis of these data and individu-

al activity changes.

Passive Avoidance

Effect of septal treatment. Significant effects on the

ability to perform the passive avoidance task were found

only in the group receiving electrolytic lesions of the septum.

A Kruskul-Wallace analysis of variance (Siegel, 1956) re-

vealed that the subjects of the septal lesioned group were

significantly worse on this measure than those of the other

septal treatment and normal control groups (p <.005). Treat-

ment of the septal area with 6-OH-dopamine did not affect the

ability to avoid the shock chamber on the test trial. Com-

parisons of individual latency scores for these groups can

be obtained by referring to Table 4 (text).

Effect of Hippocampal lesions. Hippocampal lesions im-

paired the ability to perform the passive avoidance response

in all groups. A Kruskul-Wallace analysis of variance re-

vealed that no protection was afforded by intraseptal injec-

tion of 6-OH-dopamine prior to hippocampal ablation. Laten-

cy scores are presented in Table 5 (text).

Water Consumption

As in Experiment I, no consistent effect on the amount

of water consumed was induced by any septal treatment, aspira-

tive lesion (of hippocampus or neocortex) or combined surgical

Ir ,


C 0 0 0 C 0C
0 0 0 0 0 0C
ci cY i ( Y )

0 0 0
I 0 C 0
ci) ci (n



-r -I

4 >

a) 0
> P
W (



O CD 0 D Co o O O
o OC) ,- O O C CC (D

H N m n m m 0 m ON
r-oi cM c oo c Lo uO t" o O' O

o o 'r o -CD I o
0 0 00 r-A 0

nI I I I I I I I I Ir-
U) co nU) Um u U) m) U)

H 4

0 )

0) --

r-i (N cl "T )n k No Co or H0i

> > > > > 0> >
CO f- CD 0 Co O fl CD CD CD
m H Cim ( Y H Y n m

-I oi i v in o r- co o H-
z z z z Z z 0 z z





-,-- -I 0 1 -I

0 o
'0 0 4-1 I
1 0 U
10 04
0 Q -r I I
I -- I I
0 0 0 Q

in (N (N C)o in m
CN C v



in w

) 0 I co Ln r-
HO (N k-. H

040 4-)
0) 0 U
Mn 0 H (N (N *W1 in
a 4




m N

O. Q

U l
n *-H





I 0

N 0

r- co

I o

l l

O o I N o0


U) 4 Cl
HO 0 U3
)i 0 H

U d
0r- J U


U) 0 io


0 0 Lo
r-A u
(0 U)


0 0
04 n 1



Q4 U

H-i t.
r4 v


procedure. Differences were not seen in comparisons made

either within groups across treatments or days, or in those

made between groups. These observations are summarized in

Appendix B, Tables 8 and 9.


In rats with noradrenergic depletion of the septal area

produced by bilateral intraseptal injections of 6-OH-dopa-

mine, hippocampal ablation 12-14 days later did not result

in the expected increase in response rate on a DRL-20 sched-

ule, even though the usual decrease in correct responding on

this task was observed. These subjects also evidenced a re-

duced impairment in a passive avoidance task when measured

15-20 days after hippocampal damage. Another group of rats

received hippocampal lesions 30-35 days after the drug in-

jection. Pretreatment with 6-OH-dopamine in the septum did

not protect these subjects from lesion-produced increases

in open field activity or deficits in passive avoidance re-

sponding (measured 15-20 days after the hippocampal ablation).

No effect on water consumption was seen on any of these drug-

treated, hippocampal-lesioned subjects.

Injection of 6-OH-dopamine in the septal area prior to

hippocampal ablation resulted in an inability of the treated

subjects to improve DRL performance. Although the subjects

receiving vehicle control injections significantly increased

their efficiency (increased correct responses and decreased

response rate) on this task, the drug treated group maintained

levels acquired during the previous period of testing. No

consistent effects on water consumption, open field activi-

ty or passive avoidance were seen in the drug treated sub-

jects prior to hippocampal removal.

Animals with electrolytic lesions of the septal area

did show a deficit on passive avoidance; however, this

lesion did not significantly affect either water consumption

or open field activity. In addition, septal lesions did not

protect these behaviors from manifesting usual hippocampal

lesion effects.

Bilateral intraseptal injections of 6-OH-dopamine in

the amounts used here were previously shown to induce the

destruction of adrenergic terminals in an area of 2 mm

radius around the tip of each injection cannula (Ungerstadt,

1968). This could have resulted in at least four possible

secondary effects in the affected area; 1) alteration in the

functional balance within the septum maintained between the

adrenergic input and that of other catecholaminergic (e.g.,

serotonergic) projections, 2) physiological changes in the

denervated neurons (Stein et al., 1972), 3) morphological

changes (e.g., sprouting and regeneration ) in the intact

fibers and terminals within the area, and in the lesioned

adrenergic fibers themselves (Anden et al., 1966; Nygren and

Olson, 1971), and 4) prevention of sprouting of septal nor-

adrenergic terminals normally seen after hippocampal lesions

(Moore et al., 1971). These possible changes are utilized

in the following analysis to provide a basis for the observed

behavioral changes reported above.

The most impressive effects in this study were observed

on DRL-20 behavior after hippocampal ablation. Septal pre-

treatment with 6-OH- dopamine prevented the expected increase

in response rate without affecting the lesion induced de-

crease in reinforcements. This phenomenon has been observed

previously both in rats with hippocampal lesions treated

with chlorpromazine, a major tranquilizer which is most ef-

fective in blocking adrenergic receptors (Van Hartesveldt,

in preparation), and peripherally blinded rats with hippo-

campal ablation (Schmaltz and Isaacson, 1968). These re-

sults have been interpreted to indicate that there are at

least two, more or less independent, factors involved in

producing the DRL deficit after limbic lesions. One is re-

lated to the rate of responding and the other to successful

performance by temporal spacing of responses (Van Hartes-

veldt, in preparation).

The DRL task requires that the subject learn to withold

a response for a defined interval. This observation led to

the initial assumption that performance of this task was

dependent upon an active inhibitory process (Kramer and

Rilling, 1970) such as that mediated through the hippocampal-

septal system (see Introduction). This interpretation is

consistent with the deficits in performance on this task

observed after lesions to either of these structures (see

Fried, 1972 and Rickert et al., 1973 for review). Although

there are other interpretations which could explain the

deficit in efficient responding, it may be the case that

lesion-induced rate increases are due to interruption of

this active behavioral inhibitory process. This could pro-

vide a basis for the observed protection of the response

rate after hippocampal damage, as indicated by the follow-


Noradrenalin has been observed to have an inhibitory

effect on limbic forebrain mechanisms which mediate be-

havioral response suppression (Grossman, 1960; 1964; Stein

and Wise, 1969). For example, Margules (1968) showed that

direct application of norepinephrine to the amygdala produced

a passive avoidance deficit. Serotonin on the other hand

appears to facilitate these mechanisms as shown by a marked

elevation in both flinch and jump shock thresholds and

facilitated learning of a T-maze position-habit reversal

task after local septal injection of serotonin. Intraseptal

injection with cinnanserin (2'-(3-dimethylaminopropylthio)

cinnamanilida hydrocholride), a seratonergic antagonist,

reduced these effects (Persnip and Hamilton, 1973). In

addition there is evidence to suggest that certain types of

behavioral output are dependent upon the maintainance of a

functional balance between these two transmitter systems

(see Stein et al., 1972). Lesions to either the adrenergic

system or serotonergic system produce greater behavioral

deficits than simultaneous destruction of both (Bresler and

Ellison, 1972). Moreover, histochemical studies have in-

dicated that electrolytic or pharmacologically produced

lesions of adrenergic fiber systems can result in increases

in serotonin levels within certain brain areas (Blondeaux

et al., 1973), and sprouting of local serotonergic fibers

(Bjorklund aet al., 1971; 1973; Bjorklund and Stenevi, 1971;

Nygren and Olson, 1971; Hokfelt et al., 1972). Thus,

6-OH-dopamine introduced into the septal area could result

in both a decrease in the functional level of a system

(adrenergic) which antagonizes the response suppression

function of the hippocampal-septal network and an increase

in a system (serotonergic) which augments this function.

This could account in part for the lack of increased respond-

ing after hippocampal lesions in this study. These factors

are consistent with other results mentioned above, that

effects on this task are similar to those seen in rats treat-

ed with chlorpromazine after hippocampal destruction.

A second set of factors that may be considered as pro-

viding a potential basis for the protection of DRL response

rate after hippocampal ablation concerns the denervation in-

duced changes in electrical properties in the affected septal

cells. Partial denervation results in increases in both the

excitability and sensitivity of neurons to stimuli(e.g.,

Sharpless, 1964; 1969). It has been postulated that the

denervation-induced changes in the response characteristics

of neurons (e.g. supersensitivity) may account for the re-

duction over time of some lesion effects such as recovery

from the lateral hypothalamic syndrome (Berger et al., 1971).

Pretreatment with pharmacological agents such as 6-OH-dopa-

mine (Stein et al., 1972; Mandell et al., 1972) which induce

a partial denervation, prior to lesions, has resulted in

the attenuation of some lesion effects. For example, Glick

et al. (1972) and Balagura et al. (1973) showed that pre-

treatment with drugs that have effects similar to 6-OH-dopa-

mine, effectively reduced the recovery period following

lateral hypothalamic damage. It is suggested that the in-

crease in sensitivity produced by this procedure allows the

cells to adapt to a new level of activity. Thus the change

in neural activity following the subsequent lesion is not as

great and the behavioral effects are reduced. In this study,

it is suggested that the excitability of septal cells pro-

duced by drug-induced denervation could be adaptive in a

similar manner and could partially produce the observed


This interpretation is supported by the observation

that hippocampal input is facilitatory to the cells of the

septum (DeFrance et al., 1971; 1972a; 1972b; 1973a; 1973b;

1973c). Moreover, it has been demonstrated that serotonin

terminals in the septum (which, as indicated above, are in-

creased and functionally augmented in the septum by 6-OH-

dopamine injection) are also facilitatory upon the cells of

the septal area (DeFrance et al., 1973d). These functional

effects would be further augmented by the proposed increase

in neural activity.

Taken together, the above considerations suggest that

the lack of increased responding on the DRL task after drug

injections and hippocampal lesions is due to the augmentation

of the influence of the hippocampus on the septum by 1) the

decrease in the adrenergic inhibitory influence on the nor-

mal functioning of the hippocampal-septal behavioral response

suppression system, and the increase in the serotonergic

influence that is facilitatory to this response suppression

system, and 2) the increase in the activity of the septal

cells after 6-OH-dopamine, and the facilitatory effect of

the increased serotonergic input on these cells (both of

which tend to increase the normal excitatory effects of

hippocampal input to the septal cells).

The DRL task was the only behavior measured in this

study which appeared to be directly affected by septal 6-OH-

dopamine treatment. Furthermore, this was the only behavior-

al measure used in which the animal responded in order to

receive a discrete reward for a correct response. In this

experiment, if the rat paused 20 or more seconds before re-

sponding, it received a food reinforcement. On the basis of

studies in which various limbic structures were lesioned, it

has been postulated that successful performance (i.e.,

correct responding) on this task is dependent upon the ani-

mal's ability to receive and utilize response-produced pro-

prioceptive feedback as discriminative cues (e.g., Ellen and

Powell, 1962; Pelligrino and Clapp, 1971; Rickert et al.,

1973). This enables the animal to determine whether or not

an adequate duration of time has passed since the last re-

inforced response (see Fried, 1972). The observation that

subjects with 6-OH-dopamine injected into the septum did not

improve their performance (as compared to the controls)

suggests that adrenergic terminals may be involved in mediat-

ing some aspect of the proposed "response produced feedback."

It has been postulated that the forebrain catecholamine

fiber pathways, especially those composed of noradrenergic

and serotonergic fibers, mediate information concerning

positive and negative reinforcement, respectively (see Stein,

1969, for review). Since there is some evidence that nor-

adrenalin is released during rewarding intercranial electri-

cal stimulation (Stein and Wise, 1967; 1969), it can be pos-

tulated that the adrenergic fiber system may supply informa-

tion that a response has been rewarded, if not concerning

nature of the reward. This could, in effect, be considered

as "reinforcement produced feedback." In this study, de-

struction of the adrenergic input into the septal cells

could reduce the availability of this information. The

activity of these cells would no longer be directly corre-

lated with the reinforcement. The resulting desynchrony

(between firing pattern of septal cells and adrenergically

mediated reward information) could impair the rat's ability

to base its internally timed interval on the occurance of

a reinforcement. However, since other elements of the neural

network (see Introduction) mediating this behavioral measure

are still intact and presumably receiving information provid-

ed by the adrenergic fiber system, one would not necessarily

see a deficit in DRL responding as a result of the drug

injection alone. This is supported by the present result of

maintainence of, but not improvement over, previously ac-

quired performance levels after septal injection.

On the basis of these considerations, subsequent de-

struction of another element of the network mediating the

response, which eliminates other routes whereby the infor-

mation carried by the adrenergic fiber system gains access

into this network, would cause a further decrease in the

correspondence between the time of reinforcement and the

internally timed response. This could result, as is seen in

this study in a deficit in correct responding without neces-

sarily altering the rate of response after hippocampectomy.

It can thus be postulated that the differential effect

observed in this study between rate of responding and correct

responding are due to differential mechanisms by which the

drug injection affects this system. The protective rate

effect arises as a result of the changes in the organization

and functioning of the neural elements and neurochemical

systems within the septum. The decrement in correct respond-

ing after hippocampal lesion results from the elimination

of reinforcement producedfeedback information necessary to

coordinate responses with reinforcement.

Intraseptal injection of 6-OH-dopamine had no apparent

effect on passive avoidance responding. The drug may have

produced changes, however, that would be consistent with this

result. For example, in this study, animals were trained to

enter the passive avoidance shock chamber in order to obtain

a water reward. If, as previously suggested, information

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E4TBFJM7J_LM35PW INGEST_TIME 2012-02-29T16:05:10Z PACKAGE AA00009130_00001