The effects of potential trophic or cytoprotective agents in cholinergically hypofunctional rat brains

MISSING IMAGE

Material Information

Title:
The effects of potential trophic or cytoprotective agents in cholinergically hypofunctional rat brains
Physical Description:
xiii, 163 leaves : ill. ; 29 cm.
Language:
English
Creator:
Sjak-Shie, Nelida Natasha, 1965-
Publication Date:

Subjects

Subjects / Keywords:
Brain -- drug effects   ( mesh )
Muridae   ( mesh )
Nerve Growth Factors -- pharmacology   ( mesh )
Protirelin -- pharmacology   ( mesh )
Department of Pharmacology and Therapeutics thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Pharmacology and Therapeutics -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 141-162).
Statement of Responsibility:
by Nelida Natasha Sjak-Shie.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001664867
notis - AHX6642
oclc - 25506638
System ID:
AA00009069:00001

Full Text












THE EFFECTS OF POTENTIAL TROPHIC OR CYTOPROTECTIVE AGENTS IN
CHOLINERGICALLY HYPOFUNCTIONAL RAT BRAINS















BY


NELIDA NATASHA SJAK-SHIE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1991































This dissertation is dedicated to my parents Michel and Mena,

and to my sister, Ilonka.















Sometimes it takes courage.........


To dream the impossible dream
To fight the unbeatoabe foe
To bear with unbearable sorrow
To ride where the brave dare not go
To right the unrightable wrong
To love pure and chaste from afar
To try when your arms are too weary
To reach the unreachable star

This is my quest
To follow that star
No matter how hopeless
No matter how far
To fight for the right
Without question or pause
To be wiping to march into helt for a heavenly cause

And I know
If I only be true
To this glorious quest
That my heart
Will ie peaceful and calm
When 1 am laid to my rest

And the world
Wi1 be better for this
That one man
Scorned and covered with scars
Still strolled
With his last ounce of courage
To reach
The unreachable star.















ACKNOWLEDGEMENTS


I wish to sincerely thank my doctoral advisor Dr. Edwin

Meyer for the contributions he has made to my education and

for his encouragement of my professional goals. I

respectfully thank the members of my thesis committee, Dr.

Fulton Crews, Dr. Stephen Baker, Dr. Thomas Rowe, and Dr.

William Millard, for their thoughtful suggestions and

enthusiasm for this work. I also wish to thank Drs. Gary

Arendash, Marieta Heaton, and Jennifer Poulakos for their

technical assistance. Many thanks go to my fellow graduate

students (past and present), especially Walter Folger, Daniel

Danso, Fan Xie, and Sukanya Kantawathana, for their

invaluable friendship and encouragement. I also deeply thank

Drs. Richard Lottenberg, James Deyrup, and Allen Neims for

their confidence in me and for helping me reach "the

unreachable star."















TABLE OF CONTENTS

ACKNOWLEDGEMENTS .......................................... iv

LIST OF TABLES ............................................ vii

LIST OF FIGURES ........................................... viii

ABSTRACT .................................................. xii


CHAPTER 1
INTRODUCTION ............................................ 1
Background .......................................... 1

Specific Aims ........................................ 25


CHAPTER 2
THE UNILATERAL PARTIAL FIMBRIA-LESION MODEL ............. 27
Introduction ......................................... 27

Materials and Methods ................................ 28
Chemicals and Reagents ............................ 28
Animals ........................................... 29
Unilateral Partial Fimbria-Lesions ................ 29
The Isolation and Purification of Nerve Growth
Factor .......................................... 30
Trophic Activity Assay ............................ 31
Passive Avoidance Behavior Testing ................ 32
Determination of CAT Activity ..................... 33
Scintillation Counting ............................ 33
Statistics ........................................ 33

Results .............................................. 34
Purification of 7S NGF from Mouse Submaxillary
Glands .......................................... 34
The Effects of NGF and Thyrotropin Releasing
Hormone on Cholinergic Function ................. 35

Discussion .......................................... 36


CHAPTER 3
THE NUCLEUS BASALIS MAGNOCELLULARIS-LESION MODEL........ 51
Introduction ......................................... 51

Materials and Methods ................................ 52









Chemicals and Reagents ............................ 52
Animals ........................................... 53
Nucleus Basalis Magnocellularis-Lesions ........... 53
Determination of CAT Activity ..................... 54
Determination of Nicotine Receptor
Concentration ................................... 55
Determination of Neuropeptide Y Levels ............ 55
RNA Purification and Analysis ..................... 56
Passive Avoidance Behavior Testing ................ 58
Determination of Neuronal Density ................. 59
Scintillation Counting ............................ 61
Statistics ........................................ 61

Results .............................................. 61
The Long-Term Effects of Nucleus Basalis
Lesions ......................................... 61
The Effects of Chronic Cholinergic Treatment of
Rats with Nucleus Basalis Magnocellularis-
Lesions ......................................... 67
The Effects of Chronic Nicotine Treatment of
Rats with Nucleus Basalis Magnocellularis-
Lesions ......................................... 69
The Effects of Chronic Pilocarpine Treatment of
Rats with Nucleus Basalis Magnocellularis-
Lesions ......................................... 73
The Effects of Chronic Nicotine/Pilocarpine Co-
treatment of Rats with Nucleus Basalis-
Magnocellularis Lesions ......................... 74

Discussion .......................................... 76


CHAPTER 4
SUMMARY AND FUTURE EXPERIMENTS .......................... 138

LIST OF REFERENCES ........................................ 141

BIOGRAPHICAL SKETCH ....................................... 163















LIST OF TABLES


Table 1. The long-term behavioral, neurochemical, and
neuropathological effects of bilateral nucleus basalis
magnocellularis-lesions.................................. 134

Table 2. The long-term neurochemical, and
neuropathological effects of unilateral nucleus
basalis magnocellularis-lesions ......................... 135

Table 3. The effects of chronic nicotine and/or
pilocarpine treatment on the long-term behavioral,
neurochemical, and neuropathological changes
associated with bilateral nucleus basalis
magnocellularis-lesions.................................. 136

Table 4. The effects of chronic nicotine treatment on
the long-term neuropathological changes associated
with unilateral nucleus basalis magnocellularis-
lesions .................................................. 137


vii















LIST OF FIGURES

Figure 1. Electrophoretic analysis of the purity of a
substance isolated from mouse submaxillary glands....... 42

Figure 2. Western blot analysis of a substance isolated
from mouse submaxillary glands .......................... 43

Figure 3. Trophic activity of a substance isolated from
mouse submaxillary glands............................... 44

Figure 4. Trophic activity of the isolated NGF. ........... 45

Figure 5. The effects of unilateral partial fimbria-
lesions on hippocampal CAT activity..................... 46

Figure 6. The effects of unilateral partial fimbria-
lesions on passive avoidance behavior................... 47

Figure 7. The effects of NGF and thyrotropin releasing
hormone on hippocampal CAT .............................. 48

Figure 8. The effects of NGF and thyrotropin releasing
hormone on frontal cortex CAT activity.................. 49

Figure 9. The effects of NGF and thyrotropin releasing
hormone on septal CAT activity.......................... 50

Figure 10. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on passive avoidance
behavior ................................................ 98

Figure 11. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on frontal cortex CAT
activity ................................................ 99

Figure 12. The long-term effects of unilateral nucleus
basalis magnocellularis-lesions on frontal cortex CAT
activity ................................................ 100

Figure 13. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on occipital cortex
neuropeptide Y levels................................... 101

Figure 14. The long-term effects of unilateral nucleus
basalis magnocellularis-lesions on occipital cortex
neuropeptide Y levels................................... 102


viii









Figure 15. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on parietal cortex
neuropeptide Y-encoding mRNA levels..................... 103

Figure 16. The long-term effects of unilateral nucleus
basalis magnocellularis-lesions on parietal cortex
neuropeptide Y-encoding mRNA levels..................... 104

Figure 17. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on frontal cortex
nicotine receptor levels ................................ 105

Figure 18. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on parietal cortex
neuronal density in layer 2............................. 106

Figure 19. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on parietal cortex
neuronal density in layer 2............................. 107

Figure 20. The long-term effects of bilateral nucleus
basalis magnocellularis-lesions on parietal cortex
neuronal density in layer 6............................. 108

Figure 21. The long-term effects of unilateral nucleus
basalis magnocellularis-lesions on parietal cortex
neuronal density in layer 2............................. 109

Figure 22. The long-term effects of unilateral nucleus
basalis magnocellularis-lesions on parietal cortex
neuronal size distribution in layer 2................... 110

Figure 23. The long-term effects of unilateral nucleus
basalis magnocellularis-lesions on parietal cortex
glial density in layer 2................................ 111

Figure 24. The effects of nicotine on passive avoidance
behavior of normal unoperated rats...................... 112

Figure 25. The effects of chronic nicotine,
pilocarpine, or combined nicotine/pilocarpine
treatment on the total body weight gain of rats with
bilateral nucleus basalis magnocellularis-lesions....... 113

Figure 26. The effects of chronic nicotine or
pilocarpine treatment on the total body weight gain of
rats with unilateral nucleus basalis magnocellularis-
lesions ................................................. 114

Figure 27. The effects of chronic nicotine treatment on
passive avoidance behavior of rats with bilateral
nucleus basalis magnocellularis-lesions................. 115









Figure 28. The effects of chronic nicotine treatment on
occipital cortex neuropeptide Y levels in rats with
bilateral nucleus basalis magnocellularis-lesions....... 116

Figure 29. The effects of chronic nicotine treatment on
parietal cortex neuropeptide Y-encoding mRNA levels in
rats with bilateral nucleus basalis magnocellularis-
lesions ................................................. 117

Figure 30. The effects of chronic nicotine treatment on
frontal cortex nicotine receptor levels in rats with
bilateral nucleus basalis magnocellularis-lesions....... 118

Figure 31. The effects of chronic nicotine treatment on
parietal cortex neuronal density in layer 2 of rats
with bilateral nucleus basalis magnocellularis-
lesions ................................................. 119

Figure 32. The effects of chronic nicotine treatment on
parietal cortex neuronal density in layer 2 of rats
with bilateral nucleus basalis magnocellularis-
lesions ................................................. 120

Figure 33. The effects of chronic nicotine treatment on
parietal cortex neuronal density in layer 6 of rats
with bilateral nucleus basalis magnocellularis-
lesions ................................................. 121

Figure 34. The effects of chronic nicotine treatment on
parietal cortex neuronal density in layer 2 of rats
with unilateral nucleus basalis magnocellularis-
lesions ................................................. 122

Figure 35. The effects of chronic nicotine treatment on
the parietal cortex neuronal size distribution in
layer 2 of rats with unilateral nucleus basalis
magnocellularis-lesions................................. 123

Figure 36. The effects of chronic nicotine treatment on
parietal cortex glial density in layer 2 of rats with
unilateral nucleus basalis magnocellularis-lesions...... 124

Figure 37. The effects of chronic pilocarpine treatment
on occipital cortex neuropeptide Y levels in rats with
bilateral nucleus basalis magnocellularis-lesions....... 125

Figure 38. The effects of chronic pilocarpine treatment
on parietal cortex neuropeptide Y-encoding mRNA levels
in rats with bilateral nucleus basalis
magnocellularis-lesions................................. 126

Figure 39. The effects of chronic pilocarpine treatment
on frontal cortex nicotine receptor levels in rats









with bilateral nucleus basalis magnocellularis-
lesions ................................................. 127

Figure 40. The effects of chronic pilocarpine treatment
on parietal cortex neuronal density in layer 2 of rats
with bilateral nucleus basalis magnocellularis-
lesions ................................................. 128

Figure 41. The effects of chronic nicotine/pilocarpine
co-treatment on occipital cortex neuropeptide Y levels
in rats with bilateral nucleus basalis
magnocellularis-lesions................................. 129

Figure 42. The effects of chronic nicotine/pilocarpine
co-treatment on parietal cortex neuropeptide Y-
encoding mRNA levels in rats with bilateral nucleus
basalis magnocellularis-lesions ......................... 130

Figure 43. The effects of chronic nicotine/pilocarpine
co-treatment on frontal cortex nicotine receptor
levels in rats with bilateral nucleus basalis
magnocellularis-lesions................................. 131

Figure 44. The effects of chronic nicotine/pilocarpine
co-treatment on parietal cortex neuronal density in
layer 2 of rats with bilateral nucleus basalis
magnocellularis-lesions................................. 132

Figure 45. The effects of chronic nicotine/pilocarpine
co-treatment on parietal cortex neuronal density in
layer 6 of rats with bilateral nucleus basalis
magnocellularis-lesions................................. 133















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


THE EFFECTS OF POTENTIAL TROPHIC OR CYTOPROTECTIVE AGENTS IN
CHOLINERGICALLY HYPOFUNCTIONAL RAT BRAINS

By

Nelida Natasha Sjak-Shie

May, 1991




Chairman: Edwin M. Meyer
Major Department: Pharmacology and Therapeutics


Lesions of the cholinergic projections from the nucleus

basalis magnocellularis to the neocortex or from the septum

to the hippocampus are important models for studying the

effects of potential trophic or cytoprotective agents on

cholinergic hypofunction. We investigated the abilities of

nerve growth factor (NGF) and thyrotropin releasing hormone

to modulate cholinergic activity of the septal-hippocampal

pathway, using choline acetyltransferase (CAT) activity as a

marker. Continuous administration of NGF increased CAT

activity in the unlesioned hippocampal cholinergic neurons.

Though thyrotropin releasing hormone itself had no effect on

hippocampal CAT activity, it blocked the NGF-induced

increase, and reduced neocortical CAT activity. After

characterizing the long-term neuropathological,


xii










neurochemical, and behavioral changes associated with lesions

of the nucleus basalis-neocortex cholinergic pathway, we

investigated how attempts to re-establish this cholinergic

pathway pharmacologically would affect the lesion-induced

changes. Lesioned rats suffered from a loss of neocortical

neurons, a transient deficit in passive avoidance behavior,

and displayed transient elevations in neocortical high-

affinity [3H-acetyl] choline binding to nicotine receptors.

Chronic nicotine treatment (from 5 to 8 months post-

lesioning) appeared to ameliorate the lesion-induced cell

loss and elevate nicotine receptor binding, suggesting that

nicotine exerted a cytoprotective influence on those neurons

expressing nicotine receptors. Similar pilocarpine treatment

exerted no cytoprotective influence on neocortical neurons,

and reduced nicotine receptor binding in sham-operated rats.

Bilateral nucleus basalis-lesions elevated levels of

neocortical neuropeptide Y while unilateral lesions

transiently reduced neuropeptide Y levels. The changes in

neuropeptide Y levels appeared not to be preceded by

elevations in neuropeptide Y-encoding mRNA. Cholinergic

hypofunction is associated with changes in a variety of

behavioral, neurochemical and neuropathological parameters,

some of which may be modulated pharmacologically as

demonstrated by these studies.


xiii















CHAPTER 1
INTRODUCTION


Background


Alzheimer's disease is a common neurodegenerative

disorder, characterized clinically by a progressive loss of a

wide range of cognitive abilities including memory,

attention, language, praxis, and perception (Cummings et al.,

1980). Once Alzheimer's disease begins to progress, the

projected life span of the individual is reduced

substantially and death usually occurs within 4 to 8 years

(Bartus, 1986). One of the earliest and most pronounced

neurochemical deficits consistently associated with

Alzheimer's disease is a marked cortical cholinergic

hypofunction, primarily in the neocortex and hippocampus

(Davies and Maloney, 1976; Perry et al., 1977; Whitehouse et

al., 1982). This cholinergic hypofunction results in a

reduction of several pre-synaptic cholinergic markers

including CAT activity (the enzyme responsible for

acetylcholine synthesis), acetylcholine release and turnover,

acetylcholinesterase activity, as well as high affinity

choline uptake. These early cholinergic abnormalities appear

to be due to a loss of neurons originating from the basal

forebrain (Whitehouse et al., 1982). Specifically, the










neocortical and hippocampal cholinergic deficits seem to be

due to progressive degeneration or atrophy of cholinergic

neurons located in the nucleus basalis of Meynert and the

medial septum respectively (Bartus et al., 1982; Coyle et

al., 1983; Rinne et al., 1987).

Later stages of Alzheimer's disease are characterized by

severe cerebral atrophy due to a substantial loss of neurons,

primarily in the neocortex and hippocampus (Ball et al.,

1985; Terry et al., 1981). In addition to cell loss and cell

shrinkage, there also is a high incidence of neurofibrillary

tangles and senile plaques in these brain regions (Ball,

1977; Brun and Englund, 1981). It appears that degenerative

changes in cortical cholinergic terminals are involved in the

pathogenesis of neuritic plaques. And there is a positive

correlation between the extent of cell death of the

cholinergic neurons in the nucleus basalis of Meynert, the

degree of the dementia of Alzheimer's patients, and the

density of cortical and hippocampal plaques and

neurofibrillary tangles (Blessed et al., 1968; Perry et al.,

1978; Wilcock and Esiri, 1982).

The more advanced stages of Alzheimer's disease are also

characterized by abnormalities in several neurotransmitter

systems, including the noradrenergic, peptidergic

somatostatinn and corticotropin releasing factor), and

serotoninergic systems (Gottfries et al., 1983; Hardy et al.,

1985). More recently, deficits in other peptidergic

neurotransmitter systems, including those using neuropeptide










Y, have been reported (Ferrier et al., 1983). The greatest

decreases in neuropeptide Y-like immunoreactivity occur in

the neocortex, the amygdala, and the hippocampus (Beal et

al., 1986a). This regional distribution parallels the

incidence of histopathological abnormalities such as neuritic

plaques and neurofibrillary tangles (Chan-Palay et al.,

1985a; Chan-Palay et al., 1985b). Neuropeptide Y

immunoreactive neurons are also abnormally swollen and

bulbous in shape (Chan-Palay et al., 1985a; Kowall and Beal,

1987; Nakamura and Vincent, 1986), and their axons contribute

to neuritic plaque formation (Chan-Palay et al., 1985a; Chan-

Palay et al., 1985b; Dawbarn and Emson, 1985). Furthermore,

a recent study shows that neuropeptide Y receptor binding

sites are significantly reduced in the temporal cortex and

hippocampus of Alzheimer's patients (Martel et al., 1990).

In normal adults neuropeptide Y-immunoreactivity is

found in the central as well as the peripheral nervous

system. Neuropeptide Y is the most abundant peptide found in

the central nervous system. The cerebral cortex, limbic

system, hypothalamus, striatum, brainstem, and the dorsal

horn of the spinal cord all display neuropeptide Y-

immunoreacitvity. Several classical and putative

neurotransmitters have been found co-localized with

neuropeptide Y. In the locus coeruleus, for example,

neuropeptide Y co-localizes with norepinepherine or with

norepinepherine and Fmrfamide. In addition to

catecholamines, neuropeptide Y is found co-localized with










gamma-aminobutyric acid (GABA) or somatostatin (Everitt et

al., 1984; Hendry et al., 1984) in other parts of the brain.

In the neocortex approximately 80% of neuropeptide Y

containing neurons also contain somatostatin. Neuropeptide Y

containing neurons account for about 2% of all cortical

neurons. Neocortical neuropeptide Y containing neurons are

usually circuit, intrinsic neurons that have multipolar,

bipolar, or stellate shapes (Beal et al., 1987).

The progressive nature of Alzheimer's disease, the lack

of conclusive diagnostic tools, the inability to determine

the degree of neurochemical and neuropathological deficits

until autopsy, and the variability in the extent of the

abnormalities commonly observed in Alzheimer's patients make

it difficult to determine whether there is a causal

relationship between the initial loss of basal forebrain

cholinergic neurons and the subsequent behavioral,

neurochemical, and neuropathological changes associated with

more advanced stages of Alzheimer's disease. Therefore,

attempts have been made to develop appropriate animal models

to study the consequences of cholinergic hypofunction.

Cholinergic abnormalities similar to those associated

with Alzheimer's disease are induced in the neocortex and

hippocampus of rats by respectively lesioning the nucleus

basalis magnocellularis (the rodent equivalent to the nucleus

basalis of Meynert; Fibiger, 1982; Lehmann et al., 1980;

Mesulam et al., 1983; Wenk et al., 1984) and the septal-

hippocampal pathway (Gage et al., 1986; Hefti et al., 1984;





5



Meibach and Seigel, 1977; Segal and Landis, 1979). These

animal models are important tools to study whether decreases

in cholinergic transmission in the central nervous system can

precipitate the abnormalities associated with more advanced

stages of Alzheimer's disease. Furthermore, these models

allow an examination of the ability of potentially trophic or

cytoprotective agents to augment cholinergic transmission,

and the trans-synaptic effects of this augmentation in

cholinergic transmission.

Unilateral partial fimbria-lesions interrupt the

ascending cholinergic pathway that originates in the septum

and terminates in the hippocampus. These lesions

characteristically reduce hippocampal CAT activity

anterogradely and decrease the number of Nissl-staining or

diisopropylfluorophosphate-positive septal cholinergic cell

bodies retrogradely (Daitz and Powell, 1954; Gage et al.,

1986). This lesion model is particularly useful for studying

the potential of trophic neuropeptides to stimulate a

hypofunctional cholinergic system, because all hippocampal

cholinergic activity originates from one source, the septum

(Meibach and Siegel, 1977; Segal and Landis, 1979). Changes

in hippocampal cholinergic parameters can therefore be

attributed to drug-induced changes in the activity of the

septal cholinergic neurons. Using this lesion model we

propose to characterize the trophic actions of thyrotropin

releasing hormone and NGF.










NGF was initially characterized as a selective

neurotrophic agent for sympathetic and sensory peripheral

neurons (Greene and Shooter, 1980; Thoenen and Barde, 1980;

Varon and Adler, 1980; Vinores and Guroff, 1980). However,

several findings in recent years suggest that NGF also exerts

a trophic influence on ascending cholinergic neurons of the

rat basal forebrain. One of the first studies suggesting a

central action of NGF found that destruction of the

cholinergic input to the hippocampus results in an ingrowth

of peripheral sympathetic fibers that matches the previous

distribution of cholinergic terminals in the hippocampus

(Crutcher et al., 1979; Loy and Moore, 1977; Stenevi and

Bjorklund, 1978). It was suggested that the signal

attracting these sympathetic fibers to the hippocampus was

NGF since peripheral sympathetic neurons were known to

respond to NGF.

Subsequent studies showed that NGF injected into the

hippocampus or neocortex is retrogradely transported to the

cholinergic cell bodies in the septum or the nucleus basalis

magnocellularis, respectively, indicating the presence of NGF

receptors on or near these cholinergic terminals (Schwab et

al., 1979; Seiler and Schwab, 1984). Receptor

autoradiography studies confirmed the presence of NGF

receptors on basal forebrain cholinergic neurons (Richardson

et al., 1986). In vitro studies using cultured rat fetal

forebrain cholinergic neurons show that NGF stimulates the









expression of CAT (Hefti et al., 1985; Honegger and Lenoir,

1982).

Recent in vivo studies show that NGF exerts a trophic

influence on central cholinergic neurons. NGF promotes the

survival of septal cholinergic neurons following partial

fimbria-lesions (Montero and Hefti, 1988; Otto et al., 1987;

Williams et al., 1986). In addition to preventing the

lesion-induced retrograde degeneration of septal cholinergic

neurons, NGF also has a cytoprotective effect on lesioned

non-cholinergic septal neurons (Williams et al. 1986). The

ability of NGF to prevent the septal neuronal death induced

by partial fimbria-lesions is presently well characterized.

However, few studies have assessed NGF's ability to modulate

pre-synaptic cholinergic activity in vivo. For example, it

is not known whether the neurons not lesioned by these knife-

cuts express higher than normal levels of CAT activity, which

would be expected if the septal-hippocampal pathway becomes

more active. We propose to examine this question in rats

with unilateral partial fimbria-lesions using CAT as a marker

for pre-synaptic cholinergic activity.

Besides its hormonal action at the hypothalamic

pituatary level (Schally, 1978), thyrotropin releasing

hormone has many extra-hypothalamic functions. Its

involvement has been suggested in different aspects of

autonomic functions, neuronal excitability, interactions with

other neurotransmitter systems, and behavior (Morley et al.,

1981; Sharif, 1985; Yarborough, 1979). Indirect experimental










evidence supports the hypothesis that one of the functions of

thyrotropin releasing hormone in the central nervous system

is to control cholinergic systems (Yarborough, 1983).

Recently, thyrotropin releasing hormone has been used to

ameliorate some of the cognitive deficits associated with

Alzheimer's disease (Lampe et al., 1991).

Immunohistochemical and radioimmunochemical studies into

the regional distribution of thyrotropin releasing hormone in

the central nervous system show substantial thyrotropin

releasing hormone immunoreactivity in the septum (Hokfelt et

al., 1975; Winokur and Utiger, 1974). Autoradiographic

studies reveal moderate labeling in the septum and cerebral

cortex (Manaker et al., 1985; Ogawa et al., 1981). Using a

thyrotropin releasing hormone analog, Simasko and Horita

(1982) observed high levels of binding in the septal regions

and low levels in the cerebral cortex. Thyrotropin releasing

hormone receptors appear to be located primarily on

cholinergic cell bodies (Simasko and Horita, 1984).

One of the first indications that thyrotropin releasing

hormone could potentially modulate central cholinergic

transmission was the observation that thyrotropin releasing

hormone mediates its analeptic effects through stimulation of

central cholinergic pathways (Brunello and Cheney, 1981;

Yarborough, 1979). Subsequent studies with thyrotropin

releasing hormone and MK-771, a thyrotropin releasing hormone

analog, confirmed thyrotropin releasing hormone's ability to

modulate cholinergic activity. For example, MK-771 enhances









the incorporation of choline into acetylcholine and

stimulates the release of acetylcholine (Yarborough et al.,

1978; Yarborough, 1983). Thyrotropin releasing hormone

itself increases cholinergic activity in rat spinal cord

(Schmidt-Archert et al., 1984), in primary neuronal tissue

cultures (Sjak-Shie et al., 1989), and in septal cholinergic

cultures to a similar extent as NGF (Dr. Franz Hefti,

Department of Neurology, University of Miami, personal

communication).

Whether the interaction between thyrotropin releasing

hormone and cholinergic systems is mediated via thyrotropin

releasing hormone receptors or, as some studies suggest, via

muscarinic receptors is presently unresolved. Some

investigators observe an antagonism of the analeptic effects

of thyrotropin releasing hormone by atropine (Breese et al.,

1975), whereas others do not (Santori et al., 1981).

However, the previous observations suggest that thyrotropin

releasing hormone may be able to stimulate central

cholinergic function not only in vitro, but also in vivo. We

therefore propose to examine the ability of thyrotropin

releasing hormone to enhance cholinergic activity in vivo in

rats with unilateral partial fimbria-lesions using CAT as a

marker.

Neurotrophic studies using the nucleus basalis

magnocellularis-lesion model of cholinergic hypofunction are

complicated by the presence of intrinsic neocortical

cholinergic neurons in the rat (Fibiger, 1982; Johnston et









al., 1981; Smith, 1988). However, some of the long-term

trans-synaptic changes induced by nucleus basalis

magnocellularis-lesions are known, whereas the long-term

trans-synaptic hippocampal changes associated with partial

fimbria-lesions are virtually unknown. Though that knowledge

is presently limited to two time points, 2 months and 14

months post-lesioning, it allows examination of the

functional significance of a trophic or cytoprotective

influence through assessment of trans-synaptic changes.

Bilateral nucleus basalis magnocellularis-lesions are

characterized by an initial reduction (2 months post-

lesioning) in several neocortical pre-synaptic cholinergic

markers including CAT activity, acetylcholine synthesis and

release, acetylcholinesterase activity, as well as high

affinity choline uptake. Neocortical low affinity choline

uptake found in all cells remains unchanged (Bartus et al.

1982; Strong et al., 1980; Watson et al., 1985). The

greatest loss in cholinergic markers occurs in the more

anterior half of the neocortex. Little or no effect on

cholinergic markers is observed in the posterior neocortex,

which receives virtually no cholinergic input from the

nucleus basalis magnocellularis (Bartus et al., 1986). These

cholinergic changes in rats with nucleus basalis-lesions are

similar to those observed in Alzheimer's disease.

Nucleus basalis-lesions also induce memory-related

deficits in a variety of learning paradigms including single

trial passive avoidance behavior, two way active avoidance









behavior, Lashley III maze performance, and reinforced T-maze

alternation (Altman et al., 1985; Flicker et al., 1983; Lever

and Nicholson, 1986; Salamone et al., 1984). These lesions

significantly impair performance on memory tasks, especially

in situations that require the use of recent, short-term

information (Bartus et al., 1986).

Fourteen months post-lesioning, the levels of all

neocortical pre-synaptic cholinergic markers are still

reduced to a similar extent as at 2 months post-lesioning.

However, these lesioned rats no longer display any behavioral

abnormalities (Arendash et al., 1987; Bartus et al., 1985).

Furthermore, several additional neocortical neurochemical

changes are observed 14 months post-lesioning. For example,

levels of neuropeptide Y are substantially elevated in the

parietal cortex, and small but statistically significant

decreases in parietal cortical norepinepherine levels are

present. Other neocortical monoamine levels, including

dopamine, serotonin, and their major metabolites remain

unaffected by bilateral nucleus basalis-lesions up to 14

months post-lesioning (Arendash et al., 1987). Recent

studies show that in addition to the dramatic neocortical

increases in neuropeptide Y there are concurrent robust

increases in neuropeptide Y immunoreactive fiber innervation

of the rat parietal cortex (Gaykema et al., 1989).

The lesion-induced increases in parietal cortical

neuropeptide Y levels appear to coincide with the behavioral

recovery. This parallel may suggest that the improvement in









behavioral performance is in part mediated by the elevation

in neuropeptide Y levels. The hypothesis of a neuropeptide

Y-mediated recovery of cognitive function is supported by

pharmacologic studies which suggest that neuropeptide Y

transmission in the cerebrum is involved in several

behaviors, including memory and cognition (Flood et al.,

1987).

Unlike rats with nucleus basalis-lesions, Alzheimer's

patients never recover from their memory-related behavior

deficits, and their neocortical neuropeptide Y levels never

increase in response to the loss of cholinergic transmission.

Neocortical neuropeptide Y levels may actually be decreased

instead of increased in Alzheimer's disease, and neuropeptide

Y immunoreactive neurons appear abnormal (Chan-Palay et al.,

1985a; Beal et al., 1986a; Nakamura and Vincent, 1986). It

is thus conceivable that the rat may compensate for

cholinergic hypofunction by increasing neocortical

neuropeptide Y fiber innervation and neuropeptide Y

transmitter levels, in a manner that humans cannot.

Neuropeptide Y is a 36 amino acid peptide that was first

isolated from porcine brain by Tatemoto (1982), using a

chemical approach for identification of peptides possessing a

C-terminal amide residue. The peptide is named neuropeptide

Y since it was first discovered in the brain and since it

possesses a high number of tyrosine residues, including one

at the N-terminal. Neuropeptide Y is a member of a family of

structurally related peptides, including polypeptide YY and










pancreatic polypeptide, which share a 69% and 50% amino acid

homology with neuropeptide Y, respectively. Human and rat

neuropeptide Y differ from porcine neuropeptide Y by a single

amino acid substitution (Corder et al., 1985; Minth et al.,

1984). Neuropeptide Y is widely distributed in the mammalian

central nervous system where it is thought to act as a

neurotransmitter or neuromodulator.

The increased levels of parietal cortex neuropeptide Y

may be due to increased synthesis or decreased breakdown of

the peptide. The synthesis of peptidergic neurotransmitters

generally involves several steps including transcription,

translation, and post-translational modification. Human pre-

pro-neuropeptide Y contains a signal peptide sequence, the 36

amino acid neuropeptide Y sequence, and a 30 residue COOH-

terminal peptide sequence. The neuropeptide Y gene consists

of 4 exons separated by 3 introns. All exon-intron junctions

conform to the "GT-AG" rule. A TATA-like element is located

25 basepairs upstream of the start of transcription and

probably serves as promoter. However, the neuropeptide Y

gene does not have a CAAT box that strictly adheres to the

canonical sequence (Allen et al., 1987; Larhammer et al.,

1987; Minth et al., 1986).

Knowledge regarding the regulation of neuropeptide Y

gene expression by hormones, neurotransmitters, and second-

messenger systems is currently limited. Recent studies

indicate that neuropeptide Y-encoding mRNA synthesis is

elevated by treatments that elevate cyclic AMP (Higuchi and










Sabol, 1987; Higuchi et al., 1988). The mechanism involved

in the cyclic AMP stimulation of neuropeptide Y gene

expression is not yet known. However, cyclic AMP regulation

of the expression of other genes including the somatostatin,

prolactin, and urokinase genes is better characterized

(Nagamine and Reich, 1985). Studies using the somatostatin

gene revealed that a cyclic AMP responsive sequence,

(T/G)ACGTCAG, upstream from the initiation site of

transcription can confer cyclic AMP sensitivity on a promoter

(Montminy et al., 1986a; Montminy et al., 1986b). Careful

examination of the rat neuropeptide Y gene reveals the

presence of a sequence, GGAGTCAC, that resembles the cyclic

AMP consensus sequence (Higuchi et al., 1988). It is

therefore possible that expression of the neuropeptide Y gene

may be modulated by the administration of agents that can

affect the endogenous levels of cyclic AMP.

Cerebral muscarinic receptors coupled to adenylate

cyclase can inhibit this enzyme (Ferrendelli et al., 1970).

Therefore, bilateral nucleus basalis magnocellularis-lesions

may increase adenylate cyclase activity by decreasing the

inhibitory muscarinic transmission potentially mediated via

pre-synaptic M2 receptors. Increased cyclase activity could

subsequently lead to increased transcription of the

neuropeptide Y gene resulting in increased levels of

neuropeptide Y-encoding mRNA, and ultimately neuropeptide Y

itself. Since the increases in neuropeptide Y may in part

account for the memory-related recovery, it is important to










examine if nucleus basalis magnocellularis-lesions induce

these large neuropeptide Y elevations by enhancing

transcription of its gene.

We propose to compare the time course of behavioral

recovery with that of the neuropeptide Y elevations, and

neuropeptide Y-encoding mRNA changes in the neocortex.

Furthermore, we propose to examine whether muscarinic or

nicotinic agonist treatment modulates the expression of

neuropeptide Y and neuropeptide Y-encoding mRNA. These

studies will enhance our knowledge concerning the regulation

of neuropeptide Y turnover. In addition, they may perhaps

lead to the development of new treatment strategies that can

alleviate some of the memory-related deficits associated with

Alzheimer's disease, potentially by increasing neocortical

neuropeptide Y levels.

Rats with bilateral nucleus basalis-lesions also display

significant neocortical cell loss 14 months post-lesioning

(Arendash et al., 1987; Arendash et al., 1989). These

lesions induce substantial neuronal loss in layers 2, 3, and

6 of the frontal and parietal cortex, with particularly

marked cell death in the parietal cortex. Rats with

unilateral nucleus basalis magnocellularis-lesions suffer

similar neuronal losses on the lesioned side, 14 months post-

lesioning. Fiber staining in the frontal and parietal cortex

of lesioned rats is lighter than that of sham-operated

controls and the brain often has a "spongy" appearance (with

vacuolation of tissue), probably due to the cell loss in









these areas. Fiber staining within cortical areas that do

not receive cholinergic projections from the nucleus basalis

magnocellularis, such as the cingulate cortex, pyriform

cortex, and posterior neocortex, is not changed.

One hypothesis is that the lesion-induced, trans-

synaptic cell loss in the neocortex is due to the loss of a

critical modulatory influence on these neocortical neurons

previously provided directly or indirectly by the cholinergic

projections from the nucleus basalis magnocellularis. Since

the trans-synaptic neuronal atrophy in Alzheimer's disease

may occur via a similar mechanism, it seems important to

characterize the chronology of the lesion-induced neocortical

neuropathology further.

Cholinergic neurons in the basal forebrain are

distributed across several classically defined nuclei,

including the nucleus basalis magnocellularis. They are

large (20 to 50 Jm) multipolar neurons with multiple (3 to 8)

primary dendrites, extensive dendritic fields, and frequent

dendritic spines (Semba et al., 1987). Based on

cytoarchitectonic, biochemical, and connectional

considerations, these widely dispersed magnocellular cells in

the basal forebrain appear to form a single continuum (Saper,

1984; Satoh et al., 1983; Schwaber et al., 1987).

Efferents of the caudal pole of this continuum, the

nucleus basalis magnocellularis, are considered to be the

primary source of cholinergic input to the neocortex (Lehmann

et al., 1980). In rats and primates, nucleus basalis









magnocellularis-neurons have terminal fields that are

restricted to a single cytotectonic area, usually less than

1.5 mm in diameter, and they project primarily to the

neocortex of the ipsilateral brain hemisphere (Saper, 1984;

Kristt et al., 1985). Although these cholinergic terminals

are typically found in all cortical layers, layer 5 appears

to be particularly heavily innervated in motor and sensory

cortical areas (Eckenstein et al., 1988; McDonald et al.,

1987).

Cholinergic transmission is mediated by muscarinic or

nicotine receptors. If neocortical neurons are destroyed

subsequent to a loss of cholinergic transmission, there may

be a selective loss of those cholinergic receptors used to

mediate the transmission between the ascending cholinergic

and the neocortical neurons. A likely candidate for this

receptor-subtype is the nicotine receptor since its levels,

unlike muscarinic receptor levels, are consistently and

dramatically reduced in the neocortex of Alzheimer's patients

(Araujo et al., 1988a; Flynn and Mash, 1986; Nordberg and

Winblad, 1986a; Perry et al., 1987; Shimohama et al., 1986;

Whitehouse et al., 1985a; Whitehouse et al., 1986). This

drastic reduction in nicotine receptor levels is also

confirmed by biopsy studies of Alzheimer's patients (DeSarno

et al., 1988; Giacobini et al., 1989).

Though nucleus basalis-lesions have no immediate effect

on neocortical nicotine receptor levels (Schwartz et al.,

1984), it is presently not known whether these lesions will










affect nicotine receptor levels over time. We hypothesize

that reestablishment of the excitatory nicotinic cholinergic

input to the neocortex may prevent the trans-synaptic

neocortical cell loss potentially induced by the lesion-

induced cholinergic hypofunction. Therefore, we propose to

characterize the long-term effects of nucleus basalis-lesions

on neocortical nicotine receptor levels. And we propose to

examine whether chronic nicotine treatment blocks the lesion-

induced neocortical cell loss.

Nicotinic receptors comprise only a small fraction of

the total number of cholinergic receptors in the neocortex

(Phillis and York, 1968). Nonetheless, they may be

disproportionately highly involved in cholinergic

neurotransmission in that brain region because, unlike their

muscarinic receptor counterparts, they are cation-channels,

and they are therefore never spare receptors requiring second

messenger-coupling (Changeux et al., 1984). Nicotinic

receptors do, however, desensitize rapidly in the presence of

some agonists, so that depolarization can be followed by a

period of relative refractoriness (Katz and Thesleff, 1957).

Cholinergic neurotransmission via nicotine receptors

appears to be important for the viability of some cell types.

One of the earliest observations to suggest that nicotinic

transmission may exert trophic influences is the atrophy of

striated muscle following loss of their cholinergic nicotinic

innervation (Birks et al., 1960; Hebb, 1962; Mark et al.,

1983). More recent studies reveal that chronic nicotine










treatment protects against mechanically as well as

neurotoxin-induced degeneration of nigrostriatal dopaminergic

neurons (Janson et al., 1988; Janson et al., 1989), and that

nicotinic antagonists enhance process outgrowth of retinal

ganglion cells in culture (Lipton et al., 1988).

The reduction in nicotine receptor levels characteristic

of Alzheimer's disease is presently well established.

However, initial nicotine receptor binding studies found no

significant changes in nicotine receptor density in

Alzheimer's brains using the classical nicotinic antagonist

alpha-bungarotoxin (Davies and Feisullin, 1981). Alpha-

bungarotoxin binds very tightly to the nicotine receptor in

Torpedo electric organ and muscle and blocks its function

(Heidmann and Changeuax, 1978). The discrepancy between

early studies and more recent findings appears to be due to

the expression of different nicotine receptor subtypes in the

brain, autonomic ganglia, and muscle (Wonnacott, 1990). The

brain apparently contains a large number of high-affinity

receptors that bind nicotinic cholinergic agonists, but not

alpha-bungarotoxin. Alpha-bungarotoxin apparently binds to a

separate class of nicotine receptors in the brain with very

low affinity (Marks and Collins, 1982).

The binding properties of these high-affinity nicotine

receptors have been characterized with [3H] nicotine and [3H]

acetylcholine. These ligands appear to recognize the same

population of nicotine receptors in the rodent brain based on

similarities in their pharmacologic specificity and the










regional distribution of their receptor sites (Marks et al.,

1986; Martino-Barrows and Kellar, 1987). In the rat brain,

dense autoradiographic labelling with [3H] nicotine is

observed in layers II and IV of the cerebral cortex, with

very sparse labelling in the hippocampus (Clarke, 1989).

Autoradiographic studies with [3H] acetylcholine have

virtually identical results (Marks et al., 1986).

There is some variability in the agonist binding

properties reported. Using [3H] nicotine, some investigators

report binding sites of lower affinity in both rat and human

brains (Kd typically > 100 nM) (Larsson and Nordberg, 1985),

in addition to the single high-affinity binding site

observed by others (Kd 1-10 nM) (Marks and Collins, 1982).

Binding studies with [3H] acetylcholine in human brain again

yield conflicting results. Some studies report a single

high-affinity binding site (Whitehouse et al., 1986), while

others find multiple high-affinity sites (Adem et al., 1987).

There appears to be a consensus, however, that in rats [3H]

acetylcholine binds to a single high-affinity nicotine

receptor (Schwartz et al., 1982; Martino-Barrows and Kellar,

1987).

An unusual characteristic of brain nicotine receptors is

their response to chronic nicotine treatment. There is a

surprising increase in the number of nicotine receptors in

the brains of animals treated chronically with nicotine

(Lapchak et al., 1989; Nordberg et al., 1989; Schwartz and

Kellar, 1985). Numbers of nicotine receptors are also higher









in postmortem brains of human smokers than of non-smokers

(Benwell et al., 1988). The effects of nicotine are specific

in that nicotine receptor numbers are elevated, whereas

muscarinic receptor levels are unchanged (Schwartz and

Kellar, 1985).

Initial studies reported that chronic treatment with

irreversible acetylcholinesterase inhibitors decreases the

number of nicotine receptors (Schwartz and Kellar, 1985),

suggesting that the endogenous agonist acetylcholine

regulates its receptor in a more conventional manner.

However, more recent studies show that the reversible

acetylcholinesterase inhibitor, physostigmine, can elevate

levels of nicotine receptors after chronic treatment (Collins

et al., 1990; DeSarno and Giacobini, 1989). Furthermore,

anabasine, (+) anatoxin-a, and cytisine each increase the

number of nicotine receptors in rodent brain (Collins et al.,

1990).

Agonist-induced depolarization-desensitization has been

suggested as the signal that triggers this receptor up-

regulation. However, depolarization desensitization occurs

in central as well as peripheral nicotine receptors.

Contrary to the central effects of nicotinic agonists,

nicotine down-regulates nicotine receptors on muscle

(Ashizawa et al., 1982) and autonomic ganglia (Berg et al.,

1989). Neuronal nicotine receptors are more profoundly

desensitized by nicotine than muscle receptors are (Paton and

Savini, 1968). It has therefore been suggested that the









brain nicotine receptor may be up-regulated selectively by

agonist treatment due to its greater propensity to

desensitize (Wonnacott, 1990).

Given the possible role of desensitization in the

agonist-induced receptor up-regulation, the question remains

whether chronic treatment with a nicotinic agonist will

result in enhanced, diminished, or unchanged responsiveness.

In vitro studies looking at nicotinic modulation of

transmitter release report both increased (Wonnacott et al.,

1990) and decreased (Lapchak et al., 1989) nicotinic

function. Similarly, in vivo behavioral studies suggest

increased (Clarke et al., 1988; Ksir et al., 1987), decreased

(Marks et al., 1985), or unchanged (Warburton et al., 1988)

responsiveness following chronic nicotine administration.

Thus our proposed chronic nicotine treatment of lesioned rats

may elevate nicotinic transmission despite these agonist-

induced changes.

Some controversy also exists concerning the question

whether the central effects of nicotine are mediated via pre-

or post-synaptic nicotine receptors. There are studies that

suggest a direct action of nicotine on cholinergic nerve

terminals (Rowell and Winkler, 1984; Araujo et al., 1988b;

Lapchak et al., 1988). However, in our laboratory nicotinic

agonists do not affect acetylcholine release from rat brain

synaptosomes, suggesting a post-synaptic location of the

nicotine receptors (Meyer et al., 1987). Other investigators

find that very high concentrations of nicotine are necessary










to induce acetylcholine release, and that tetrodotoxin, which

blocks axon potentials, blocks the nicotine-induced

acetylcholine release from guinea pig brian slices (Beani et

al., 1985). Furthermore, no alterations in the number of

neocortical nicotine receptor levels are observed in rats

shortly after lesioning the nucleus basalis magnocellularis

(Schwartz et al., 1984), while the sensitivity of

iontophoretically applied nicotine appears to be increased

(Lamour et al., 1982). These observations also suggest that

nicotine receptors are located post-synaptically in the rat

brain, suggesting that they will remain accessible to

treatment in our lesion model and possibly in early stages of

Alzheimer's disease too.

The long-term effects of nucleus basalis-lesions on

muscarinic receptor levels are presently not known. Studies

regarding the status of muscarinic receptors in Alzheimer's

disease have yielded conflicting results. Depending on the

study muscarinic receptor levels appear to remain unchanged

(Davies and Verth, 1978; DeSarno et al., 1988; Kellar et al.,

1987; Waller et al., 1986; Whitehouse et al., 1985b),

significantly elevated (Danielsson et al., 1988; DeSarno et

al., 1988; Nordberg and Winblad, 1986b; Nordberg et al.,

1983), or reduced (Reinikainen et al., 1987) in Alzheimer's

disease.

These studies were all performed using [3H] 1-

quinuclidinyl-phenyl-4-benzilate, which does not distinguish

between the muscarinic receptor subtypes. The proportion of









pirenzepine-insensitive and sensitive receptors in the human

neocortex has been estimated to be 67% and 33%, respectively

(Vanderheyden et al., 1987). Lesion studies in animals

suggest that the neocortical pirenzepine-insensitive

receptors are predominantly pre-synaptic, inhibitory

autoreceptors (Whitehouse et al., 1984), while the

neocortical pirenzepine-sensitive receptors are primarily

post-synaptic, excitatory receptors (Crews et al., 1986).

Recent studies in Alzheimer's disease report that pre-

synaptic muscarinic receptor sites are slightly reduced in

Alzheimer's disease, but that the post-synaptic receptors

remain functional at least in early stages of the disease

(Mash et al., 1985; Vanderheyden et al., 1987).

The optimal drug to restore muscarinic transmission in

Alzheimer's disease should be a selective agonist at post-

synaptic receptors as well as a selective antagonist at pre-

synaptic receptors. Stimulation of the inhibitory

autoreceptors may decrease acetylcholine release and

potentially further impair the already hypofunctional

cholinergic systems in Alzheimer's disease. Presently no

such drug has been described and one of the best candidates

commercially available appears to be pilocarpine.

Pilocarpine crosses the blood brain barrier and is a slightly

more potent agonist at post-synaptic than pre-synaptic

receptors.

Whether muscarinic transmission is involved in neuronal

viability has not been explored. Interestingly, a recent










study by Ashkenazi et al. (1989) suggests that carbachol may

exert a trophic influence, as reflected by increases in DNA

synthesis, on primary neonatal astroglia cultures via

muscarinic receptors. Since the muscarinic component of

cholinergic input to the neocortex may also be required to

maintain neocortical neurons viable, we propose to test the

ability of chronic pilocarpine and nicotine co-treatment to

block the long-term cell loss caused by nucleus basalis-

lesions.


Specific Aims


This dissertation examines the trophic or cytoprotective

potentials of thyrotropin releasing hormone, NGF, nicotine,

and pilocarpine. The specific aims of this proposal follow:

1) to examine the ability of thyrotropin releasing

hormone and NGF to maintain or elevate hippocampal choline

acetyltransferase activity in rats with unilateral partial

fimbria-lesions;

2) to characterize the long-term neurochemical,

behavioral, and neuropathological changes associated with

nucleus basalis magnocellularis-lesions;

3) to examine the ability of nicotine and/or pilocarpine

to modulate the expression of neuropeptide Y and neuropeptide

Y-encoding mRNA in rats with nucleus basalis magnocellularis-

lesions;





26



4) and to examine the ability of nicotine and/or

pilocarpine to ameliorate the lesion-induced neocortical

neuropathology in rats with nucleus basalis magnocellularis-

lesions.















CHAPTER 2
THE UNILATERAL PARTIAL FIMBRIA-LESION MODEL


Introduction


This chapter delineates the effects of NGF and

thyrotropin releasing hormone on cholinergic neurons

following unilateral partial fimbria-lesions. Partial

transactions of the fimbria result in a partial lesion of the

septal-hippocampal cholinergic pathway. Partial, rather than

complete, lesions were used to monitor changes in cholinergic

activity in the neuronal cell bodies of the septum as well as

the surviving cholinergic terminals in the hippocampus.

Unilateral partial fimbria-lesions typically induce a

reduction in the level of hippocampal CAT activity and a

reduction in the number of septal cholinergic cell bodies

(Daitz and Powell, 1954; Gage et al., 1986).

The ability of NGF to prevent the septal neuronal death

associated with partial fimbria-lesions is well characterized

(Montero and Hefti, 1988; Otto et al., 1987; Williams et al.,

1986). However, few studies have assessed the ability of NGF

or other potentially trophic agents to modulate in vivo

cholinergic neuronal activity in remaining, unlesioned

neurons, such as those that survive in the early stages of

Alzheimer's disease. Several in vitro studies suggest that










thyrotropin releasing hormone may exert a trophic effect on

cholinergic neurons (Schmidt-Archert et al., 1984; Sjak-Shie

et al., 1989; Yarborough, 1983). In the present

investigation we therefore examined the abilities of NGF and

thyrotropin releasing hormone to modulate cholinergic

activity of the septal-hippocampal pathway in rats with

unilateral partial fimbria-lesions, using CAT activity as a

marker.


Materials and Methods



Chemicals and Reagents


Choline chloride and eserine sulfate were purchased from

Sigma Chemicals, St. Louis, MO; thyrotropin releasing factor

from Calbiochem, La Jolla, CA; and male mouse submaxillary

glands from Pel-Freeze, Rogers, AK. The Pierce-protein-

estimation-reagent was purchased from Pierce Chemical

Company, Rockford, IL. [3H-acetyl] coenzyme A (200 mCi/mmol)

was purchased from New England Nuclear, Boston, MA.

Xylazine, ketamine, and penicillin G were purchased from the

University of Florida, Department of Animal Resources. All

other chemicals used were reagent grade.











Animals


Male Sprague Dawley albino rats (350-450 g) were

purchased from the University of Florida Department of Animal

Resources. The rats were professionally maintained in a 22C

environment according to NIH standards and had ad libitum

access to water and food (Purina Rat Chow).


Unilateral Partial Fimbria-Lesions


All rats were anesthetized using a combination of 1

ml/kg ketamine (100 mg/kg, intraperitoneal) and xylazine (20

mg/kg, intramuscular). Prior to lesioning the fimbria, each

animal received a 2 .l intraventricular loading dose of the

same drug that was subsequently administered continuously for

2 weeks using osmotic minipumps (flow rate 0.5 pl/hr; Alzet

model 2002, Alza, Palo Alto, CA). The loading dose was 10

times as concentrated as that administered via the pumps.

All drugs were dissolved in artificial cerebral spinal fluid

containing 0.1 % albumin. Using this diluent, NGF was shown

to maintain its biological activity for 14 days, the duration

of the study (Williams et al., 1986).

With a stereotaxic apparatus and the atlas of Paxinos

and Watson (1986) the injection cannula (23 gauge stainless

steel hypodermic tubing) was positioned at: bregma, 1.5 mm

lateral to the sagittal suture, and 5.0 mm below the surface

of the skull. The incisor bar was set at 0.0 mm. The










septal-hippocampal fibers were partially transected 10

minutes after administration of the loading dose using a

specially modified knife (designed by Dr. Franz Hefti,

Department of Neurology, University of Miami). The knife was

positioned 6.1 mm anterior to the interaural line, 1.5 mm

lateral to the sagittal suture, and 5.0 mm below the

neocortex. It was moved laterally to 5.0 mm lateral to the

sagittal suture and then retrieved.

The infusion cannula (30 gauge stainless steel

hypodermic tubing) and its connecting line were filled with

the experimental drug and attached to pre-filled osmotic

minipumps. It was lowered into the lateral ventricle

ipsilateral to the partial fimbria-lesion at the same

coordinates used to administer the intraventricular loading

dose and cemented in place with cranioplastic cement. The

osmotic minipump was placed subcutaneously in the neck of the

animal and sutured in place. The osmotic minipumps remain

patent at least 17 days post-implantation. Immediately

following surgery, all rats received an intramuscular

injection of 30,000 units penicillin G. They were housed

individually and weighed regularly.


The Isolation and Purification of Nerve Growth Factor


Adult male mouse submaxillary glands were homogenized in

deionized water using a polytron. The homogenate was

centrifuged at 12,000 rpm for 1 hour. The supernatant was










collected and dialyzed overnight against a 0.02 M sodium

phosphate, pH 6.8 buffer. The dialysate was passed over a CM

52 column equilibrated in the same buffer. The eluate was

collected and dialyzed overnight against a 0.25 sodium

phosphate, pH 6.8 buffer. The sodium chloride concentration

of the dialysate was brought to 0.4 M, and 1 to 9

(volume/volume) of 0.05 M sodium acetate, pH 4.0 buffer was

added to the dialysate to induce dissociation of the NGF

complex.

After removal of precipitates the dialysate was passed

over a CM 52 column pre-equilibrated in a 0.4 sodium

chloride, 0.05 M sodium acetate, pH 4.0 buffer. After

several washes PNGF was eluted off the column with a 0.05 M

Tris, 0.4 M sodium chloride, pH 9.0 buffer. The NGF peak

fractions were pooled and dialyzed overnight against a 0.05 M

sodium acetate, pH 5.0 buffer. The concentration of the NGF

preparation was estimated spectrophotometrically (A280 = 1.6

for 1 mg/ml NGF). After subjecting the NGF preparation to

ultrafiltration through a Milipore filter, aliquots were

stored at -800C at 0.5 mg/ml.


Trophic Activity Assay


Dorsal root ganglia from E7 chick embryos were used to

test the trophic potential of the isolated substance.

Cultures were grown in saline or untreated control medium

(negative controls), in control medium supplemented by 50









ng/ml PNGF (positive control; a generous gift of Dr. Eugene

M. Johnson, Department of Pharmacology, Washington University

School of Medicine), and in control medium supplemented by

the isolated substance.

The total number of neurons and the number of neurons

with processes were counted 24 and 48 hours post-plating in

four 0.6 mm2 regions. The number of neurons surviving was

expressed as a percent of those counted initially in each

well, and the number of cells with processes was expressed as

a percent of the total neurons present each day. All

analyses were performed in the laboratory of Dr. Marieta

Heaton (Department of Neuroscience, University of Florida).


Passive Avoidance Behavior Testing


Rats were placed in the lighted compartment of a

standard two compartment passive avoidance apparatus with the

connecting door between the two compartments opened. As soon

as the rat touched the floor of the lighted compartment, the

timer was started. The connecting door was closed when the

rat passed through the door, and the timer was stopped as

soon as he fully entered the dark compartment. Upon entering

this compartment the rat received a 0.8 mAmp footshock for 1

second. Latency to enter the dark compartment was recorded

up to a maximum of 5 minutes during this training period.

The rats were tested in similar fashion 24 hours later, but









no shock was administered upon entering the dark compartment.

Latency for this testing period was also 5 minutes.


Determination of CAT Activity


CAT activity was measured in the frontal cortex,

hippocampus, or septum of each brain hemisphere as described

previously by Meyer et al., 1985. In short, tissue

homogenates were incubated with [3H-acetyl] coenzyme A and

choline. The newly synthesized [3H-acetyl] choline was

separated from acetylcoenzyme A by ion pair extraction, and

the pmol [3H-acetyl] choline per mg protein calculated.

Protein content was estimated after sodium hydroxide

solubilization of the tissue using the Pierce-protein-

estimation-reagent with a bovine serum albumin standard.


Scintillation Counting


All radioactivity measurements were performed via liquid

scintillation spectrophotometry using Liquiscint (National

Diagnostics, Sommerville, NJ) and a Beckman 7100 A

spectrophotometer. The counting efficiency for 3H is

approximately 50% in this spectrophotometer.


Statistics


All data were expressed as the mean standard error of

the mean (S.E.M). The Student t-test was used to analyze

comparisons between two groups only. To assess the equality










of several population means, the one way analysis of variance

test with the Scheffe F-test were used.


Results



Purification of 7S NGF from Mouse Submaxillary Glands


Sodium dodecyl sulfate-polyacrylamide gel

electrophoretic analysis of the substance isolated from mouse

submaxillary glands (6 mg in 12 ml 0.05 M sodium acetate, pH

5.0 buffer) confirmed that the preparation was free of

protein contaminants (Figure 1). The isolate had a molecular

weight similar to that of the 7S mouse NGF generously

supplied by Dr. Eugene M. Johnson. Western blot analysis

showed that it was immunoreactive with a polyclonal antibody

raised in rabbit against mouse NGF (Figure 2).

As shown in Figure 3, further analysis of the isolated

substance revealed that it exerted a trophic action on

cultured dorsal spinal root ganglia. More neurons survived

in the presence of the isolated substance, and more of the

neurons that survived displayed neuronal processes. The

trophic action of the isolated substance was similar to that

of the 7S mouse NGF supplied by Dr. Eugene M. Johnson (data

not shown). Figure 4 shows photographs of representative

cultures treated with either saline or the isolated

substance. These observations confirmed that the substance

isolated was indeed mouse NGF.











The Effects of NGF and Thyrotropin Releasing Hormone on
Cholinergic Function


As shown in Figure 5, following unilateral partial

fimbria-lesions, CAT activity in the hippocampus was reduced

by 52%, indicating that the knife cuts partially transected

the fimbria. This conclusion is supported by the observation

that passive avoidance behavior was not adversely affected by

these lesions, as shown in Figure 6. Complete bilateral

fimbria-lesions as well as complete unilateral fimbria-

lesions have been reported to interfere with memory-related

behaviors (Feldon et al., 1985; Gage et al., 1983).

Unilateral partial fimbria-lesions still reduced

hippocampal CAT activity in all rats after two weeks of

continuous intraventricular administration of NGF (11 gg/ml),

thyrotropin releasing hormone (100 Jg/ml), both drugs, or

artificial cerebral spinal fluid (Figure 7). However, the

reduction in hippocampal CAT activity was significantly less

after treatment with NGF alone. Treatment with thyrotropin

releasing hormone alone did not affect this hippocampal

cholinergic marker. However, when thyrotropin releasing

hormone was co-administered with NGF, it appeared to block

the NGF induced increase in hippocampal CAT activity. For

all treatment groups the unlesioned control values for

hippocampal CAT activity were essentially similar at the

p<0.05 level with one way analysis of variance (ranging from

9.3 1.1 to 11.5 1.2 nmol/mg protein).










Surprisingly, thyrotropin releasing hormone treatment

(alone or with NGF) also reduced CAT activity in the frontal

cortex of these animals by almost 50% compared to rats

treated with artificial cerebral spinal fluid (Figure 8).

Furthermore, as shown in Figure 9, septal CAT activity was

only significantly elevated in the lesioned brain hemisphere

after co-administration of NGF and thyrotropin releasing

hormone. There was no significant interaction between the

main effects (lesioned /unlesioned versus drug treatments)

with the two variable one way analysis of variance test. For

all treatment groups the unlesioned control values for septal

CAT activity were similar at the p<0.05 level with one way

analysis of variance.


Discussion


The findings of this study indicate that continuous

administration of NGF can increase the level of cholinergic

activity in a lesion model similar to that used to

demonstrate its ability to increase survival of lesioned

cholinergic neurons. Thus, not only does NGF protect the

lesioned cells from dying, it also increases levels of CAT in

the unlesioned cholinergic neurons. This elevation in CAT

activity is observed in the hippocampus, where all CAT

activity resides in cholinergic nerve terminals from un-cut

axons. These results are supported by the in vitro

experiments of Neff and Friedman (1990), who report that NGF










increases CAT activity in existing cholinergic neurons but

does not affect cholinergic cell survival in cultured rat

septal neurons.

The effect of NGF on neuronal cholinergic activity

appears to be specific for the injured septal-hippocampal

pathway inasmuch as cholinergic neurons in the uninjured

nucleus basalis magnocellularis-frontal cortex pathway do not

display increased cholinergic activity, even though both

pathways reportedly appear to express NGF receptors

(Richardson et al., 1986; Taniuchi et al., 1986; Springer et

al., 1987). This observation is in agreement with that of

Hefti et al. (1984) who reported earlier that exogenous NGF

does not elevate CAT activity in intact adult rat brains.

Furthermore, NGF does not raise hippocampal levels of CAT on

the control sides of rats with unilateral partial fimbria-

lesions.

The observation that NGF ameliorates the lesion-induced

reduction in hippocampal CAT activity suggests that the level

of cholinergic activity of the remaining septal neurons is

increased in response to NGF administration. In vivo and in

vitro studies have demonstrated NGF's ability to increase

survival and induce sprouting of cholinergic neurons

(Williams et al., 1986; Gahwiler et al., 1987; Hartikka and

Hefti, 1988). However, the NGF-induced increase in

hippocampal cholinergic function is probably not due to

increased sprouting of the transected nerve terminals since

there presently is no evidence for cholinergic re-innervation










of the hippocampus following fimbrial transactions, even

after NGF treatment over a two-week interval.

As noted above, NGF treatment appears to increase the

CAT activity in the unlesioned cholinergic neurons. However,

this increase in activity is only reflected by increased

levels of CAT activity in the hippocampus, but not in the

septum. This observation may be due to rapid transport of

the CAT enzyme from the cell body (septum) to the nerve

terminal (hippocampus). Contrary to our results, Hefti et

al. (1984) reports a 60% increase in septal CAT activity

levels following NGF treatment of rats with unilateral

partial fimbria-lesions. This difference may be due to the

fact that their NGF treatment lasted for 4 weeks instead of 2

weeks, or that they used a 20x higher dose of NGF.

The unexpected detrimental effects of thyrotropin

releasing hormone on cholinergic activity suggest that

thyrotropin releasing hormone can potentially interfere with

normal transport of the CAT enzyme from the cell body to the

nerve terminal. According to this hypothesis, following co-

administration of NGF and thyrotropin releasing hormone, the

NGF-induced increase in CAT levels is restricted to the

septum due to thyrotropin releasing hormone's ability to

block the transport of this enzyme from the septum (cell

body) to the hippocampus (nerve terminal).

Similarly, the thyrotropin releasing hormone-induced

decrease in frontal cortex CAT activity may be due to a

blockade of enzyme transport from the nucleus basalis










magnocellularis to the neocortex. The above hypothesis would

also suggest that cholinergic activity in the nucleus basalis

magnocellularis would be elevated. Levels of CAT activity in

the nucleus basalis magnocellularis were not measured,

because this nucleus is diffuse with cells found in the

globus pallidus and elsewhere, making dissection difficult

and poorly reproducible. The levels of frontal cortex CAT

activity were measured to assess whether thyrotropin

releasing hormone would be selective for injured cholinergic

pathways, as was demonstrated for NGF.

The apparently detrimental effects of thyrotropin

releasing hormone on cholinergic neurons is contrary to

recent findings of other investigators (Horita et al.,1987;

Horita et al., 1989). Using similar lesion models these

investigators reported that MK-771, a thyrotropin releasing

hormone analog, and thyrotropin releasing hormone itself each

increased high affinity choline uptake, another marker of

cholinergic function. The apparent discrepancy in our

results may again be due to the substantial difference in

drug concentration, since the dose of thyrotropin releasing

hormone administered in the present study is approximately

40x higher than theirs.

A disturbing possibility suggested by our observations

is that high concentrations of a neurotrophic factor could

potentially have toxic effects similar to those noted with

thyrotropin releasing hormone, which are different from those

seen with lower doses. Consistent with the hypothesis of a










dose dependent toxic effect of neurotrophic factors is the

recent observation of significant hypertrophy of striatal

cholinergic cell bodies following administration of high

concentrations of NGF (Gage et al, 1988; Hagg et al., 1989).

Similar cholinergic neuronal hypertrophy in the nucleus

basalis magnocellularis is reported by Higgins et al. (1989),

following infusions of high NGF concentrations into the

forebrain. This hypertrophy could be due to excessive

accumulation of proteins such as CAT. The generation of dose

response curves of the trophic and toxic effects of

thyrotropin releasing hormone and NGF is necessary to insure

the design of optimal treatment strategies using these drugs.

Appel (1981) hypothesized that the degeneration of

selective neuronal populations as observed in several

diseases, including Alzheimer's disease, is caused by a lack

of neurotrophic factors. Since NGF is a trophic factor for

basal forebrain cholinergic neurons which degenerate in

Alzheimer's disease, it has been speculated that NGF may be

involved in the pathogenesis in Alzheimer's disease (Hefti,

1983). However, following the formulation of the latter

hypothesis, studies have shown that levels of NGF, NGF-

encoding mRNA, and NGF receptors in the central nervous

system are not affected by Alzheimer's disease (Goedert et

al., 1986; Mufson and Kordower, 1989). These observations

argue against a primary involvement of NGF in the

pathogenesis of Alzheimer's disease. The possibility remains

that NGF-related functions are affected in selective regions










of the cerebrum in Alzheimer's disease, but that these

discrete alterations are not revealed upon gross examination

of the neocortex. Furthermore, since NGF receptors continue

to be expressed by dystrophic as well as healthy cholinergic

basal forebrain neurons in Alzheimer's disease (Mufson and

Kordower, 1989), NGF treatment of Alzheimer's patients may

attenuate the neuronal degeneration and the neocortical

cholinergic hypofunction associated with their death, as

observed in animal studies.
















































Figure 1. Electrophoretic analysis of the purity of a
substance isolated from mouse submaxillary glands.

The substance isolated from mouse submaxillary
glands was subjected to electrophoresis in a sodium
dodecyl sulfate-polyacrylamide gel, and the gel was
subsequently stained with coomassie blue.
Lane 1: 5.2 gg NGF, Lane 2: 7.2 gg of the isolate,
Lane 3: 2.6 gg NGF, Lane 4: 3.6 gg of the isolate,
and Lane 5: molecular weight standards.
















































Figure 2. Western blot analysis of a substance isolated
from mouse submaxillary glands.

The substance isolated from mouse submaxillary
glands was subjected to electrophoresis in a sodium
dodecyl sulfate-polyacrylamide gel, transferred to
nitrocellulose, and probed with a polyclonal
antibody raised in rabbit against mouse NGF.
Lane 1: 0.13 gg NGF, Lane 2: 0.40 gg NGF, Lane 3:
0.18 gg of the isolate, and Lane 4: 0.54 gg of the
isolate.



















* % SURVIVAL
El % NEURONS WITH PROCESSES


ISOLATED
SUBSTANCE


SALINE


24HRS


48 HRS


24 HRS


ii-


48 HRS


Time post-plating


Figure 3.


Trophic activity of a substance isolated from mouse
submaxillary glands.

Dorsal root ganglia cultures were treated with
saline or with the isolated substance (20 ng/ml).
The total number of neurons and the number of
neurons with processes were assessed as described
in the text. All values were expressed as the mean
S.E.M. of 4 regions/plate for 4 plates (*p<0.05
compared to saline).


100 -


0

z
S50-


0 -

















































Figure 4. Trophic activity of the isolated NGF.

Dorsal root ganglia cultures were treated with TOP)
saline and BOTTOM) the isolated substance. These
representative photographs demonstrate the ability
of the isolated substance to increase the survival
of dorsal root ganglia neurons and to stimulate the
formation of neuronal processes.






















~1~


I1


UNLESIONED SIDE


LESIONED SIDE


Treatment


Figure 5.


The effects of unilateral partial fimbria-lesions
on hippocampal CAT activity.

Male Sprague Dawley albino rats received unilateral
partial fimbria-lesions by knife cut. Two weeks
post-lesioning levels of hippocampal CAT activity
were measured for each brain hemisphere as
described in the text. All values were expressed
as the mean S.E.M. of 4 animals/group (*p<0.05
compared to the unlesioned side).


12500 -


4Jr
-d -H
> u
- H-
4J LO
u r-i
(H -d
EA -H
F U 4-)-
0


04
o.


o \
040
04o f
-H 04
a4 -


10000



7500



5000



2500


0 -----






















400




300



200




100




0
SHAM OPERATED LESIONED
Treatment


Figure 6. The effects of unilateral partial fimbria-lesions
on passive avoidance behavior.

Male Sprague Dawley albino rats received unilateral
partial fimbria-lesions by knife cut or they were
sham-operated. Two weeks post-lesioning all rats
were tested for passive avoidance behavior 24 hours
after training as described in the text. All
values were expressed as the mean S.E.M. of 4
animals/group.






48















*

4 80
.H



O 60

0

0




20



0
CSF NGF TRH NGF+TRH

Treatment


Figure 7. The effects of NGF and thyrotropin releasing
hormone on hippocampal CAT.

Alzet osmotic minipumps containing NGF (11 gg/ml),
thyrotropin releasing hormone, TRH, (100 gg/ml),
both drugs, or artificial cerebral spinal fluid,
CSF, were implanted in male Sprague Dawley albino
rats with unilateral partial fimbria-lesions. Two
weeks post-lesioning levels of hippocampal CAT
activity were assessed as described in the text.
All values were expressed as the mean S.E.M. of
4 animals/group (*p<0.05 compared to the control
group treated with artificial cerebral spinal
fluid).























30000
>i
.





4 20000
0-H-


4-)

)O

0

10000"







CSF NGF TRH NGF+TRH

Treatment





Figure 8. The effects of NGF and thyrotropin releasing
hormone on frontal cortex CAT activity.

Alzet osmotic minipumps containing NGF (11 gg/ml),
thyrotropin releasing hormone, TRH, (100 gg/ml),
both drugs, or artificial cerebral spinal fluid,
CSF, were implanted in male Sprague Dawley albino
rats with unilateral partial fimbria-lesions. Two
weeks post-lesioning levels of frontal cortex CAT
activity were assessed as described in the text.
All values were expressed as the mean S.E.M. of
4 animals/group (*p<0.05 compared to the control
group treated with artificial cerebral spinal
fluid).




















CONTROL
8000- E3 LESIONED SIDE



-H LO
> 6000














CSF NGF TRH NGF+TR-
4-P
4 P 4000










CSF NGF TRH NGF+TRH

Treatment





Figure 9. The effects of NGF and thyrotropin releasing
hormone on septal CAT activity.

Alzet osmotic minipumps containing NGF (11 gg/ml),
thyrotropin releasing hormone, TRH, (100 gg/ml),
both drugs, or artificial cerebral spinal fluid,
CSF, were implanted in male Sprague Dawley albino
rats with unilateral partial fimbria-lesions. Two
weeks post-lesioning levels of septal CAT activity
were assessed as described in the text. All values
were expressed as the mean S.E.M. of 3-4
animals/group (*p<0.05 compared to the control
group treated with artificial cerebral spinal
fluid).















CHAPTER 3
THE NUCLEUS BASALIS MAGNOCELLULARIS LESION MODEL


Introduction


The nucleus basalis magnocellularis-neocortex

cholinergic pathway was particularly useful for the study of

potentially cytoprotective agents, since preliminary data

were available regarding the long-term trans-synaptic changes

associated with ibotenate lesions of this pathway (Arendash

et al., 1987). Nucleus basalis magnocellularis-lesions are

characterized by a reduction in neocortical pre-synaptic

cholinergic markers in those regions of the neocortex that

receive cholinergic input from the nucleus basalis

magnocellularis (Bartus et al., 1982; Strong et al., 1980;

Watson et al., 1985). Nucleus basalis-lesions also induce

memory-related deficits (Altman et al., 1985; Flicker et al.,

1983; Lever and Nicholson, 1986; Salamone et al., 1984), as

well as other neocortical neurochemical and neuropathological

changes (Arendash et al., 1987; Gaykema et al., 1989).

This chapter delineates some of the long-term

behavioral, neuropathological, and neurochemical changes

induced by nucleus basalis magnocellularis-lesions, as well

as the effects of chronic nicotine and/or pilocarpine

treatment on the lesion-induced neurochemical and










neuropathological changes. Frontal cortex CAT activity and

passive avoidance behavior were measured to confirm the

effectiveness of the nucleus basalis magnocellularis-lesions.

Occipital cortex neuropeptide Y, parietal cortex neuropeptide

Y-encoding mRNA, and frontal cortex nicotine receptor levels

were measured to assess the degree of the lesion-induced

trans-synaptic neurochemical change. Parietal cortex

neuronal density and size distribution as well as glial

density were measured to assess the degree of the lesion-

induced trans-synaptic neuropathology. Following chronic

cholinergic treatments, these parameters were again measured.


Materials and Methods



Chemicals and Reagents


Ibotenic acid, pilocarpine, atropine, carbamylcholine

chloride, eserine sulfate, choline chloride, (-)-nicotine,

and cresyl violet acetate were purchased from Sigma

Chemicals, St. Louis, MO. The Pierce protein-estimation-

reagent was purchased from Pierce Chemical Company, Rockford,

IL. The radiochemicals purchased from New England Nuclear,

Boston, MA, included: [3H-methyl] choline chloride (80

Ci/mmol), [3H-acetyl] coenzyme A (200mCi/mmol), [1251]

neuropeptide Y (2200 Ci/mmol), and [a-32P] dCTP (800

Ci/mmol). The nick translation system, Eco R1 and Bam H1

restriction enzymes, lambda/Hind 3 and phi/X digests, and the










RNA ladder were purchased from Bethesda Research

Laboratories, Gaithersburg, MD. Xylazine, ketamine, and

penicillin G were purchased from the University of Florida

Department of Animal Resources. All other chemicals used

were reagent grade.


Animals


Male Sprague Dawley albino rats (350-450 g) were

purchased from the University of Florida Department of Animal

Resources or from Charles River Laboratories, North

Wilmington, MA. The rats were professionally maintained in a

220C environment according to NIH standards and had ad

libitum access to water and food (Purina Rat Chow), except

following surgeries.


Nucleus Basalis Magnocellularis-Lesions


Ibotenic acid was dissolved in phosphate buffered saline

pH 7.4, at a concentration of 5 mg/ml. One .1 of the acid

was infused either bilaterally or unilaterally into the

nucleus basalis magnocellularis with a CMA/100 microinjection

pump (Bioanalytical Systems Inc., Lafayette, IN) at a flow

rate of 0.33 gl/minute. Using a stereotaxic apparatus (Model

900 or Model 1404, David Kopf Instruments, Tujunga, CA) and

the rat brain atlas of Paxinos and Watson (1986), the

infusion cannula (23 gauge stainless steel hypodermic tubing)

was positioned into the nucleus basalis magnocellularis










located at the following coordinates: 7.2 mm anterior to the

interaural line, 2.6 mm lateral to the sagittal suture, and

6.6 mm below the neocortex. The incisor bar was set at -2.6

mm.

All rats were anesthetized using a combination of 100

mg/kg ketamine (1 ml/kg, intraperitoneal) and 20 mg/kg

xylazine (1 ml/kg, intramuscular). Controls for bilaterally

lesioned rats underwent sham surgery but did not receive

ibotenic acid infusions into the nucleus basalis

magnocellularis. Immediately following surgery, the rats

received an intramuscular injection of 30,000 units

penicillin G. They were housed individually and weighed

regularly. For 2 weeks following surgery, rats that received

bilateral ibotenate infusions were fed a mixture of crushed

Purina Rat Chow and vegetable oil to maintain their body

weight and increase their rate of survival (as found in

preliminary studies).


Determination of CAT Activity


CAT activity was measured in the frontal cortex of each

brain hemisphere as described previously by Meyer et al.,

1985. In short, tissue homogenates were incubated with [3H-

acetyl] coenzyme A and choline. The newly synthesized [3H-

acetyl] choline was separated from acetylcoenzyme A by ion

pair extraction, and the pmol [3H-acetyl] choline per mg

protein calculated. Protein content was estimated after










sodium hydroxide solubilization of the tissue using the

Pierce protein-estimation-reagent with a bovine serum albumin

standard.


Determination of Nicotine Receptor Concentration


Nicotine receptor binding was measured in one half of

the frontal cortex as described previously by Schwartz et al.

(1982). In short, [3H-acetyl] choline was incubated with

plasma membrane homogenates in the presence of atropine and

eserine physostigminee). Scatchard analyses showed this

ligand to bind with a single Kd of about 20 nM, the

concentration used for these binding studies (data not

shown). Carbachol (100 gM) was added to some duplicate

samples to determine non-specific binding. Results were

normalized for the amount of protein present in each sample.

Protein content was measured as described above.


Determination of Neuropeptide Y Levels


One half of the occipital cortex was used to measure

neuropeptide Y levels with a specific double antibody

radioimmunoassay developed by Beal et al. (1986b). The

antibody was generated in rabbit against synthetic

neuropeptide Y coupled to bovine serum albumin. Iodinated

neuropeptide Y served as the labeled preparation and porcine

neuropeptide Y as the standard. All radioimmunoassay's were

performed in the laboratory of Dr. William J. Millard.











RNA Purification and Analysis


Isolation of Total RNA. Total cellular RNA was isolated

from the parietal cortex of each hemisphere by the

guanidinium/isothiocyanate phenol extraction method of

Chomczynski and Sacchi (1987). The tissue was homogenized in

the guanidinium-isothiocyanate solution with a polytron

(Brinkman Instruments, a Division of Syborn Corporation,

Westburg, NY) and then phenol-extracted. The isolated total

RNA was quantitated spectrophotometrically by absorption (A)

at 260 nm and its purity assessed by measuring A260/A280

ratios. Only total RNA with a ratio greater than 1.7 was

used for further analysis. To verify the integrity of our

total RNA, i.e. to ensure that no degradation occurred during

the isolation procedure, aliquots of the total RNA samples

were investigated on agarose/formaldehyde gels (Sambrook et

al., 1989).

The Neuropeptide Y Containing Vector. The neuropeptide

Y probe is a 551 basepair sequence that includes an 86

basepair 5' untranslated region, the 291 basepair pre-pro-

neuropeptide Y encoding region, and a 174 basepair 3'

untranslated region. The neuropeptide Y containing vector

was generously provided by Dr. Janet Allen, Massachusetts

General Hospital (Allen et al., 1987). Competent E. coli

cells were transformed and the positive colonies were

amplified in liquid culture to establish a stock of the










neuropeptide Y-containing plasmid vector. The neuropeptide Y

plasmid was then isolated by equilibrium centrifugation in a

cesium chloride/ethidium bromide density gradient. The

neuropeptide Y insert was digested from the plasmid with the

restriction endonuclease Eco RI. The insert was isolated

after purification by agarose gel electrophoresis and labeled

by primer extension with [a-32p] dCTP (Sambrook et al.,

1989). Northern blot analysis showed that with the

appropriate hybridization conditions, the neuropeptide Y

probe identified a single RNA species with a migration

pattern similar to the 550 base pairs long neuropeptide Y-

encoding mRNA.

Expression of Neuropeptide Y-encoding mRNA. To

determine the relative amounts of neuropeptide Y-encoding

mRNA, the total RNA was subjected to dot-blot analysis.

Total RNA (1, 3, 10 pg) was applied to nitrocellulose paper

(Nitroplus 2000, Fisher, Orlando, FL) and hybridized with the

32p labeled cDNA probe encoding the rat neuropeptide Y gene.

The blots were prehybridized in 50% formamide, 5x SSC, 0.1%

SDS, 5x Denhardts solution and 150 99/ml salmon sperm DNA at

420C overnight. Hybridization was carried out under

conditions previously shown to result in recognition of a

single band in Northern blots, with the addition of a minimum

of 1x106 cpm per ml of the 32P labeled neuropeptide Y probe.

The blots were washed with 2x SSC, 0.1% SDS at room

temperature and at 600C for 30 minutes each. They were

subsequently washed twice with 0.2x SSC, 0.1% SDS at 600C for









30 minutes each. The blots were exposed to X-ray film with

intensifying screens at -700C. The intensity of the

hybridizing signal was quantitated by scanning laser

densitometry. Other blots were analyzed directly using the

Betascope Model 603 Blot Analyzer (Betagen Corporation,

Waltham, MA). All samples analyzed in this manner displayed

radiation densities that were proportional to their RNA-

concentrations (1 to 10 gg).


Passive Avoidance Behavior Testing


Rats were placed in the lighted compartment of a

standard two compartment passive avoidance apparatus with the

connecting door between the two compartments opened. As soon

as the rat touched the floor of the lighted compartment, the

timer was started. The connecting door was closed when the

rat passed through the door, and the timer was stopped as

soon as it fully entered the dark compartment. Upon entering

this compartment the rat received a 0.8 mAmp footshock for 1

second. Latency to enter the dark compartment was recorded

up to a maximum of 5 minutes during this training period.

The rats were tested in similar fashion 24 hours later, but

no shock was administered upon entering the dark compartment.

Latency for this testing period was also 5 minutes.

For the acute studies of nicotine-induced behavioral

changes using unoperated rats, the following modifications

were made: the footshock level was reduced to 0.5 mAmps, and









the time interval between training and testing was increased

to 72 hours. These changes, which were based on several

preliminary studies, allowed the intact and therefore more

intelligent rats sufficient time to forget their training so

that nicotine's ability to enhance this memory-related

behavior could be assessed.


Determination of Neuronal Density


Rats with either unilateral or bilateral lesions of the

nucleus basalis magnocellularis were sacrificed by

decollation and their brains were dissected. Upon removal,

the brains were fixed in a 10% formalin, 0.15 M sodium

acetate, pH 7.0 buffer. To reduce the risk of tissue damage

caused by the formation of ice crystals and to facilitate

sectioning of the tissue in the frozen state, the brains were

placed in a 30% sucrose, 10% formalin, 0.15 M sodium acetate,

pH 7.0 buffer two days prior to sectioning. Initially, 30 pm

frozen sections were collected from the entire rat brain

using a sliding microtome ( Model 860, American Optical

Corporation, Buffalo, NY). The brain sections were placed on

gelatinized slides and stained with a Nissl-selective stain

containing: 0.1% cresyl violet dye, 0.003% glacial acetic

acid, and 1.5 mM sodium acetate, as described previously

(Arendash et al., 1987).

In addition to frozen brain sections, 10 pm paraffin

sections were also generated. To generate paraffin sections,










the fixed rat brains were first trisected. The three tissue

slices were then embedded in paraffin with a Tissue Tek 2

Embedding Center (American Scientific Products, Miami, FL).

Using a rotary microtome (Model 820, American Optical

Corporation, Buffalo, NY), 10 pm brain sections were

subsequently collected. Neuronal density was determined

manually or with the aid of a CUE 2 image analysis system (C2

corporation, Tamarac, FL) at a magnification of 500x. Manual

assessment of neuronal density was accomplished using a

redicule with 1 mm squares mounted in the eyepiece, a Nikon

microscope,and a Nikon drawing tube attachment (Southern

Micro Instruments, Orlando, FL). Neurons were manually

counted in a 0.3 by 0.3 mm area in layers 2 and 6 of the

parietal cortex.

The image analysis system counted neurons in a 0.26 by

0.19 mm2 area in the same brain regions. In addition to

neuronal density the image analysis system allowed the

measurement of changes in the size distribution of the

neocortical neuronal populations as well as changes in glial

density. Only those neurons with a distinct nucleolus and

more than half of their cell body within the boundaries of

the square were included in the analysis. For each animal, 3

to 5 consecutive brain sections were analyzed. All brain

sections included in the study were at least 30 pm apart to

prevent analysis of the same neuron more than once.










Scintillation Counting


All radioactivity measurements were performed via liquid

scintillation spectrophotometry using Liquiscint (National

Diagnostics, Sommerville, NJ) and a Beckman 7100 A

spectrophotometer. The counting efficiencies for 3H and 32p

approximated 50% and 95%, respectively, in this

spectrophotometer.


Statistics


All data were expressed as the mean standard error of

the mean (S.E.M). The Student t-test was used to analyze

comparisons between two groups only. To assess the equality

of several population means, the one way analysis of variance

test and the Scheffe F-test were used.


Results



The Long-Term Effects of Nucleus Basalis Magnocellularis-
Lesions


Passive avoidance behavior. To determine the time

course of the behavioral recovery that ultimately resulted in

normal passive avoidance behavior of rats with bilateral

nucleus basalis magnocellularis-lesions 14 months post-

lesioning (Arendash et al., 1987), sham-operated and lesioned

animals were tested 2 months, 5 months, and 8 months post-

lesioning. As shown in Figure 10, all animals displayed a









detriment in passive avoidance behavior at 2 months post-

lesioning. However, by 5 months post-lesioning half of the

animals, and by 8 months post-lesioning all lesioned rats

displayed normal passive avoidance behavior.

Neurochemical parameters. As shown in Figure 11, levels

of frontal cortex CAT activity were significantly reduced 2

months, 5 months, and 8 months post-lesioning in rats with

bilateral nucleus basalis-lesions. CAT activity levels of

both sham-operated and bilaterally lesioned rats at the 8

month post-lesioning time point were about 4x lower than

those at the other time points. This difference was due to a

change in our CAT assay procedure. The amount of [3H-acetyl]

coenzyme A was reduced to approximately 1/4 the original

amount. The sham-operated control values for frontal cortex

CAT activity of the other time points (2 and 5 months post-

lesioning) were essentially similar at the p<0.05 level with

the student t-test. The frontal cortex CAT activity deficits

increased (from 17 to 33%) as the time post-lesioning

increased.

As shown in Figure 12, rats with unilateral nucleus

basalis-lesions had lower levels of frontal cortex CAT

activity 2 weeks, 1 month, 2 months, and 5 months post-

lesioning. The unlesioned control values of frontal cortex

CAT activity were essentially similar at the p<0.05 level

with one way analysis of variance (ranging from 154 10 to

178 11 pmol/gg protein/15 minutes) of all time points

measured. The frontal cortex CAT activity deficits tended to









decrease (from 30 to 17%) as the time post-lesioning

increased.

As shown in Figure 13, in rats with bilateral nucleus

basalis-lesions occipital cortex levels of neuropeptide Y

remained unchanged up to 8 months post-lesioning. However,

between 8 and 10 months post-lesioning occipital cortex

neuropeptide Y levels increased dramatically. The sham-

operated control values for occipital cortex neuropeptide Y

levels were essentially similar at the p<0.05 level with one

way analysis of variance (ranging from 214 30 to 218 23

pg/mg wet weight), with the exception of the 8 month post-

lesioning time point (382 33 pg/mg wet weight). The

neuropeptide Y levels in both bilaterally lesioned and sham-

operated rats at 8 months post-lesioning were higher compared

to the other time points.

As shown in Figure 14, occipital cortex neuropeptide Y

levels were decreased by 1 month post-lesioning in rats with

unilateral nucleus basalis-lesions. However, by 2 months

post-lesioning neuropeptide Y levels were again similar to

those of the unlesioned control side. The unlesioned control

values for occipital cortex neuropeptide Y levels were

essentially similar at the p<0.05 level with one way analysis

of variance (ranging from 5200 1200 to 7900 1100 pg/mg

protein) at all time points. There was more variability in

the neuropeptide Y data than usual as reflected by the larger

S.E.M. compared to the data presented in Figure 13. This









increased variability may be due to the accidental storage of

these samples at -200C, instead of -800C, for 48 hours.

As shown in Figure 15, expression of neuropeptide Y-

encoding mRNA in the parietal cortex was reduced 2 months

after bilateral nucleus basalis lesioning. However, by 5

months post-lesioning neuropeptide Y-encoding mRNA levels

were again similar to those of sham-operated controls. By 8

months post-lesioning neuropeptide Y-encoding mRNA levels

were slightly increased, but this elevation was not

statistically significant. To avoid variability due to

potential differences in the specific activity of the probe,

the lesion-induced changes in neuropeptide Y-encoding mRNA

were expressed as a percentage of sham-operated control

samples studied concurrently. Due to these experimental

limitations comparisons between different time-points or

treatments may not be significant unless the samples were

studied concurrently. The potential drug-induced changes in

parietal cortex neuropeptide Y-encoding mRNA were measured

concurrently; that data was therefore expressed as

densitometry units per gg total RNA.

As shown in Figure 16, unilateral nucleus basalis-

lesions did not affect parietal cortex neuropeptide Y-

encoding mRNA levels at the time points (2 weeks, 1 month, 2

months, 5 months) investigated. Neuropeptide Y-encoding mRNA

levels were again expressed as a percentage of the respective

unlesioned control values to avoid variability due to









potential differences in the specific activity of the probe,

as discussed previously.

Frontal cortex nicotine receptor concentrations were

elevated by 5 months post-lesioning in rats with bilateral

nucleus basalis-lesions (Figure 17). However, by 8 months

post-lesioning nicotine receptor concentrations in lesioned

rats were slightly reduced (25%) compared to sham-operated

controls, though this reduction was not statistically

significant at the p<0.05 level with student t-test. The

sham-operated control values for frontal cortex nicotine

receptor concentrations were essentially similar at the

p<0.05 level with the student t-test. A recent study

performed in the laboratory of Dr. Edwin M. Meyer found no

significant change in frontal cortex nicotine receptor

concentrations in bilaterally lesioned rats, 2 months post-

lesioning. Instead of [3H-acetyl] choline, these studies

used [3H-methyl] carbachol as the radioligand (unlesioned

controls: 19.7 4.2 fmol/mg protein and bilaterally lesioned

rats: 17.0 1.5 fmol/mg protein). Acetylcholine and

methylcarbachol appear to bind to the same sub-type of high

affinity nicotine receptors, under the binding conditions

used in these experiments.

Neuropathological parameters. As shown in Figure 18,

analysis of 30 pm frozen brain sections of rats with

bilateral nucleus basalis-lesions revealed that neuronal

density in layer 2 of the parietal cortex was significantly

reduced (17%), 8 months post-lesioning. Analysis of 10 pm










paraffin brain sections of additional rats at this time

point, 8 months post-lesioning, revealed a similar decrease

(23%) in neuronal density in layer 2 of the parietal cortex

(Figure 19). Layer 6 of the parietal cortex of these rats,

revealed a similar reduction (18%) in neuronal density

(Figure 20).

As shown in Figure 21, analysis of 10 jim paraffin brain

sections of rats with unilateral nucleus basalis-lesions

revealed that neuronal density in layer 2 of the parietal

cortex of the lesioned brain hemisphere was reduced to a

similar extent (17%) as following bilateral lesions, at 8

months post-lesioning.

The mean neuronal density of the sham-operated rats

derived by examination of frozen sections was lower than that

derived by examination of paraffin sections. This

discrepancy may be due to the different embedding procedures.

The frozen brain sections expand more than the paraffin

embedded sections while being mounted on the microscope slide

from the water bath. Furthermore, the paraffin embedding

procedure itself causes some shrinkage of the tissue due to

the high temperatures necessary to ensure complete perfusion

of the rat brain.

Unilateral nucleus basalis-lesions did not affect

neuronal size distribution in layer 2 of the parietal cortex

by 8 months post-lesioning, as shown in Figure 22. There was

a small but not statistically significant shift toward the










left (smaller neurons). Nine size bins were used to allow

detection of subtle changes in neuronal size distribution.

As shown in Figure 23, glial density in layer 2 of the

parietal cortex of the lesioned brain hemisphere was not

significantly affected by unilateral nucleus basalis-lesions.

The QUE 2 image analysis system discriminates between neurons

and glia solely by comparing differences in size. This

system thus contains the built-in danger of misclassifying

shrunken or small neurons as glia. All neurons with a

diameter less than 40 pm2 will thus be included in the glia

counts, a factor that could potentiate a decrease or mask an

increase in neuronal density.

Summarizing these results, nucleus basalis

magnocellularis-lesions induced a wide array of behavioral,

neurochemical, and neuropathological changes in rats. Tables

1 and 2 summarize the changes associated with long-term

bilateral and unilateral nucleus basalis-lesions,

respectively.


The Effects of Chronic Cholinergic Treatment of Rats with
Nucleus Basalis Magnocellularis-Lesions


Determination of the optimal dose of nicotine.

Preliminary studies were performed to determine which dose of

nicotine could optimally improve memory-related passive

avoidance behavior, a behavior known to be dysfunctional in

rats with nucleus basalis-lesions. It was first established

that normal, unoperated animals no longer recalled receiving










a footshock during the training period and therefore readily

entered the dark compartment upon testing 72 hours later

(data not shown). Increasing doses of nicotine were tested

for their ability to enhance passive avoidance behavior in

these rats. Nicotine was administered intraperitoneally 5

minutes prior to training and again 5 minutes prior to

testing of the animals.

As shown in Figure 24, nicotine induced a dose-dependent

improvement in passive avoidance behavior. One ml/kg of 0.2

mg/kg nicotine was the highest dose of nicotine that improved

passive avoidance behavior. Rats that received the dose of

0.5 mg/kg nicotine appeared uncoordinated and could not be

trained to enter the dark compartment of the passive

avoidance behavior apparatus. In contrast, latency to enter

the dark compartment during the training phase was unaffected

by the other nicotine concentrations, 0.05 and 0.2 mg/kg,

administered although these doses improved test-behavior.

These latency values were essentially similar at the p<0.05

level with one way analysis of variance (ranging from 16

8.2 to 20 8.2 seconds).

Starting 5 months post-lesioning, sham-operated rats and

rats with bilateral nucleus basalis magnocellularis-lesions

received daily injections of 1 ml/kg 0.9% saline diluent; 0.2

mg/kg nicotine; 0.2 mg/kg pilocarpine; or 0.2 mg/kg

nicotine/pilocarpine intraperitoneally until their sacrifice

3 months later (8 months post-lesioning). As shown in Figure

25, total body weight gain of rats with bilateral nucleus










basalis-lesions was decreased during the first weeks of

chronic nicotine treatment at the p<0.05 level with one way

analysis of variance compared to sham-operated rats treated

with saline. However, toward the end of the study this

decrease in weight gain was no longer significant.

Furthermore, total body weight gain was also significantly

decreased in lesioned rats treated chronically with

nicotine/pilocarpine. This decrease in weight gain was only

significant toward the end of the study. In addition, none

of the treatments significantly affected total body weight

gain of sham-operated rats.

Starting 5 months post-lesioning rats with unilateral

nucleus basalis magnocellularis-lesions received daily

injections of 1 ml/kg 0.9% saline diluent; 0.2 mg/kg

nicotine; or 0.2 mg/kg pilocarpine intraperitoneally until

their sacrifice 3 months later (8 months post-lesioning). As

shown in Figure 26, none of the treatment paradigms

significantly affected total body weight gain with the one

way analysis of variance test.


The Effects of Chronic Nicotine Treatment of Rats with
Nucleus Basalis Magnocellularis-Lesions


Passive avoidance behavior. As shown in Figure 27,

chronic nicotine treatment interfered with passive avoidance

behavior of rats with bilateral nucleus basalis-lesions

tested at 8 months post-lesioning. However, passive

avoidance behavior of sham-operated rats treated with









nicotine was similar to that of sham-operated rats treated

with saline at the level of p<0.05 with the student t-test.

Neurochemical parameters. As shown in Figure 28,

chronic nicotine treatment (from 5 to 8 months post-

lesioning) did not affect occipital cortex neuropeptide Y

levels in sham-operated rats or rats with bilateral nucleus

basalis-lesions. As noted previously, nucleus basalis-

lesions alone also did not induce any changes in neuropeptide

Y levels at this time point. All values were essentially

similar at p<0.05 level with the one way analysis of variance

(ranging from 323 20 to 386 60 pg/mg wet weight).

As shown in Figure 29, levels of parietal cortex

neuropeptide Y-encoding mRNA were significantly elevated to a

similar extent in rats with bilateral nucleus basalis-lesions

as well as sham-operated rats following chronic nicotine

treatment. As noted previously, nucleus basalis-lesions

alone did not affect the expression of neuropeptide Y-

encoding mRNA at 8 months post-lesioning.

As shown in Figure 30, frontal cortex nicotine receptor

concentrations were increased in rats with bilateral nucleus

basalis-lesions following chronic nicotine treatment.

However, nicotine receptor levels in sham-operated rats

remained unchanged. The sham-operated control values for

frontal cortex nicotine receptor concentrations were

essentially similar at the p<0.05 level with the student t-

test. The nicotine-induced elevation in nicotine receptor

concentrations in lesioned rats was similar in size to the










lesion-induced increase observed 5 months post-lesioning. As

reported previously (see Figure 17), the lesion-induced

elevation in nicotine receptor levels was no longer present 8

months post-lesioning.

Neuropathological parameters. Analysis of 30 gm frozen

brain sections of rats with bilateral nucleus basalis-lesions

revealed that chronic nicotine treatment attenuated the

lesion-induced trans-synaptic cell loss in layer 2 of the

parietal cortex (Figure 31). Analysis of 10 pm paraffin

sections of additional rats treated in a similar fashion

confirmed the ability of chronic nicotine treatment to

attenuate the neuronal cell loss in layer 2 induced by

bilateral nucleus basalis-lesions (Figure 32). Furthermore,

chronic nicotine also ameliorated the lesion-induced neuronal

cell loss in layer 6 of the parietal cortex of these rats

(Figure 33). Chronic nicotine treatment did not

significantly affect neuronal density in layers 2 or 6 of the

parietal cortex of sham-operated rats in these studies.

Sham-operated values of each study were essentially similar

at the p<0.05 level with the student t-test.

The effects of nicotine on neuronal density in

unilaterally lesioned rats were more complex. As shown in

Figure 34, chronic nicotine treatment appeared not to

ameliorate neuronal cell loss in layer 2 of the parietal

cortex in the lesioned brain hemisphere of rats with

unilateral nucleus basalis-lesions. On the other hand,

neuronal density in this layer was also reduced to a similar









extent (14%) in the unlesioned brain hemisphere of rats

chronically treated with nicotine. Thus, one interpretation

is that the nicotine treatment itself had a detrimental

effect on apparent neuronal density, even as it blocked the

trans-synaptic cell loss.

Along this line, there was a significant shift toward

the left in the neuronal size distribution in layer 2 of the

parietal cortex in the lesioned as well as the unlesioned

brain hemispheres of rats treated chronically with nicotine

(Figure 35). The percentage of neurons within the 40 to 60

Im2 and 65 to 90 gm2 size bins increased substantially, while

the percentage of neurons in the larger size bins (90 to 290

im2) decreased. This shift in the neuronal size distribution

was more pronounced in the unlesioned brain hemisphere of

rats chronically treated with nicotine.

As shown in Figure 36, chronic nicotine treatment

induced a small increase in glial density in layer 2 of the

parietal cortex of the lesioned as well as the unlesioned

brain hemispheres of rats with unilateral nucleus basalis-

lesions. However, like the lesion-induced elevation in glial

density, the nicotine-induced increase was not statistically

significant.











The Effects of Chronic Pilocarpine Treatment of Rats with
Nucleus Basalis Magnocellularis-Lesions


Neurochemical parameters. As shown in Figure 37,

occipital cortex levels of neuropeptide Y in rats with

bilateral nucleus basalis magnocellularis-lesions were

unaffected by chronic treatment with pilocarpine. Though

neuropeptide Y levels were slightly reduced in both lesioned

and sham-operated rats treated with pilocarpine, this

reduction was not statistically significant. All values were

essentially similar at the p<0.05 level with one way analysis

of variance (ranging from 316 50 to 386 60 pg/mg wet

weight).

Chronic pilocarpine treatment did not affect levels of

parietal cortex neuropeptide Y-encoding mRNA in rats with

bilateral nucleus basalis-lesions either, as shown in Figure

38. All values were essentially similar at the p<0.05 level

with the one way analysis of variance (ranging from 280 58

to 410 58 densitometry units/3 gg total RNA).

As shown in Figure 39, frontal cortex nicotine receptor

concentrations were reduced in sham-operated rats (25%) as

well as rats with bilateral nucleus basalis-lesions (20%)

following chronic pilocarpine treatment. However, the

reduction in nicotine receptor concentrations was only

statistically significant at the p< 0.05 level with the

student t-test in sham-operated rats treated with

pilocarpine. As noted previously, at this time point (8










months post-lesioning) bilateral nucleus basalis-lesions

alone no longer affected frontal cortex nicotine receptor

concentrations.

Neuropathological parameters. As shown in Figure 40,

analysis of 10 pm paraffin brain sections of rats with

bilateral nucleus basalis-lesions revealed that chronic

pilocarpine treatment did not ameliorate the lesion-induced

cell loss in layer 2 of the parietal cortex. Furthermore,

chronic pilocarpine treatment did not adversely affect

neuronal density in layer 2 of the parietal cortex of sham-

operated rats.


The Effects of Chronic Nicotine/Pilocarpine Co-treatment of
Rats with Nucleus Basalis Magnocellularis-Lesions


Neurochemical parameters. As shown in Figure 41,

occipital cortex levels of neuropeptide Y in rats with

bilateral nucleus basalis-lesions and sham-operated rats were

not affected by chronic nicotine/pilocarpine co-treatment.

All values were essentially similar at the p<0.05 level with

one way analysis of variance (ranging from 352 48 to 386

60 pg/mg wet weight).

Parietal cortex neuropeptide Y-encoding mRNA levels of

rats with bilateral nucleus basalis magnocellularis-lesions

and sham-operated rats were not affected by chronic

nicotine/pilocarpine co-treatment (Figure 42). All values

were essentially similar at the p<0.05 level with one way









analysis of variance (ranging from 210 34 to 410 26

densitometry units/3 gg total RNA).

As shown in Figure 43, frontal cortex nicotine receptor

concentrations were elevated following chronic

nicotine/pilocarpine co-treatment in rats with bilateral

nucleus basalis-lesions. This elevation was similar in size

as that observed in bilaterally lesioned rats treated

chronically with nicotine. However, frontal cortex nicotine

receptor concentrations were reduced in sham-operated rats

treated chronically with nicotine/pilocarpine to a similar

extent (29%) as sham-operated rats treated chronically with

pilocarpine alone.

Neuropathological parameters. Analysis of 10 pm

paraffin brain sections of rats with bilateral nucleus

basalis-lesions revealed that chronic nicotine/pilocarpine

co-treatment ameliorated the lesion-induced cell loss in

layers 2 (Figure 44) and 6 (Figure 45) of the parietal cortex

to a similar extent as nicotine treatment alone.

Furthermore, chronic nicotine/pilocarpine co-treatment did

not adversely affect neuronal density in these layers in the

parietal cortices of sham-operated rats.

Attempts to re-establish central cholinergic function

pharmacologically using nicotine and/or pilocarpine affected

a variety of the neurochemical and neuropathological changes

otherwise associated with long-term nucleus basalis

magnocellularis-lesions. Tables 3 and 4 summarize how

chronic nicotine and/or pilocarpine treatment affected the










lesion-induced behavioral, neurochemical and

neuropathological changes in rats with bilateral or

unilateral nucleus basalis-lesions, respectively.


Discussion


The results of this study are significant because they

describe for the first time the time course by which several

behavioral, neurochemical and neuropathological changes occur

in a model for Alzheimer's disease: rats with neocortical

cholinergic hypofunction induced by nucleus basalis

magnocellularis-lesions. These results also demonstrate for

the first time that attempts to re-establish central

cholinergic function pharmacologically using nicotine and/or

pilocarpine affect a variety of the neurochemical and

neuropathological changes otherwise associated with long-term

nucleus basalis magnocellularis-lesions.

We investigated the neocortical neuropathology

associated with nucleus basalis-lesions by measuring neuronal

and glial density as well as neuronal size distribution.

Frontal cortex CAT activity and passive avoidance behavior

were measured to confirm the effectiveness of the nucleus

basalis magnocellularis-lesions. Occipital cortex

neuropeptide Y and frontal cortex nicotine receptor levels

were measured to assess the degree of the lesion-induced

trans-synaptic neurochemical change. These parameters were

again measured following nicotine and/or pilocarpine










administration to examine the treatments' ability to modulate

the lesion-induced changes.

That the ibotenic acid infusions destroyed cholinergic

neurons in the nucleus basalis magnocellularis is

demonstrated by the significant loss of CAT activity in the

frontal cortex. Previous studies from the laboratory of Dr.

Edwin M. Meyer demonstrated a consistent 90% loss of

acetylcholinesterase staining cholinergic neurons in the

nucleus basalis magnocellularis following similar lesions

(Arendash et al., 1987). The residual CAT activity in the

neocortex derives primarily from intrinsic neocortical

neurons that are not affected by the lesions (Fibiger, 1982;

Johnston et al., 1981; Smith, 1988). Though some

investigators (Wenk and Olton, 1984) report some recovery of

CAT function in rats with unilateral nucleus basalis-lesions

over extended intervals (6 months), in this study the

neocortical CAT deficits persist over time in rats with

unilateral as well as bilateral nucleus basalis-lesions.

Later stages of Alzheimer's disease are characterized by

severe cerebral atrophy due to a substantial loss of and a

reduction in the size of neurons, primarily in the neocortex

and hippocampus (Arai et al., 1987; Ball et al., 1985; Terry

et al., 1981). The cerebral atrophy in Alzheimer's disease

has been characterized by postmortem as well as biopsy

studies. Postmortem studies may reflect changes found only

at the end of dementia, while biopsy studies allow an

assessment of the early stages of Alzheimer's disease.









Though both methods report similar neuronal losses, the

autopsy studies find no correlation between the severity of

dementia and neuronal loss (Mountjoy et al., 1986), whereas

the biopsy studies report significant correlations between a

number of psychological tests and cell loss in layers 3 and 5

of the neocortex (Neary et al., 1986).

The extent of neuronal loss in the cerebral cortex in

Alzheimer's disease is somewhat controversial. Whereas Terry

et al. (1981) reports a 40% reduction in neurons larger than

9.5 gm diameter in the frontal cortex and a 46% decrement in

the temporal region, Mountjoy et al., 1983, finds deficits to

average 23% for the frontal and 31% for the superiotemporal

gyrus. More recent studies (Terry et al., 1987) again report

a selective loss of large pyramidal neurons in layers 3 and 5

of the neocortex, particularly in association areas, in

Alzheimer's disease, while Hubbard and Anderson (1985) note a

26% reduction in neurons larger than 12 gm in diameter in the

frontal cortex, and a 33% reduction in the temporal gyrus.

Small neurons (less than 12 pm in diameter) are found to be

unchanged in the latter study. In contrast, Braak and Braak,

1986, report that small, nonpyramidal neurons are

preferentially reduced in number in Alzheimer's disease.

These discrepancies may be due to differences in the ages of

the patients, duration of the disease, and possibly different

sampling and counting procedures.

There is now unanimity that the entire cortical mantle

receives its major source of cholinergic innervation from the









basal forebrain system. Though other neurotransmitters are

present in the basal forebrain, including galanin and GABA,

the nucleus basalis projections to the neocortex (i.e. motor,

somatosensory, association, and visual cortices) are 80 to

90% cholinergic in nature (Wainer and Mesulam, 1990).

Different cytoarchitectonic regions of the neocortex are

innervated by different subsets of cells within the nucleus

basalis and colateralization is limited (Saper, 1984). In

rats and primates, nucleus basalis neurons have terminal

fields that are restricted to a single cytotectonic area,

usually less than 1.5 mm in diameter (Price and Stern, 1983),

and their targets are largely limited to the ipsilateral

brain hemisphere (Kristt et al., 1985; Saper, 1984).

Although the ascending cholinergic terminals are

typically found in all cortical layers, layer 5 appears to be

particularly heavily innervated in motor and sensory cortical

areas (Eckenstein et al., 1988; McDonald et al., 1987).

Knowledge concerning the precise relationship between nucleus

basalis efferents and various types of cortical neurons is

limited and appears complex. Whereas in motor cortex

pyramidal neurons of layer 5 are likely to receive

cholinergic innervation from the nucleus basalis, in primary

sensory cortex granular cells of layer 4 appear to be the

principle recipients (Koliatsos et al., 1990).

In addition to the cholinergic innervation from the

nucleus basalis, the rodent neocortex also possesses

intrinsic cholinergic neurons (Fibiger, 1982; Johnston et










al., 1981; Smith, 1988). These cholinergic neurons are

bipolar in morphology and exist in all layers, although they

tend to congregate in layers 2 and 3 (Eckenstein and

Baughman, 1984; Levey et al., 1984). The axonal projections

of intrinsic cholinergic neurons remain to be elucidated.

The presence of intrinsic cholinergic neurons in human or

other primate cerebral cortices has not been demonstrated.

The hypothesis that chronic cortical cholinergic

hypofunction may be sufficient to induce trans-synaptic cell

loss is supported by our observation that bilateral nucleus

basalis magnocellularis-lesions significantly reduce neuronal

density in layers 2 and 6, by 8 months post-lesioning. This

is the earliest lesion-induced, trans-synaptic cell loss

reported to date, although Dr. Arendash cites similar cell

losses as early as 5 months post-lesioning in personal

communications. This deficit in Nissl staining cell bodies

is similar in extent as that observed 14 months post-

lesioning (Arendash et al., 1987), suggesting that the trans-

synaptic cell loss has reached a plateau by 8 months post-

lesioning. The question remains whether this plateau

represents the maximum degree of lesion-induced neocortical

atrophy, or whether chronic neocortical cholinergic

hypofunction will result in more extensive neocortical

atrophy at time points greater than 14 months post-lesioning.

It should be noted that in this study Nissl staining

cell bodies were only quantified in layers 2 and 6 of the

parietal cortex. Therefore, changes in other brain regions










or cortical layers may be different from those described

here. Currently studies are in progress to examine the long-

term effects of nucleus basalis-lesions on neuronal density

in other areas of the neocortex. Of particular interest is

the assessment of neuronal density within cortical areas that

do not receive cholinergic projections from the nucleus

basalis magnocellularis, such as the cingulate cortex,

pyriform cortex, and posterior neocortex.

Unlike Alzheimer's disease where there appears to be a

selective loss of large pyramidal neurons, nucleus basalis-

lesions do not appear to induce a selective loss of large

neurons at 8 months post-lesioning, since no significant

shift in the neuronal size distribution curve is observed.

However, the neuronal size distribution has only been

measured in layer 2 of the parietal cortex; other layers may

yet reveal lesion-induced shifts. Furthermore, neuronal size

distribution has only been measured 8 months post-lesioning,

again allowing for potential lesion-induced changes at later

time points.

The morphological identity of the neurons lost following

nucleus basalis-lesions remains to be investigated. Though

Nissl staining allows easy assessment of changes in neuronal

density, little information can be gathered concerning the

morphological features of the neurons (i.e. pyramidal versus

granular, polymorphic multiform, horizontal, or martinotti

cells). To gather histologic information concerning neurons,

their processes, and intracortical relationships special









silver impregnation techniques, such as the Golgi method, are

more appropriate.

Investigation of neuronal cell loss in individual layers

is important, since the different layers are involved in

different cortical functions (Pansky et al., 1988). In

general the cortical layers 1 to 3, also known as the

molecular, external granular, and external pyramidal layers,

are particularly important for association and higher

functions such as memory, interpretation of sensory input,

and certain discriminating functions. Layer 4, also known as

the internal granular layer, is mainly a receptive layer

where the thalamocortical fibers primarily terminate. Layers

5 and 6, also known as the ganglionic and multiform layers,

are primarily efferent layers that contain nerve cell bodies

whose axons enter the corticospinal tract. Our studies

concentrated on neuropathological changes in the parietal

cortex, since the greatest degree of neuronal loss was

observed in this area, 14 months post-lesioning (Arendash et

al., 1987). We studied layers 2 and 6, since these layers

are the most readily identifiable layers in brain sections

stained with cresyl violet dye (a Nissl-selective stain).

We hypothesized that the loss of neocortical neurons

subsequent to the loss of cholinergic transmission may induce

a selective loss of those cholinergic receptors used to

mediate this transmission between the ascending cholinergic

and neocortical neurons. A likely candidate for this

receptor-subtype is the nicotinic receptor since its levels,










unlike muscarinic receptor levels, are consistently and

dramatically reduced in the neocortex of Alzheimer's patients

(Araujo et al., 1988a; Flynn and Mash, 1986; Nordberg and

Winblad, 1986a; Perry et al., 1987; Shimohama et al., 1986;

Whitehouse et al., 1985a; Whitehouse et al., 1986). Studies

regarding the status of muscarinic receptors in Alzheimer's

disease have yielded conflicting results. Depending on the

study muscarinic receptor levels appear to remain unchanged

(Davies and Verth, 1978; DeSarno et al., 1988; Kellar et al.,

1987; Waller et al., 1986; Whitehouse et al., 1985b),

significantly elevated (Danielsson et al., 1988; DeSarno et

al., 1988; Nordberg and Winblad, 1986b; Nordberg et al.,

1983), or reduced (Reinikainen et al., 1987) in Alzheimer's

disease.

Our results show that nicotine receptor binding is

initially elevated 5 months following nucleus basalis-

lesions, only to return to sham-operated control levels by 8

months post-lesioning. The elevation in receptor binding is

most likely due to an increase in nicotine receptor

concentrations, since studies by Schwartz and Kellar (1985)

found no change in Kd following similar nucleus basalis

lesions. Nucleus basalis magnocellularis-lesions thus appear

to exert a complicated, time-dependent effect on neocortical

nicotine receptor concentrations. The initial rise in

nicotine receptor levels may be due to post-synaptic

denervation supersensitivity typically associated with the

loss of agonist stimulation. The apparent lesion-induced









elevation in nicotine receptor concentrations may have failed

to persist due to the neocortical atrophy also associated

with the nucleus basalis magnocellularis-lesions. A more

detailed comparison between the time-dependent changes in

nicotine receptor concentrations and the degree of

neocortical neuropathology will help to evaluate the validity

of the latter hypothesis.

Our hypothesis that chronic cortical cholinergic

hypofunction may be sufficient to induce trans-synaptic

changes in neocortical neuropeptide Y levels is supported by

the observation that bilateral nucleus basalis-lesions

significantly elevate neuropeptide Y levels by 10 months

post-lesioning, and that unilateral lesions significantly

reduce neuropeptide Y levels by 1 month post-lesioning. The

trans-synaptic effects of cholinergic transmission appear

complex. Clearly the direction of change depends on the type

of lesion, bilateral versus unilateral, which raises the

question of the role of inter-hemispheric communication in

the neuropeptide Y response to cholinergic deafferentation

(as is known to occur).

Following our discovery that long-term neocortical

cholinergic hypofunction induces an elevation in neocortical

neuropeptide Y levels, we proceeded to examine the mechanism

underlying this increase. As discussed in chapter one,

changes in peptide levels may be due to changes in the level

of synthesis or breakdown of the peptide. We hypothesized

that the dramatic elevations in neocortical neuropeptide Y









levels in rats with bilateral nucleus basalis-lesions may be

preceded by an increase in the degree of transcription of the

neuropeptide Y gene. We used neuropeptide Y-encoding mRNA

levels as a measure of the level of transcription of the

neuropeptide Y gene. Our studies indicate that changes in

the degree of neuropeptide Y gene transcription do not

consistently correspond to similar changes in neocortical

neuropeptide Y levels in rats with unilateral or bilateral

nucleus basalis-lesions at the time points investigated.

The lack of a correlation between peptide levels and

levels of the corresponding mRNA is not an uncommon

phenomena. Several other peptidergic neurotransmitter

systems investigated, including the opioid system (Howells et

al., 1986), fail to demonstrate a consistent correlation

between peptide and mRNA levels. The robust elevation in

neuropeptide Y levels induced by long-term bilateral nucleus

basalis-lesions appears to occur between 8 and 10 months

post-lesioning. Though levels of neuropeptide Y-encoding

mRNA are not elevated by 8 months post-lesioning, the

possibility remains that the robust neuropeptide Y elevations

may correlate with increases in neuropeptide Y gene

transcription since levels of neuropeptide Y-encoding mRNA

have not been investigated at 10 months post-lesioning.

As mentioned in the introduction, the early literature

and unpublished data available (data limited to 2 months and

14 months post-lesioning) suggested the that the lesion-

induced increase in neocortical neuropeptide Y coincides with










the onset of behavioral recovery (Arendash et al., 1987). We

therefore hypothesized that the neocortical neuropeptide Y

elevation may in part mediate the behavioral recovery.

However, our time course studies indicate that the behavioral

recovery is completed (between 5 and 8 months post-lesioning)

before elevations in neuropeptide Y (between 8 and 10 months

post-lesioning) are observed. It therefore seems unlikely

that the recovery in cognitive function is mediated by

neuropeptide Y. Nevertheless we cannot completely exclude a

role of neuropeptide Y in the recovery of cognitive function,

since we have not investigated the status of neuropeptide Y

release. The possibility remains that neuropeptide Y release

from neocortical neurons is elevated, or reduced, over a

similar time course as the behavioral recovery occurs

(between 5 and 8 months post-lesioning).

Our studies now suggest that the lesion-induced

neocortical cell loss may coincide with the onset of

behavioral recovery, since both appear to occur between 5 and

8 months post-lesioning. It therefore appears that the

neocortical atrophy is more likely to be involved in the

recovery of cognitive function than the neocortical

neuropeptide Y elevations. Unfortunately, as noted

previously, the neurochemical nature of the cell loss is

unclear at present, with no decrease seen in the

neurotransmitter systems to date including neuropeptide Y,

somatostatin, corticotropin releasing hormone, dopamine,

serotonin (Arendash et al., 1987), and glutamate (personal









communications from Dr. Ralph Dawson, Department of

Pharmacodynamics, University of Florida).

It has been suggested that GABAergic neurons are most

likely to account for the lesion-induced loss of Nissl

staining cell bodies in the neocortex. GABAergic inhibitory

interneurons are the most numerous neurons in the cerebral

cotex. Approximately 50% of the neurons in the cerebral

cortex are GABAergic in nature (Curtis and Johnston, 1974;

Nistri and Constanti, 1979). It is therefore very likely

that GABAergic neurons are among those lost following nucleus

basalis-lesions, particularly since the levels of glutamate,

the second most abundant neurotransmitter in the cerebral

cortex (approximately 30% of cerebral cortical neurons are

glutamatergic in nature; Curtis and Johnston, 1974; Nistri

and Constanti, 1979; Roberts, 1974), are not affected by

nucleus basalis-lesions. The lesion-induced loss of

cholinergic input may increase the level of inhibitory

activity of GABAergic neurons, directly or indirectly, which

in turn may interfere with passive avoidance behavior. A

decrease in the number of GABAergic neurons could reduce the

level of the GABA-mediated inhibitory influence and perhaps

ultimately restore normal passive avoidance behavior.

This hypothesis is supported by our observation that

chronic nicotine treatment ameliorated the lesion-induced

cell loss in rats with bilateral nucleus basalis-lesions, but

that these rats continue to suffer passive avoidance

behavioral deficits. The observation that administration of