JUKKA PUOLIVÄLI. An experimental study on the cholinergic modulation of cortical arousal and cognitive functions. With special emphasis on apolipoprotein E. Series of Reports, No 51, Department of Neurology, University of Kuopio. 2000.

ISBN 951-781-743-6
ISSN 0357-6043

ABSTRACT

The brain cholinergic system is involved in a number of behaviors including arousal and cognitive processes. In Alzheimer’s disease (AD) patients, the function of the cholinergic system is severely compromised. AD patients with apolipoprotein E (apoE) e4 allele(s) suffer more pronounced cholinergic deficits which are associated with severe cognitive impairments, pronounced cortical electroencephalogram (EEG) slowing, impaired synaptic plasticity, and decreased responsiveness to cholinomimetic treatment compared to AD patients without apoE e4 allele. The purpose of this study was to investigate the function of the brain cholinergic system and its interaction with apoE in the regulation of cortical arousal and cognitive functions. First, a rat EEG model of thalamocortically generated high-voltage spindles (HVS) was used to study the cholinergic modulation of neocortical arousal. The effects of intrathalamic (reticular nucleus of thalamus (NRT) and ventroposteromedial nucleus of thalamus (VPM)) infusions of a muscarinic M₁ receptor agonist, McN-A-343, a muscarinic M₂ receptor antagonist, methoctramine, and a muscarinic receptor agonist, oxotremorine, on HVSs were studied. Second, the role of apoE in the regulation of biochemical, cognitive and electrophysiological functions was investigated. The effects of quisqualic acid induced nucleus basalis (NB) -lesion and ageing on cholinergic activity, spatial navigation in water maze (WM), cortical EEG activity and EEG response to a muscarinic receptor antagonist, scopolamine, were investigated in apoE-deficient and control mice. The major findings of this study were: 1) intrathalamic NRT and/or VPM infusions of McN-A-343, methoctramine, and oxotremorine, decreased HVSs in rat, indicating increased neocortical arousal. These results suggest that intrathalamic infusions of cholinergic muscarinic drugs modulate neocortical arousal via muscarinic M₁ and M₂receptors in thalamus. Furthermore, these results suggest that intracerebroventricularly and systemically administered cholinergic drugs may also modulate neocortical arousal via the thalamus. 2) The apoE-deficient mice did not suffer any apparent impairment in their cortical and hippocampal choline acetyltransferase (ChAT) -activity, number of ChAT-positive neurons in NB, or in WM spatial navigation during ageing. Furthermore, apoE-deficiency did not increase the sensitivity to cholinergic, WM spatial navigation or EEG deficits induced by NB-lesion. However, the apoE-deficient mice had slightly altered cortical EEG activity during ageing and blunted EEG response to scopolamine treatment. These results suggest that apoE does not have to be present to preserve the viability of cholinergic neurons and that spatial navigation and behavioral recovery during ageing or after NB-lesion is not affected by apoE-deficiency. However, apoE-deficiency might alter the regulation of cortical arousal during ageing and after cholinergic drug (scopolamine) manipulation. Taken together, this study provides new information about the function of the brain cholinergic system and its interaction with apoE. This may have relevance in the development of new experimental models and therapeutic treatments for clinical disorders such as AD.

National Library of Medicine Classification: WL 150, WL 155

Medical Subject Headings: Alzheimer’s disease; apolipoproteins E; arousal; cholinergic agents; electroencephalography; maze learning; memory; nucleus basalis of Meynert

To my parents

ACKNOWLEDGEMENTS

This work was performed in the Department of Neuroscience and Neurology during the years 1995-2000.

I wish to thank Docent Paavo Riekkinen Jr. and my supervisors Docent Pekka Jäkälä; and Docent Riitta Miettinen for their guidance and instructions during these years.

I wish to thank Docent Jouni Sirviö; and Professor Dag Stenberg, the official reviewers of this study, for their constructive criticism and suggestions to improve the manuscript.

I also express my gratitude to Professor Hilkka Soininen, Docent Heikki Tanila, Docent Aarne Ylinen, Docent Antero Salminen, Docent Irina Alafuzoff, Doctor Thomas van Groen, Professor Tuula Pirttilä; and Professor Juhani Sivenius, for their excellent teaching in neuroscience.

My thanks to Doctor Laurent Pradier for his significant contribution to the study. I owe my thanks to Anna-Liisa Gidlund and Päivi Räsänen for their first-rate technical assistance. I would also like to thank Esa Koivisto, Sari Palviainen, Mari Tikkanen and Hanna Turkki for their significant help during these years, and Doctor Ewen MacDonald for revising the language of the manuscript. I am also grateful to the personnel of National Laboratory Animal Center of the University of Kuopio.

My deepest and warmest thanks to Juhana Aura, Markus Björklund, Mikko Hiltunen, Jouni Ihalainen, Sami Ikonen, Petri Kerokoski, Petri Kolehmainen, Pauliina Korhonen, Erkki Kuusisto, Mia Mikkonen, Maaria Roschier, Tero Tapiola and Iain Wilson for their friendship and creating an encouraging and inspiring community spirit among us students.

I also wish to thank my post-doc friends Seppo Helisalmi and Raimo Pussinen for the fruitful academic discussion.

I wish to thank the players of our sähly team for the victorious and glorious moments that I shared with you. I also like to thank Markku for nice hiking company and Tommi and Anne-Mari for every Tuesday’s swimming practice and refreshing Y.E.T. evenings.

I would also like to thank my sister Päivi for her love and support.

My dearest thanks to Mia for her love and encouragement.

Finally, I owe my deepest gratitude to my parents, Mirja and Niilo, for their love and support during these years.

This study was financially supported by the European Commission (Biomed 2 grant), the Finnish Academy of Sciences, the Finnish Cultural Foundation of Northern Savo, the Kuopio University Foundation, and the University of Kuopio.

Kuopio, September 2000

Jukka Puoliväli

ABBREVIATIONS

Ab	b-amyloid
AChE	acetylcholinesterase
AD	Alzheimer’s disease
AMPA	a-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid
ANOVA	analysis of variance
ApoE	apolipoprotein E
APP	amyloid precursor protein
ChAT	choline acetyltransferase
CNS	central nervous system
EEG	electroencephalogram
GABA	gamma-aminobutyric acid
HACU	high affinity choline uptake
HDB	horizontal diagonal band of Broca
HVS	high-voltage spindle
i.c.v.	intracerebroventricular
IgG	immunoglobulin G
i.p.	intraperitoneal
LC	locus coeruleus
LDT	laterodorsal tegmental nucleus
LTP	long-term potentiation
MANOVA	multivariate analysis of variance
MAP-2	microtubule-associated protein 2
MS	medial septum
NB	nucleus basalis
NRT	reticular nucleus of thalamus
PPT	pedunculopontine tegmental nucleus
PS-1	presenilin-1
PS-2	presenilin-2
THA	tetrahydroaminoacridine
VDB	vertical diagonal band of Broca
VPM	ventroposteromedial nucleus of thalamus
WM	water maze

LIST OF ORIGINAL PUBLICATIONS

This thesis is based on the following original publications that are referred to in the text by the Roman numerals I-IV.

I Puoliväli J, Jäkälä P, Koivisto E, and Riekkinen P Jr.: Muscarinic M1 and M2 receptor subtype selective drugs modulate neocortical EEG via thalamus. NeuroReport 9: 1685-1689, 1998.

II Puoliväli J, Jäkälä P, Koivisto E, and Riekkinen P Jr.: Oxotremorine suppresses thalamocortical oscillations via thalamic muscarinic acetylcholine receptors. Psychopharmacology 140: 285-292, 1998.

III Puoliväli J, Pradier L, and Riekkinen P Jr.: Moderate cortical EEG changes in apolipoprotein E-deficient mice during ageing and scopolamine treatment but not after nucleus basalis lesion. Psychopharmacology 145: 386-392, 1999.

IV Puoliväli J, Miettinen R, Pradier L, and Riekkinen P Jr.: Apolipoprotein E-deficient mice are not more susceptible to the biochemical and memory deficits induced by nucleus basalis lesion. Neuroscience 96: 291-297, 2000.

CONTENTS

1. INTRODUCTION

2. REVIEW OF THE LITERATURE

2.1 THE BRAIN CHOLINERGIC SYSTEM

2.1.1 Anatomy of the brain cholinergic system
2.1.2 Pharmacology of the cholinergic synapse
2.1.3 Neurophysiology of the brain cholinergic system
2.1.4 Cognitive and attentional effects of cholinergic manipulations
2.1.5 Brain cholinergic system during ageing and in Alzheimer's disease

2.2 APOLIPOPROTEIN E

2.2.1 ApoE structure and function
2.2.2 ApoE and Alzheimer's disease
2.2.3 ApoE, the cholinergic system and synaptic plasticity in Alzheimer's disease
2.2.4 ApoE-mutant mice

3. AIMS OF THE STUDY

4. MATERIALS AND METHODS

4.1 ANIMALS

4.2 IMPLANTATION OF EEG ELECTRODES AND INFUSION CANNULAS

4.3 NUCLEUS BASALIS LESION

4.4 EEG RECORDINGS

4.5 DRUG STUDIES

4.6 BEHAVIOR

4.7 BIOCHEMISTRY

4.8 HISTOLOGY AND IMMUNOHISTOCHEMISTRY

4.9 STATISTICS

5. RESULTS

5.1 EEG RECORDINGS IN RATS

5.1.1 High-voltage spindle activity
5.1.2 Total recording time

5.2 EEG RECORDINGS IN APOE-DEFICIENT MICE


5.2.1 Nucleus basalis lesion study
5.2.2 Ageing study

5.3 BEHAVIOR

5.4 BIOCHEMISTRY

5.5 HISTOLOGY AND IMMUNOHISTOCHEMISTRY

6. DISCUSSION

6.1 METHODOLOGICAL CONSIDERATIONS

6.1.1 Nature of transgenic and control mice
6.1.2 Quisqualic acid induced nucleus basalis lesion
6.1.3 EEG recordings in rats and mice
6.1.4 Behavioral testing in water maze

6.2 CHOLINERGIC MODULATION OF CORTICAL AROUSAL IN RATS

6.3 NEUROPHYSIOLOGICAL FINDINGS IN APOE-DEFICIENT MICE

6.4 BEHAVIORAL FINDINGS IN APOE-DEFICIENT MICE

6.5 BIOCHEMICAL AND NEUROPATHOLOGICAL FINDINGS IN APOE-DEFICIENT MICE

7. CONCLUSIONS

REFERENCES

---

1. INTRODUCTION

Dementia is defined as the development of multiple cognitive deficits that include memory impairment and at least one of the following: aphasia, apraxia, agnosia, or a disturbance in executive functioning (American Psychiatric Association, 1994). The prevalence of Alzheimer's disease (AD), the most common cause of dementia, doubles every five years after the age of 60, rising from a prevalence of 1 % among 60-years-olds to greater than 40 % among 85-years-olds. AD is characterized by unknown etiology and a progressive course of the disease (Cummings et al. 1998). Neuritic plaques and neurofibrillary tangles are typical histopathological findings in AD (Cummings et al. 1996, Arendt et al. 1998). These lesions are predominantly located in the cerebral cortex, although they are also observed in certain subcortical structures. In addition to these two classical markers, massive loss of synapses (Terry et al. 1991) and neurons (Gomez-Isla et al. 1997) in the hippocampal formation and association regions of the neocortex is observed in AD patients.

Both genetic and environmental factors are associated with AD. Mutations in amyloid beta precursor protein (APP), presenilin-1 (PS-1) and presenilin-2 (PS-2) genes result in increased production of b-amyloid (Ab) peptide which is deposited in the brain. These gene defects are associated with early-onset familial AD (Cummings et al. 1998). In addition, several other genes, including the e4 allele of apolipoprotein E (apoE) gene, have been found to increase the risk for AD (Strittmatter and Roses 1996). Other risk factors for AD include advanced age and history of head trauma. On the other hand, higher levels of education, the apoE e2 allele, and the use of nonsteroidal anti-inflammatory drugs and estrogen replacement therapy in post-menopausal women, might work as "protective" factors and decrease the risk for and delay the onset of AD (Breitner et al. 1995, Cummings et al. 1998).

AD is also associated with dysfunction of several neurotransmitter systems, including the cholinergic (Whitehouse et al. 1982), noradrenergic (Mann 1983, Marcyniuk et al. 1986) and serotonergic systems (Palmer et al. 1987, Nazarali and Reynolds 1992). The brain cholinergic system is involved in a number of behaviors, including arousal, attention, sleep-wake cycles as well as learning and memory. The basal forebrain and brainstem cholinergic systems play also an important role in the regulation of cortical and thalamic electrical activity (McCormick 1990, McCormick 1992b). In AD, the loss of cortical cholinergic markers and the degeneration of basal forebrain cholinergic neurons are the most consistent and severe neurochemical deficits. The severity of the cholinergic deficits has been shown to correlate with the degree of cognitive decline (Bartus et al. 1982) and electroencephalogram (EEG) slowing in AD patients (Soininen et al. 1992). Those AD patients carrying the apoE e4 allele suffer more profound cholinergic pathology (Poirier 1994, Soininen et al. 1995) which is associated with more severe cognitive impairment (Lehtovirta et al. 1996b), more pronounced EEG slowing (Lehtovirta et al. 1996a, Lehtovirta et al. 2000), and decreased responsiveness to cholinesterase inhibitor therapy (Poirier et al. 1995) when compared to AD patients without the apoE e4 allele. Furthermore, AD patients with the apoE e4 allele have an earlier onset of the disease (Strittmatter and Roses 1996) and impaired neuronal remodelling capacity (Arendt et al. 1997).

The in vivo rodent models of cholinergic hypofunction have been useful in clarifying the role of the brain cholinergic system in cognitive and electrophysiological functions (Wenk 1997). Recently, transgenic animal models of AD that have modifications in APP, PS and apoE genes have been developed (Hsiao et al. 1996, Holcomb et al. 1999, Hamanaka et al. 2000). These models make it possible to determine the role of specific molecules in the pathogenesis and neurodegenerative processes of AD, and, most importantly, to test novel therapies in the treatment of AD. In the present study, the effects of cholinergic manipulation on cortical arousal and cognitive functions were investigated. Moreover, the role of apoE in the regulation of biochemical, behavioral, and electrophysiological functions were studied.

2. REVIEW OF THE LITERATURE

2.1 THE BRAIN CHOLINERGIC SYSTEM

2.1.1 Anatomy of the brain cholinergic system

The distribution of cholinergic cells and processes in the rat brain is well characterized (Mesulam et al. 1983, Wainer and Mesulam 1990, Butcher 1995). However, there are fewer anatomical studies on the mouse brain cholinergic system (Hohmann and Ebner 1985, Kitt et al. 1994). The brain cholinergic system has two major anatomically different regions: the cholinergic nuclei in the basal forebrain and the cholinergic nuclei in the brainstem. The cholinergic nuclei of the basal forebrain consist of large magnocellular cholinergic cells situated in the medial septum (MS), the vertical and horizontal branches of the diagonal band of Broca (VDB and HDB, respectively) and the nucleus basalis (NB). The brainstem cholinergic cell groups are situated in the pedunculopontine tegmental nucleus (PPT), the laterodorsal tegmental nucleus (LDT), the medial habenular nucleus and the parabigeminal nucleus (Wainer and Mesulam 1990). In addition, striatal intrinsic cholinergic neurons can be found in both rodents and primates. Although cortical ChAT-positive neurons have been found in the rat, those neurons are not associated with ChAT mRNA. This finding has indicated that intrinsic cholinergic neurons do not exist in rat cortex, which is consistent with findings in other mammals (Butcher 1995).

Both mice and rats show regional variations in the distribution and densities of cholinergic innervation. The major source of cholinergic innervation of the hippocampus derives from the MS and VDB. The patterns of cholinergic innervation of the hippocampus seem to be quite similar in rats and mice (Wainer and Mesulam 1990, Kitt et al. 1994). HDB neurons project mainly to olfactory bulb, and the NB neurons project to the entire cerebral cortex and to amygdala (Wainer and Mesulam 1990). In mice, the cholinergic axons in the somatosensory cortex are prominent in layers I and IV, and within the lower portion of layer V and the upper segment of layer VI (Kitt et al. 1994). A similar cholinergic innervation pattern in rat sensory cortex has also been reported (Lysakowski et al. 1986). However, Eckenstein et al. (1988) reported that in rat sensory cortex the layers I-III and the upper portion of layer V have the highest densities of cholinergic innervation. The potential differences in the laminar structure of cortical innervation between mice and rats suggest that there might be some differences in the ways in which acetylcholine modulates cortical activity in these two species (Kitt et al. 1994). Moreover, there is a strong hereditary variability within the cholinergic system in mouse strains. For example, the number of cholinergic neurons in MS and VDB, and the density of cholinergic fibers in hippocampus differ significantly between different inbred mouse strains (Ingram and Corfman 1980, Schwegler et al. 1996b), which might explain some of the behavioral variation observed between strains (Schwegler et al. 1996a).

Figure 1. Sagittal view of basal forebrain and brainstem ascending cholinergic pathways in rat (modified from Mesulam et al. 1983). Abbreviations: AMG = amygdala; CB = cerebellum; H = hippocampus; HDB = horizontal diagonal band of Broca; LDT = laterodorsal tegmental nucleus; MS = medial septum; NB = nucleus basalis; NC = neocortex; OB = olfactory bulb; PPT = pedunculopontine tegmental nucleus; TH = thalamus; VDB = vertical diagonal band of Broca.

The basal forebrain cholinergic neurons also provide innervation to the thalamus. In rats, over one third of the total number of basal forebrain neurons projecting to reticular nucleus of thalamus (NRT) are cholinergic (Levey et al. 1987). However, the major source of cholinergic innervation to the thalamus derives from the cholinergic neurons of the PPT and LDT in the brainstem (Wainer and Mesulam 1990). The cholinergic innervation of the thalamus is virtually identical in mice and rats. In both species, the anteroventral, intralaminar and reticular nuclei show intense choline acetyltransferase (ChAT) immunoreactivity (Levey et al. 1987, Kitt et al. 1994). There are also less dense connections from the brainstem cholinergic nuclei to parts of the cerebral cortex, basal forebrain, and a number of extrapyramidal structures, such as the striatum, globus pallidus, and substantia nigra (Butcher 1995).

Although the basal forebrain cholinergic neurons project to many cortical areas, they receive projections from only a very limited set of limbic and paralimbic cortical areas (Wainer and Mesulam 1990). There is input to basal forebrain cholinergic neurons from the amygdala, the hypothalamus and the brainstem nuclei, including the PPT, the LDT, the substantia nigra, the ventral tegmental area, the raphe nuclei, and the locus coeruleus (LC) (Koliatsos et al. 1990, Wainer and Mesulam 1990).

2.1.2. Pharmacology of the cholinergic synapse

The synthesis of acetylcholine from choline and acetyl coenzyme A takes place in nerve terminals of cholinergic cells. The formation of acetylcholine is catalysed by ChAT, which is produced in the neuronal soma and transported by slow axonal transport down the axon to the synapses. In the cholinergic nerve terminals, the acetylcholine is stored either in the cytosol or in the synaptic vesicles from where it is released. The immediate source of the calcium dependent, evoked release of acetylcholine is the vesicular rather than the cytosolic fraction. The hydrolysis of acetylcholine into choline and acetate is catalysed by acetylcholinesterase (AChE) and takes place in the synaptic cleft. Specific acetylcholinesterase inhibitors like donepezil, galanthamine, tetrahydroaminoacridine (THA), physostigmine, rivastagmine and metrifonate, can be used to increase the amount of available acetylcholine in the cholinergic synapse (Francis et al. 1999).

The acetylcholine receptors are divided into muscarinic and nicotinic acetylcholine receptors. Different subtypes of muscarinic acetylcholine receptors have been identified by using both pharmacological and molecular techniques. On the basis of pharmacological properties, four subtypes of muscarinic acetylcholine receptors (M₁-M₄) have been identified, whereas molecular cloning has revealed five different muscarinic acetylcholine receptor subtypes (m₁-m₅). Muscarinic acetylcholine receptors transduce their signals by coupling with G-proteins, which then modulate the activity of a number of effector enzymes and ion channels. The stimulation of muscarinic receptor M₁, M₃ and m₅ subtypes can activate multiple signalling effectors simultaneously including phospholipases C, D and A₂ as well as different ion channels. The receptor-mediated activation of phospolipases results in activation of several different second messenger systems. For example, the activation of phospholipase C results in the release of inositol trisphosphate and diacylglycerol. The main cellular response to muscarinic M₂ and M₄ receptor stimulation is the inhibition of adenylate cyclase, which reduces the levels of cyclic AMP Ehlert et al. 1995; Felder 1995).

In the rat brain, the M₁ receptor is most abundant in cortex and gradually declines in number caudally towards spinal cord. For example, the thalamus contains fewer M₁ receptors (Wall et al. 1991, Wei et al. 1994). The M₂ receptor is most abundant in cerebellum, pons/medulla, and thalamus/hypothalamus. In contrast, forebrain regions (cortex, hippocampus, striatum and olfactory bulb) contain markedly fewer M₂ receptors (Li et al. 1991, Wei et al. 1994). The highest levels of the M₄ receptors are found in the striatum and in lower amounts in cortex and hippocampus (Waelbroeck et al. 1990, Wei et al. 1994). Also in mouse brain, the M₁ receptors are abundant in cortex and hippocampus, M₂ receptors in thalamus, and M₄ receptors in striatum (Hohmann et al. 1995). In hippocampus and several regions of neocortex in human brain, the M₁ receptor constitutes from 35-60 % of all muscarinic receptors, whereas M₂ and M₄ receptors each account for about 15-25 % of the muscarinic receptors in the same areas. In contrast, the M₂ receptor is the predominant subtype in the basal forebrain, and M₄ receptor is the most abundant in the caudate and putamen (Flynn et al. 1995, Levey 1996). The M₃ and m₅ receptors are found consistently throughout the brain, although for M₃ there is a modest rostral-caudal decline (Wei et al. 1994). Muscarinic receptors can be located on both pre- and post-synaptic sides. The M₁ receptors are largely post-synaptic and their activation increases the cholinergic neurotransmission. In contrast, the M₂ receptors are predominantly presynaptic autoreceptors, which control the cholinergic tone in negative manner. Much less is known about the precise location of muscarinic M₃, M₄ and m₅ receptor subtypes (Levey 1996).

Nicotinic acetylcholine receptors belong to the group of ligand-gated receptors. Nicotinic receptors are composed of several combinations of a- and b-subunits with considerable molecular diversity, suggesting that many functional subtypes of neuronal nicotinic receptors are possible. Nicotinic receptors are abundant in cerebral cortex, thalamus, hippocampus and hypothalamus (Arneric et al. 1995).

Table 1. Acetylcholine receptor subtypes: main locations in the brain and examples of second messenger systems and biochemical mediators. Abbreviations: cAMP = cyclic adenosine monophosphate; CNS = central nervous system; DAG = diacylglycerol; IP₃ = inositol trisphosphate (Arneric et al. 1995, Ehlert et al. 1995, Felder 1995).

Receptor subtype	Brain region	Second messenger system	Biochemical mediator
M1	Hippocampus, cortex, olfactory bulb, striatum	Phospholipase C activation	DAG, IP3
M2	Cerebellum, pons/medulla, thalamus/hypothalamus	Adenylate cyclase inhibition	cAMP
M3	Cortex, hippocampus, striatum, cerebellum	Phospholipase C activation	DAG, IP3
M4	Striatum, cortex, hippocampus	Adenylate cyclase inhibition	cAMP
m5	Evenly distributed throughout the CNS	Phospholipase C activation	DAG, IP3
Nicotinic	Cortex, hippocampus, thalamus, hypothalamus	Ligand-gated ion channel	Na+, K+, Ca2+ conductance

2.1.3 Neurophysiology of the brain cholinergic system

The brain cholinergic system is involved in the generation of various EEG patterns, and pharmacological manipulations or pathological changes in the cholinergic system affect neocortical EEG activity. Different EEG rhythms are related to distinct behavioral states. The slow-waves (delta waves) with a frequency below 4 Hz prevail during the deep stage of normal sleep, whereas the normal theta activity within the frequency range 4-8 Hz occurs during activated behavioral states. Other EEG patterns include for example alpha rhythm that occurs during wakefulness and EEG fast-waves (beta activity) that occur during epochs of increased alertness (Steriade et al. 1990). Also different patterns of synchronized spindle oscillations, which take place during the early stages of sleep or low arousal and vigilance can be found in cortical EEG activity of humans and experimental animals (Buzsáki et al. 1988a, Steriade et al. 1990, Riekkinen et al. 1991b). In pathological situations, such as AD, the regulation of EEG activity is altered. The decline of cholinergic markers is correlated with EEG slowing observed in patients with AD (Soininen et al. 1992). The slowing of EEG is characterized by enhanced theta power in patients with mild and moderate AD, and delta waves increase in more advanced stages of AD. Moreover, the EEG slowing in AD is also manifested by a decrease of alpha and beta activities (Coben et al. 1983, Penttilä et al. 1985).

The ascending cholinergic systems of basal forebrain and brainstem regulate the neocortical and thalamic electrical activity (McCormick 1992b). The frequency changes in NB neurons correlate well with changes in behavioral and neocortical arousal of the animal. In rats, the highest frequency activity of NB neurons has been observed during running, followed by drinking, and immobility (Buzsáki et al. 1988a). NB neurons are also active when the animal is waiting to respond to visual or auditory stimuli that are associated with the expectation of a food reward (Richardson and DeLong 1991). The increase in activity of NB neurons results in neocortical desynchronization, whereas the decrease in activity results in increased slow-wave activity in the neocortex (Wenk 1997, Détári et al. 1999). In a recent study, it was also shown that the discharge pattern of rat basal forebrain cholinergic neurons shifted from an irregular tonic discharge pattern during EEG slow irregular activity to a rhythmic burst discharge pattern during EEG rhythmic slow activity (Manns et al. 2000). These results indicate that the cholinergic basal forebrain neurons have the capacity to modulate their cortical target areas rhythmically during cortical activation (Manns et al. 2000).

The electrophysiological effects of acetylcholine are mediated via cholinergic receptors. Muscarinic receptor activation results in a variety of hyperpolarizing and depolarizing responses, dependent upon the cell type or the tissue under investigation (Jones 1993). For example, in cortex, the extracellular application of acetylcholine to cortical neurons has been found to result in both excitation as well as inhibition. In cortical pyramidal cells, acetylcholine induces a slow excitatory response associated with an increase in neuronal excitability. This is due to suppression of potassium current, which appears to be mediated through muscarinic M₁ receptors (McCormick 1990, McCormick 1992b). On the other hand, the activation of muscarinic M₂ receptors in cortical GABAergic inhibitory interneurons by acetylcholine results in depolarization of the neurons through activation of cation conductance. This can lead to inhibition of cortical pyramidal neurons (McCormick 1990, McCormick 1992b). Thus, the cholinergic system regulates the excitability of cortical pyramidal neurons both directly and indirectly via cortical GABAergic interneurons. This is in accordance with the anatomical findings showing that basal forebrain cholinergic neurons have synapses both with cortical pyramidal neurons and GABAergic interneurons (Beaulieu and Somogyi 1991).

The thalamic neurons can exhibit two basic modes of action potential generation: single spike activity and rhythmic burst generation (McCormick 1990). The activity mode of thalamic neurons correlates with changes in behavior and arousal of the animal. Typically, rhythmic activity occurs only during states of low arousal and vigilance, being virtually absent during high vigilance states. When the animal is aroused, the increased activity of the ascending systems, such as the basal forebrain and brainstem cholinergic systems, is thought to suppress rhythmic burst firing, and the typical cortical electrical activity is characterized by low-voltage fast-waves. Whereas in the drowsy or sleeping animal, large-amplitude slow-waves are observed in the delta band (Buzsáki et al. 1988b). Furthermore, in waking-immobile rats, a highly synchronized EEG pattern, characterized by large-amplitude rhythmic waves in theta/alpha band is observed. It is believed that the loop between thalamocortical relay neurons and GABAergic neurons in NRT is responsible for the generation of rhythmic high-voltage spindle (HVS) wave activity (Figures 2 and 3). During the spindle activity, rhythmically active GABAergic neurons in NRT can phasically hyperpolarize their thalamocortical target neurons. In the absence of other depolarizing inputs, voltage- and time- dependent rebound Ca²⁺ spike bursts occur in thalamocortical relay neurons in a phase-locked manner. The rhythmical bursts of the thalamocortical relay neurons are transferred to the cortex, where they cause excitatory post-synaptic potentials in cortical pyramidal neurons, thereby generating the EEG neocortical HVSs. The corticothalamic neurons may also participate in the generation of HVSs by having an excitatory influence on NRT and thalamocortical neurons (Steriade and Deschenes 1984, Buzsáki et al. 1988b, Steriade and Llinás 1988, McCormick 1990, McCormick and Bal 1997).

The basal forebrain and brainstem cholinergic projection cells innervate thalamic nuclei and cortical areas, and active functioning of these neurons inhibits HVSs and neocortical slow-waves (Buzsáki et al. 1988b, Steriade and Buzsáki 1990). Systemically administered muscarinic and nicotinic acetylcholine receptor agonists (Jäkälä et al. 1997) as well as intrathalamic infusion of nicotine into a relay nucleus, ventoposteromedial nucleus (VPM) (Riekkinen et al. 1995), suppress neocortical HVSs in adult rats. Furthermore, a partial excitotoxic lesion of NB in adult rats can increase neocortical HVSs and slow-waves, and this can be alleviated by administering an anticholinesterase drug, THA (Riekkinen et al. 1991e). In aged rats, the increase in HVSs has also been shown to correlate with the decrease in the number of ChAT-positive cells in the NB (Riekkinen et al. 1992).

The response of a particular thalamic neuron to acetylcholine depends on both the location and type of the neuron as well as the species of the animal being studied. In thalamic slices, the application of acetylcholine in cat or guinea pig lateral geniculate nuclei relay neurons results in a pronounced inhibition of rhythmic burst firing and the appearance of single spike activity (McCormick 1990, McCormick 1992b). This change in firing mode is achieved by depolarization of the neurons through activation of both nicotinic and muscarinic receptors. The nicotinic response typically results in a rapid depolarization associated with an increase in membrane cationic conductance, while the muscarinic response is slower, being related to a reduction in potassium conductance. In addition to these two depolarizing responses, a muscarinic hyperpolarizing response, due to an increase in potassium conductance, can occur between the fast nicotinic and slow muscarinic depolarizing responses (McCormick 1990, McCormick 1992b, McCormick and Bal 1997). Furthermore, in vitro application of acetylcholine in cat NRT results in marked inhibition of neuronal discharge mediated by muscarinic M₂ receptors via a post-synaptic increase in potassium conductance. This inhibition of action potential activity in NRT GABAergic neurons can prevent the hyperpolarization in thalamic relay neurons and may inhibit the rhythmic burst firing (McCormick 1990, McCormick 1992b).

Figure 2. A) Example of non-spindling background EEG. B) Example of a typical neocortical HVS episode (separated from background EEG by vertical bars) recorded above the frontal cortex during a period of relaxed behavioral waking-immobility (eyes open, head held up) in rat. The amplitude scale (vertical bar) indicates 300 mV and time scale (horizontal bar) indicates 1 s.

Figure 3. Schematic circuitry of the hypothesized mechanisms of HVS activity (left) and non-spindling EEG activity (right) (modified from Buzsáki et al. 1988b). Pathways and structures playing an important role in the HVS activity and non-spindling EEG activity are indicated by bold lines and shading. Arrows, excitatory pathways; bars, inhibitory pathways. When the NRT-thalamocortical system is released from the suppressive effect of the NB and brainstem nuclei, rhythmic activity of the NRT neurons phasically inhibits thalamocortical neurons and induces rhythmic rebound burst firing in these neurons, resulting in rhythmic depolarization of cortical neurons.

2.1.4 Cognitive and attentional effects of cholinergic manipulations

The findings from experimental animals, and ageing and AD research have provided an empirical foundation for the cholinergic hypothesis of learning and memory. The cholinergic hypothesis asserts that significant, functional disturbances in brain cholinergic activity play an important role in the memory loss and related cognitive problems associated with old age and AD. Furthermore, the prediction of the cholinergic hypothesis is that the enhancement or restoration of central cholinergic function should reduce the severity of the cognitive loss (Bartus et al. 1982, Francis et al. 1999).

Several pharmacological studies have investigated the effects of cholinergic receptor antagonists and cholinomimetic drugs on cognitive functions. The administration of the muscarinic antagonist, scopolamine, induces a performance deficit in several kinds of learning and memory tasks (Blokland 1995, Ebert and Kirch 1998). However, the impairments in cognitive functions observed after scopolamine treatment reflect more likely worsening in sensory and attentional processes than in learning and memory processes per se (Blokland 1995, Stanhope et al. 1995, Ebert and Kirch 1998). The cholinomimetic drugs are used to enhance the cholinergic transmission and thus improve the cognitive functions. Recent evidence from clinical studies shows that the treatment of AD patients with acetylcholinesterase inhibitors (e.g. THA, donezepil, rivastagmine, and galanthamine) can improve cognitive functioning and decrease neuropsychiatric problems (Francis et al. 1999, Perry et al. 1999). Similarly, in aged rats and MS-lesioned rats and mice, the acetylcholinesterase inhibitors stimulate spatial navigation in water maze (WM) task (Riekkinen et al. 1991a, Riekkinen et al. 1996, Ikonen et al. 1999). More recently, specific muscarinic M₁ agonists and M₂ antagonists, which might be useful in the treatment of cognitive and non-cognitive symptoms of AD, have been developed (Francis et al. 1999, Levy et al. 1999).

The early reports showing a correlation between dysfunction of the basal forebrain cholinergic system and the cognitive deficits in normal ageing and AD suggested that the basal forebrain projections to cortex and hippocampus were particularly important in learning and memory functions (Bartus et al. 1982, Francis et al. 1999). The findings that lesions in the basal forebrain with excitatory amino acid analogues, such as ibotenic acid, result in cognitive impairments and reduction of cholinergic markers in experimental animals, supported the role of the basal-cortical system in learning and memory (Dekker et al. 1991, Dunnett et al. 1991, Blokland 1995, Wenk 1997). However, the results from studies showing that the injection of other neurotoxins, such as quisqualic acid and AMPA (a-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) into the same site caused a greater reduction in cholinergic activity but less learning and memory deficits, did not support this hypothesis (Dekker et al. 1991, Dunnett et al. 1991, Blokland 1995, Wenk 1997). Only recently, has it become possible to destroy cholinergic neurons selectively in rats by using immunoglobulin (IgG)-saporin (Wiley et al. 1991). The antibody attaches to the low-affinity nerve growth factor (NGF) receptors, which are enriched only in the cholinergic neurons, and is internalized into the neurons allowing the cytotoxin saporin to destroy the neurons (Winkler et al. 1998). The direct infusion of IgG-saporin into the NB has only minor or no effects on spatial learning and memory, despite the fact that there is extensive loss of cortical ChAT-activity. These results suggest that deficits in learning and memory tasks after basal forebrain lesions are not well correlated with the degree of cholinergic loss (Torres et al. 1994, Wenk et al. 1994, Dornan et al. 1996, Wenk 1997, Winkler et al. 1998). Compared to the infusion of IgG-saporin into the NB, the intracerebroventricular (i.c.v.) IgG-saporin administration impairs cognitive functions in several tasks. However, it is possible that the damage to striatal cholinergic neurons and cerebellar Purkinje cells can explain some of the behavioral deficits produced by i.c.v. IgG-saporin (Leanza et al. 1995, Leanza et al. 1996, Winkler et al. 1998).

Recently, it has been proposed that the neuronal networks between distinct basal forebrain cholinergic nuclei and their target areas are active during different mnemonic and attentional processes. The septohippocampal projection is thought to be involved in the modulation of short-term spatial memory, and the projections from VDB to cingulate cortex may impact on the ability to utilize response rules through conditional discrimination (Everitt and Robbins 1997). Rather than being involved in learning and memory processes per se, the NB-neocortical cholinergic system contributes greatly to visual attentional function (Everitt and Robbins 1997). Non-selective lesions of NB in rat have been shown to result in performance deficits in 5-choice serial reaction time task, which measures attentional abilities (Robbins et al. 1989). Some of these deficits can be ameliorated by anticholinesterase treatment or by transplantation of cholinergic-rich embryonic cells into the cortex of lesioned animals (Muir et al. 1992). More recently, it has been shown that the selective lesioning of NB cholinergic neurons with immunotoxin IgG-saporin in rats also results in impairment of certain aspects of attentional processes (Wenk 1997). For example, McGaughly et al. (1999) found that infusion of IgG-saporin into the area of the NB/substania innominata in rats produced small but significant impairments in a 5-choice serial reaction time task under baseline conditions, and even more severe impairments under conditions of increased attentional demands. In addition to behavioral data, also the electrophysiological evidence implicates NB cholinergic neurons in attentional processes. An increase of activity of basal forebrain neurons is observed when the animal is waiting to respond to visual or auditory stimuli that are associated with the expectation of a food reward (Richardson and DeLong 1991). Whereas the basal forebrain cholinergic nuclei are involved in attentional, and learning and memory functions, the brainstem cholinergic nuclei play an important role in regulation of arousal and sleep-wake cycle (McCormick 1990, Steriade and Buzsáki 1990). It seems likely that the brainstem cholinergic nuclei are also involved in behavioral activation but at a more basic level than the modulation of cognitive functioning (Everitt and Robbins 1997).

2.1.5 Brain cholinergic system during ageing and in Alzheimer's disease

Ageing in rodents. Studies on the effects of age on the integrity of the basal forebrain cholinergic system in rats and mice have revealed inconsistent findings. Some rat studies have reported significant neuron loss in NB (Riekkinen et al. 1992, De Lacalle et al. 1996) while others have found no loss at all (Armstrong et al. 1993). Moreover, both increases and decreases during ageing in the size of neurons and in the overall size of the region of cholinergic cell bodies in basal forebrain in rats, have been reported (Sarter and Bruno 1998). In one study, the number of basal forebrain cholinergic neurons was found to be the same in young and aged mice, though these neurons became substantially smaller with increasing age (Mesulam et al. 1987). Also the data on cortical ChAT-activity in rodents during ageing is inconclusive: both decreases and no changes in cortical ChAT-activity in rats (Sarter and Bruno 1998), and increase (Sherman and Friedman 1990) and no change (Bernstein et al. 1985) in cortical ChAT-activity in mice, have been reported. The results from studies measuring cortical sodium-dependent high affinity choline uptake (HACU), which is the rate-limiting step in the synthesis of acetylcholine, also do not reveal any consistent changes in aged rodents. In one study, no changes in cortical HACU were found between 5- to 6- and 23- to 25- month-old rats (Meyer et al. 1984), whereas in another study, rats aged 10-11 months had lower cortical HACU than 3-4 months old rats but no further decline in HACU was found in 24- to 25-month-old rats (Sirviö et al., 1988). Moreover, Lebrun et al. (1990) did not find differences in HACU between young and aged mice, whereas in the study of Sherman et al. (1990), HACU was significantly decreased in aged mice. The differences can partly be explained by the variability in the age, strains and methods between the studies. Taken together, the above data does not consistently support the hypothesis that normal ageing in rodents extensively affects the basal forebrain cholinergic system.

Ageing in humans. Biochemical measurements of cortical cholinergic enzymes have revealed inconsistent age-related changes in non-demented individuals. Some studies have shown significant decreases in cortical ChAT- and AChE-activities (Davies 1979, Perry et al. 1981), while others have reported no loss at all (Wilcock et al. 1982). Also the data on basal forebrain cholinergic neuron loss during normal ageing is inconsistent: some investigations have reported up to 50 % decrease in the number of NB cholinergic neurons (De Lacalle et al. 1991) while others have found no age-related changes in neuron numbers (Chui et al. 1984). These findings suggest that there might be a slight age-related impairment in cholinergic system during normal ageing in humans. However, this decline is at best modest when compared to the cholinergic loss in AD.

Alzheimer's disease. The loss of cortical cholinergic markers and the degeneration of basal forebrain cholinergic neurons are the most consistent and severe biochemical deficits in AD. Both biochemical measurements of ChAT- and AChE-activities as well as ChAT immunohistochemistry and AChE histochemistry have demonstrated severe but regionally variable loss of cholinergic activity in AD brain (Bowen et al. 1976, Davies and Maloney 1976, Davies 1979, Reinikainen et al. 1990, Geula and Mesulam 1996). The loss of cholinergic innervation is greatest in cortical structures within the temporal lobe (Geula and Mesulam 1996). In AD, the number of cholinergic neurons is markedly decreased in all basal forebrain cholinergic cell groups. For example, in NB, neuronal loss ranging from 30 to 95 % has been reported (Whitehouse et al. 1982, Rinne et al. 1987, Geula and Mesulam 1999). The density of muscarinic receptors in AD brain has been reported to be slightly decreased or unchanged depending on the receptor subtype and brain region studied. In hippocampal formation, the loss of muscarinic M₂ receptors has been well documented (Rinne et al. 1989, Aubert et al. 1992, Rodriguez-Puertas et al. 1997). Some studies have also reported a decrease in the numbers of hippocampal M₁ and M₃ receptors, whereas the markers for M₄ receptors remained unchanged (Rodriguez-Puertas et al. 1997). In cortex, a decrease in the numbers of M₁ (Rodriguez-Puertas et al. 1997) and M₂ (Quirion et al. 1989, Aubert et al. 1992) receptors has also been documented. However, there are several reports showing no changes in cortical M₁ (Quirion et al. 1989, Rinne et al. 1989, Aubert et al. 1992) and M₂ (Rinne et al. 1989, Rodriguez-Puertas et al. 1997) or hippocampal M₁ (Quirion et al. 1989, Rinne et al. 1989, Aubert et al. 1992) receptor numbers in AD. In contrast, several studies suggest that the number of cortical nicotinic receptors is decreased in AD brain (Rinne et al. 1991, Svensson et al. 1997). In addition, the coupling of residual muscarinic receptors to G protein might be affected in AD, implying alterations in signalling pathways (Ferrari-DiLeo et al. 1995). Taken together, the cholinergic loss in AD is very extensive, involving the degeneration of cholinergic neurons, processes and receptors as well as changes in cholinergic signal transduction.

2.2 APOLIPOPROTEIN E

2.2.1 ApoE structure and function

Apolipoprotein E (ApoE) is a 35-kD glycoprotein containing 299 amino acids encoded by a four-exon gene on chromosome 19. In humans, the three isoforms of apoE, apoE2, E3 and E4 are encoded by respective apoE alleles (e2, e3 and e4). These three apoE alleles give rise to six different genotypes: e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4. The apoE isoforms differ in the presence of cysteine/arginine residues in the receptor-binding domain: apoE2 isoform has cysteine at residues 112 and 158, apoE3 has cysteine at site 112 and arginine at site 158, while apoE4 has arginine at both sites (Table 2) (Beffert et al. 1998).

Table 2. The polymorphism in apoE.

Isoform	Allele	Amino acids
ApoE2	ApoE e2	Cys-112; Cys-158
ApoE3	ApoE e3	Cys-112; Arg-158
ApoE4	ApoE e4	Arg-112; Arg-158

The major sites for apoE synthesis are the liver and the brain. In the brain, apoE is synthesized in both astrocytes and microglia, but not in neurons. The apoE protein has two major functional domains: the carboxy- and amino-terminal domains. The binding of apoE to lipoproteins occurs via the carboxy-terminal domain and the binding of apoE to the receptors occurs via the amino-terminal domain. After being released from the liver into plasma, the apoE is complexed to three classes of lipoproteins: very large lipoproteins, such as chylomicron remnants, very low-density lipoproteins (VLDL) and high-density lipoproteins (HDL). In the CNS, only HDL-like lipoprotein particles are found. The apoE-containing lipoproteins are bound and internalized by an apoE-dependent receptor mediated mechanism (Beffert et al. 1998). In the plasma, apoE regulates the lipid and cholesterol metabolism by mediating the transport of lipids to and from the bloodstream. In humans, some of the allelic variations of apoE are associated with lipid metabolism changes and increased susceptibility to atherosclerosis and coronary heart disease. For example, the apoE e4 allele increases the risk of coronary artery and cerebrovascular diseases and ischemic stroke (Wilson et al. 1996).

In the peripheral nervous system, apoE is involved in the mobilization and redistribution of cholesterol during development and after injury. Following sciatic nerve injury in rats, large amounts of lipids are released from degenerating myelin fibers. At the same time, apoE synthesis and release from macrophages are increased by 100- to 200-fold (Ignatius et al. 1986). At the injury site, the apoE is needed for the scavenging of cholesterol from cellular and myelin debris and for delivery of the lipids for storage in macrophages. Much of the stored cholesterol appears to be reused during axonal regeneration and remyelination (Boyles et al. 1989).

In the CNS, the apoE plays a pivotal role in the redistribution of cholesterol for sprouting neurons. Compared to the peripheral nervous system, the regenerative capabilities of CNS neurons are very limited. However, specific brain areas, such as hippocampal formation, have the ability to induce proliferation of presynaptic extensions from axons or terminals derived from undamaged neurons in order to compensate for the loss of specific input (Poirier 1994). For instance, a lesion to the rat entorhinal cortex removes the cortical connections to the hippocampus and causes a loss of the synaptic input to the granule cell layer. However, when over half of the synaptic input is destroyed, almost complete restoration of synapses can be observed within two months (Matthews et al. 1976). The lesion in the entorhinal cortex is followed by increased synthesis of apoE. It is thought that the apoE transport system and neuronal apoE receptors are involved in the release of lipids from the degenerating synaptic terminals and their storage in astrocytes, as well as in the transportation of lipid components to synaptic terminals for the synthesis of new synapses and terminals (Poirier et al. 1991, Poirier et al. 1993).

2.2.2 ApoE and Alzheimer's disease

Several research groups have recently studied the frequency distribution of different apoE alleles in AD patients. These studies show that the prevalence of the apoE e4 allele is increased in both sporadic and familial late-onset AD, whereas the frequency of apoE e2 allele is decreased in familial late-onset AD (Corder et al. 1994, Strittmatter and Roses 1996). In addition to being an important risk factor for AD, the apoE e4 allele is also associated with several other characteristics of the disease, like the age of onset, severity of pathology and cognitive impairment.

The AD patients with apoE e4 allele have an earlier average age of onset than patients not carrying an apoE e4 allele (Strittmatter and Roses 1996). In a recent study, it was found that the inheritance of two copies of apoE e4 alleles decreased the average age of onset by about seven years (Gomez-Isla et al. 1996). In contrast, the apoE e2 allele might function as a protective factor and decrease the risk for AD and delay the clinical appearance of its symptoms (Corder et al. 1994, Hyman et al. 1996). The relation between the apoE e4 allele and the presence of amyloid-containing senile plaques or intraneuronal tangles has also been examined in detail. It has been shown that the amyloid burden is increased in AD patients with the apoE e4 allele (Schmechel et al. 1993, Gomez-Isla et al. 1996). On the other hand, a lack of correlation between apoE genotype and senile plaques has been demonstrated (Heinonen et al. 1995). The data showing the effect of apoE genotype on accumulation of neurofibrillary tangles in AD brain is more controversial. A positive correlation between neurofibrillary tangles and the apoE e4 allele has been reported (Nagy et al. 1995, Ohm et al. 1995), whereas there are also reports showing no relationship between neurofibrillary tangles and the apoE e4 allele (Gomez-Isla et al. 1996). The apoE genotype is also associated with the severity of hippocampal damage and memory impairment. The AD patients homozygous for apoE e4 allele have the most prominent volume loss in the medial temporal lobe structures and more severe memory loss than those subjects without the apoE e4 allele (Lehtovirta et al. 1996b, Lehtovirta et al. 1996c).

2.2.3 ApoE, the cholinergic system and synaptic plasticity in Alzheimer's disease

The function of brain cholinergic system is largely dependent on the production of acetylcholine from membrane lipid components. Since the apoE is the major lipoprotein involved in lipid transport and metabolism in the brain, it has an important role in cholinergic transmission. It has been shown that AD patients with apoE e4 allele suffer more pronounced cholinergic deficits compared to AD patients without the apoE e4 allele. For example, the cortical and hippocampal ChAT-activity is reduced in AD patients carrying the apoE e4 allele when compared to patients without the apoE e4 allele (Poirier 1994, Soininen et al. 1995). Furthermore, the number of cholinergic neurons in NB in AD patients with apoE e4 is decreased (Arendt et al. 1997). Recently, it was also demonstrated that AD patients with the apoE e4 allele tended to have more pronounced cortical EEG slowing when compared to patients without an apoE e4 allele, which may reflect the more severe cholinergic deficit in these patients (Lehtovirta et al. 1996a, Lehtovirta et al. 2000). The data from clinical drug trial studies suggest that there are also differences in the efficacy of cholinomimetic treatments between AD patients with or without the apoE e4 allele. The anticholinesterase drug, THA, suppressed EEG slowing less effectively in AD patients with the apoE e4 allele (Riekkinen et al. 1997b). Poirier et al. (1995) also found that the treatment with THA resulted in a marked improvement of cognitive symptoms in more than 80 % of apoE e4 negative AD patients, whereas 60 % of apoE e4 carriers performed worse compared to the baseline situation. These results suggest that the limited efficacy of anticholinesterase treatment to stimulate the cholinergic system might be due to the more severe cholinergic hypofunction observed in AD patients with the apoE e4 allele.

In AD, the compensatory growth of fibers taking place in subcortical neurons undergoing degeneration is very limited when compared to regular dendritic growth seen during ageing (Arendt et al. 1995). Since apoE is involved in neuronal growth and synaptic remodelling after peripheral and CNS injuries, it has been suggested that apoE genotype might also have an effect on dendritic plasticity in AD patients. Indeed, Arendt et al. (1997) showed that in AD patients carrying one apoE e4 allele, the increase in dendritic length of subcortical neurons was much less pronounced when compared to apoE e3/e3 individuals, and that these plastic dendritic changes were reduced further in patients homozygous for the apoE e4 allele. Also the data from in vitro studies supports the idea that synaptic remodelling and plasticity might be more compromised in AD patients with apoE e4 allele than in e4 allele non-carriers. Holtzman et al. (1995) showed that in a CNS-derived neuronal cell line, apoE3, but not apoE4, increased neurite extension. Similarly, in cultured dorsal root ganglion neurons, apoE3 increased but apoE4 decreased neurite outgrowth (Nathan et al. 1994).

2.2.4 ApoE-mutant mice

The importance of apoE in lipid metabolism was first associated with etiology of atherosclerosis. The apoE-deficient mice were created to understand the role of apoE in lipid metabolism in vivo and in the genetic and environmental determinants of atherosclerosis. The common features for the apoE-deficient mice, irrespective of the source and genetic background, are an elevated plasma cholesterol level (5-8 times higher total cholesterol level compared to controls) and atherosclerotic lesions in the aorta, and coronary and pulmonary arteries (Plump et al. 1992, Zhang et al. 1992). Only recently, have apoE-deficient mice been used to study the role of apoE in nervous system and AD.

Following a sciatic nerve crush, the apoE-deficient mice revealed a similar peripheral nerve regeneration as the control mice (Popko et al. 1993, Goodrum et al. 1995). These results suggest that in the peripheral nervous system other lipoproteins may compensate for the loss of apoE. However, in the brain, the apoE is the main apolipoprotein and it is largely responsible for the neuronal plasticity and repair. Masliah et al. (1995b) found that there was an age-dependent 15 to 40 % loss of synaptophysin-immunoreactive nerve terminals and microtubule-associated protein 2-immunoreactive (MAP-2) dendrites in the neocortex and hippocampus of apoE-deficient mice when compared to controls. This suggests that apoE-deficient mice might display age-dependent disruption of the synaptic and dendritic organization of the neocortex and limbic system. Further testing showed that apoE-deficient mice also had abnormal neuronal and synaptic regeneration after perforant pathway transection and closed head injury (Masliah et al. 1995a, Chen et al. 1997). However, Anderson et al. (1998) failed to detect any differences in MAP-2 or synaptophysin immunoreactivity between the apoE-deficient and control mice, and showed similar reactive synaptogenesis in apoE-deficient and control mice after entorhinal cortex lesion. Electrophysiologically, apoE-deficient mice show either reduced (Krugers et al. 1997, Veinbergs et al. 1998, Krzywkowski et al. 1999) or normal (Anderson et al. 1998) long-term potentiation (LTP) in hippocampus. Several reports have also demonstrated cognitive impairments in WM (Gordon et al. 1995, Oitzl et al. 1997) and cholinergic deficits in cortex and hippocampus (Gordon et al. 1995, Chapman and Michaelson 1998) as well as impairments in other neurotransmitter systems, such as the noradrenergic (Chapman and Michaelson 1998, Puoliväli et al. 2000a) and serotonergic systems (Chapman and Michaelson 1998). However, other groups have not found cholinergic (Anderson and Higgins 1997), neuropathological (Anderson et al. 1998, Fagan et al. 1998) or behavioral (Anderson and Higgins 1997, Anderson et al. 1998) alterations in apoE-deficient mice.

More recently, transgenic mice expressing different human apoE isoforms have been developed. Raber et al. (1998 and 2000) showed that mice expressing human apoE4 isoform were impaired in WM when compared to apoE3 and control mice. Moreover, apoE4 transgenic mice showed a similar age-dependent degeneration of synaptophysin-positive presynaptic terminals and MAP-2 positive neuronal dendrites in neocortex and hippocampus as apoE-deficient mice, whereas apoE3 mice did not differ from controls (Buttini et al. 1999). In the same study, it was also reported that expression of apoE3, but not of apoE4, protected against kainic-acid induced neuronal damage (Buttini et al. 1999). These findings demonstrate that human apoE isoforms have different effects on brain functions in vivo and that transgenic mice expressing apoE4 develop cognitive impairments and neurodegenerative changes that resemble those observed in AD patients with the apoE e4 allele (Raber et al. 1998, Buttini et al. 1999).

3. AIMS OF THE STUDY

The present study was designed to investigate the cholinergic modulation of neocortical arousal and the interaction between the brain cholinergic system and apoE in the modulation of cortical EEG activity, cognitive processes and neuronal plasticity. The specific aims of this study were:

To study the action site of cholinergic drugs to modulate neocortical arousal by investigating the effects of thalamic (NRT and VPM) infusions of a muscarinic M₁ receptor agonist, McN-A-343, a muscarinic M₂ receptor antagonist, methoctramine, and a muscarinic receptor agonist, oxotremorine, on thalamocortically generated HVSs in rat. The hypothesis was that thalamic muscarinic M₁ and M₂ receptors are involved in the regulation of neocortical arousal and HVS activity (I and II).

To study the role of apoE in modulation of cortical EEG activity by investigating the EEG effects of ageing, NB-lesion and scopolamine treatment in apoE-deficient mice. The hypothesis was that the lack of apoE would affect cortical EEG activity during ageing and after scopolamine treatment, and aggravate EEG defects induced by NB-lesion (III).

To study the role of apoE in modulation of cognitive processes and neuronal plasticity by investigating the effects of ageing and NB-lesion on spatial navigation and cholinergic activity in apoE-deficient mice. The hypothesis was that the lack of apoE would disrupt spatial navigation and the integrity of cholinergic system during ageing, and aggravate behavioral and biochemical defects induced by NB-lesion (IV).

4. MATERIALS AND METHODS

4.1 ANIMALS

Male Han:Wistar rats (n=83), and male apolipoprotein E-deficient and control mice (n=180) were used in this study. The rats were 16-18 months old (I and II) and the mice 3-21 months old when tested (III and IV). The animals were singly housed in a controlled environment (National Animal Center, Kuopio, Finland; temperature 22 °C, humidity 50-60 %, lights on from 0700 to 1900 hours) with water and food freely available. The apoE-deficient mice were initially obtained from a colony produced by Dr. N. Maeda and colleagues (University of North Carolina, Chapel Hill, USA) (Piedrahita et al. 1992, Zhang et al. 1992). The apoE gene was inactivated in mouse embryonic stem (ES) cells of 129/Ola origin by homologous recombination. The targeted ES cells were injected into C57BL/6J blastocysts. Chimeras were mated to C57BL/6J to obtain (C57BL/6J x 129/Ola) F1 animals heterozygous for the disrupted apoE gene. These hybrid mice were backcrossed to C57BL/6J mice for nine generations. Homozygous mice were obtained by intercrossing. Progeny from this homozygous line was transferred to Transgenic Alliance, L'Arbresle Cedex, France. The colony was maintained at Transgenic Alliance from which the experimental groups were transferred to Finland. Age-matched C57Bl/6J mice (National Animal Center, Kuopio, Finland) were used as controls. A summary of the experimental design is presented in table 3.

Table 3. The summary of the experimental design: species, animal number, operation and measurements conducted in each publication (I-IV). Abbreviations: ChAT = choline acetyltransferase; n = animal number; NB-lesion = NB-lesion or sham-lesion; WM = water maze; X = measurement/operation conducted; - = measurement/operation not conducted.

Publication	Species	n	NB-lesion	EEG	WM	ChAT-activity	Histology / Immunohistochemistry
I	rat	41	-	X	-	-	X
II	rat	42	-	X	-	-	X
III	mouse	34	X	X	-	X	-
IV IV IV	mouse mouse mouse	97 17 32	X - X	- - -	X X -	X X -	X - X

4.2 IMPLANTATION OF EEG ELECTRODES AND INFUSION CANNULAS

The rats and mice used for EEG recordings were anaesthetized with Equithesin (3 or 5 ml/kg, i.p., respectively) and placed in a stereotaxic frame. Two active EEG recording electrodes (stainless steel screws 0.5 mm in diameter) were implanted symmetrically on both sides above the frontal cortex for both rats and mice. For the rats, the guide cannulas were bilaterally implanted in the NRT, VPM, and ventral hippocampus or unilaterally in the lateral ventricle during the same operation when the EEG electrodes were being implanted. The implantation of EEG electrodes and infusion cannulas is described in detail in publications I-III.

4.3 NUCLEUS BASALIS LESION

The mice used for NB-lesion studies were anaesthetized with Equithesin (5 ml/kg, i.p.) and placed in a stereotaxic frame. The NB-lesion was performed with intracerebral quisqualic acid (Sigma, USA) infusions when the mice were 4 months old. In publication III, the quisqualic acid was infused unilaterally into NB (7.6 mg/0.2 ml/lesion site; two lesion sites per hemisphere; volume of 0.2 ml was delivered in 3 min) and the other hemisphere was sham-lesioned. In publication IV, the mice were either NB-lesioned with bilateral quisqualic acid infusion (3.8 m g/0.2 ml/lesion site; two lesion sites per hemisphere; volume of 0.2 ml was delivered in 3 min) or bilaterally sham-lesioned. The sham-lesion was performed identically to the NB-lesion, but no infusions were made. In publication III, the unilateral NB-lesion was performed during the same operation as the EEG electrodes were implanted. The NB-lesion procedure is described in detail in publications III and IV.

4.4 EEG RECORDINGS

Publications I and II. The IBM-compatible software separated HVSs from background EEG. A cumulative incidence of HVSs or HVS total duration (incidence * mean duration) during a 20-min cumulative waking-immobility (eyes open, head held up) period was used as a marker for HVS activity. The EEG epochs were considered as spike-and-wave HVS periods if they fulfilled the following criteria: 1. the amplitude of the EEG was more than twice that of the background EEG, 2. the duration of each epoch was > 0.5 s, 3. the frequency of HVS epoch was 6-12 Hz, 4. the time between two separate spindles was > 0.5 s (if the time between two spindles was < 0.5 s, it was considered as one HVS), 5. no movement registered by the magnetic coil binding on the head except vibrissal or head tremor was allowed 1 s before or during each HVS epoch. The total recording time needed to achieve a 20 min period of behavioral waking-immobility related EEG was recorded to monitor the possible motor effects of drug treatments.

Publication III. Five 4 s long artifact-free EEG episodes were recorded from both NB-lesioned and sham-lesioned hemispheres during each recording session in awake and not moving mice. EEG samples were converted to a digital form using a 250 Hz sample rate and Fast Fourier Transformation. The EEG spectrum was divided into the following frequency bands: delta (1-4 Hz), theta (4-8 Hz), alpha (8-12 Hz) and beta (12-20 Hz). The relative amplitudes of these EEG bands were used. For example, the relative delta amplitude = absolute value of delta amplitude/sum of absolute values of all the frequency bands (= delta, theta, alpha and beta). To change the data into a Gaussian distribution, a natural logarithm transformation was applied for relative amplitude values prior to the statistical analyses.

The EEG recording systems are described in detail in publications I-III.

4.5 DRUG STUDIES

Publication I. McN-A-343, a muscarinic M₁ agonist (Micheletti and Schiavone 1990, Lambrecht et al. 1993) (RBI, USA), was administered into the NRT (bilaterally; 2.0 and 20.0 mg/0.25 ml per hemisphere), i.c.v. (unilaterally; 10 and 100 mg/5.0 ml) and for control purposes intrahippocampally (bilaterally; 4.0 mg/0.5 ml per hemisphere). Methoctramine tetrahydrochloride, a muscarinic M₂ antagonist (Giraldo et al. 1988, Michel and Whitning 1988) (RBI, USA), was administered into the NRT (bilaterally; 5.0 and 20.0 mg/0.25 ml per hemisphere) and for control purposes intrahippocampally (bilaterally; 40.0 mg/0.5 ml per hemisphere). Intrathalamic, intrahippocampal or i.c.v. administrations of vehicle were used as the control treatments. The infusion speed in intrathalamic and intrahippocampal administrations was 0.2 ml/min and in i.c.v. administrations 1.0 ml/min. The effects of McN-A-343 and methoctramine on neocortical HVSs and total recording time were studied.

Publication II. Oxotremorine, a muscarinic agonist (Tayebati et al. 1999) (RBI, USA), was administered into the NRT or VPM (bilaterally; 3.0 and 15.0 mg/0.5 ml per hemisphere), i.c.v. (unilaterally; 10, 30 and 100 mg/5.0 ml) or for control purposes intrahippocampally (bilaterally; 3.0 mg/0.5 ml per hemisphere). Methoctramine tetrahydrochloride, a muscarinic M₂ antagonist (Giraldo et al. 1988, Michel and Whitning 1988) (RBI, USA), was administered i.c.v. (unilaterally; 10, 30 and 100 mg/5.0 ml). THA, an anticholinesterase drug (Freedman and Dawson, 1991, Soares and Gershon 1995) (Sigma, USA), was administered systemically (3 mg/kg). Intrathalamic, intrahippocampal, i.p. or i.c.v. administrations of vehicle were used as the control treatments. The infusion speed in intrathalamic and intrahippocampal administrations was 0.2 ml/min and in i.c.v. administrations 1.0 ml/min The effects of oxotremorine, methoctramine and THA, either alone or in combination, on HVSs and total recording time were studied.

Publication III. Scopolamine hydrocloride, a muscarinic antagonist (0.05 and 0.2 mg/kg; Sigma, USA), was administered systemically. I.p. injections of vehicle were used as control treatments. In NB-lesion study, the effects of NB-lesion and scopolamine, and in the ageing study, the effects of ageing and scopolamine on cortical EEG activity were examined in control and apoE-deficient mice. The NB-lesioned mice were tested 40 days after the lesion at the age of 5 months, and the mice included in the ageing study were tested at the age of 3, 12 and 21 months.

The methods for drug testing are described in detail in publications I-III.

4.6 BEHAVIOR

The Morris WM was used to study the spatial learning in control and apoE-deficient mice. In standard use of the WM task, the animal is placed into the pool at one of the four randomly ordered start locations (e.g. north, south, west and east ‘poles’) near the wall and allowed to swim to a submerged platform (e.g. centre of the north-west quadrant) maintained in a fixed position throughout the training. In this study, the platform location was kept constant for five days (initial acquisition) after which the platform was placed in the opposite quadrant for one more day (reversal learning). The percentage of found platforms and the latency to find the platform were measured during training days. The probe trial accuracy was tested after initial acquisition and reversal learning training. During the probe trial, the animal is permitted to swim freely in the pool without any platform present. The use of a place strategy is inferred if the animal crosses the old platform position more often than the equivalent positions in the other three quadrants. A computer connected to an image analyser (HVS Image Ltd., Hampton, UK) monitored the swim pattern. The timing of the latency to find the submerged platform was started and ended by the experimenter. Variables used in the data analysis were: the percentage of found platforms, the latency to find the platform and the number of counter crossings during the spatial probe test. The NB-lesioned mice were lesioned at the age of four months and tested 25 (short-term recovery) or 106 (long-term recovery) days after the lesion at the age of 5 or 8 months, respectively, and the mice included in the ageing study were tested at the age of 16 months. The training paradigm and testing system are described in detail in publication IV.

4.7 BIOCHEMISTRY

The mice in publication III were decapitated at the age of 5 or 21 months, and in publication IV at the age of 5, 8 or 16 months. The brain was quickly removed; the hippocampus and/or frontal cortex were dissected on ice and frozen in liquid nitrogen. Tissue samples were stored at -75 °C until the hippocampal and cortical ChAT-activity were measured according to the method of Fonnum (1975).

5. RESULTS

5.1 EEG RECORDINGS IN RATS

5.1.1 High-voltage spindle activity

The NRT infusion of methoctramine 20 µg, McN-A-343 2 and 20 µg, and oxotremorine 3 and 15 µg as well as VPM infusion of oxotremorine 15 µg decreased the HVS activity (I, Fig. 2A,B; II, Fig. 4A,B). Intrahippocampal infusions of methoctramine 40 µg, McN-A-343 4 µg, and oxotremorine 3 µg had no effect on HVS activity (I, Fig. 2C; II, Fig. 4C). The i.c.v. administered McN-A-343 100 µg, and oxotremorine 30 and 100 µg decreased the HVS activity (I; II, Fig. 2A). However, the i.c.v. administered methoctramine 100 µg increased the HVS activity (II, Fig. 2B). The i.c.v. administered methoctramine 100 µg blocked the decrease in HVS activity induced by THA 3 mg/kg treatment or i.c.v. administered oxotremorine 100 µg (II, Fig. 3A,B). A summary of the drug effects on HVS activity is presented in table 4.

5.1.2 Total recording time

The NRT infusion of methoctramine 20 µg decreased the total recording time but the NRT infusion on McN-A-343 (2 and 20 µg), and NRT (3 and 15 µg) or VPM (3 and 15 µg) infusions of oxotremorine had no effect on total recording time (I; II, Table 1). Intrahippocampal infusions of methoctramine 40 µg, McN-A-343 4 µg or oxotremorine 3 µg had no effect on total recording time (I; II, Table 1). The i.c.v. administered McN-A-343 100 µg and oxotremorine 30 µg decreased total recording time, whereas the i.c.v. administered methoctramine (10, 30 and 100 µg) had no effect on total recording time (I; II, Table 1). The i.c.v. administered methoctramine 100 µg blocked the decrease in total recording time induced by i.c.v. administered oxotremorine 100 µg but failed to block the decrease in total recording time induced by THA 3 mg/kg treatment (II, Table 1).

Table 4. The effects of methoctramine (METHO), McN-A-343 (McN), oxotremorine (OXO), and tetrahydroaminoacrine (THA) on high-voltage spindle (HVS) activity during waking-immobility in rats. Abbreviations: ­ = increase in HVS activity; ¯ = decrease in HVS activity; 0 = no change in HVS activity; - = not measured.

Drug	NRT	VPM	Hippocampus	i.c.v.	i.p. THA 3 mg/kg + i.c.v. METHO	i.c.v. OXO 100 mg + i.c.v. METHO
METHO
0 m g	0	-	0	0	¯	¯
5 m g	0	-	-	-	-	-
10 m g	-	-	-	0	¯	¯
20 m g	¯	-	-	-	-	-
30 m g	-	-	-	0	¯	¯
40 m g	-	-	0	-	-	-
100 m g	-	-	-	­	0	0
McN
2 m g	¯	-	-	-	-	-
4 m g	-	-	0	-	-	-
10 m g	-	-	-	0	-	-
20 m g	¯	-	-	-	-	-
100 m g	-	-	-	¯	-	-
OXO
3 m g	¯	0	0	-	-	-
10 m g	-	-	-	¯	-	-
15 m g	¯	¯	-	-	-	-
30 m g	-	-	-	¯	-	-
100 m g	-	-	-	¯	-	-

5.2 EEG RECORDINGS IN APOE-DEFICIENT MICE

5.2.1 Nucleus basalis lesion study

NB-lesion decreased the relative amount of delta activity and increased the relative amount of alpha activity similarly in control and apoE-deficient mice (III, Table 3). Scopolamine 0.2 mg/kg decreased the relative amount of theta activity similarly in sham-lesioned and NB-lesioned hemispheres in both strains (III, Fig. 1B). A summary of the effects of NB-lesion and scopolamine treatment on cortical EEG activity is presented in table 5.

Table 5. Cortical baseline EEG activity and the effects of scopolamine (SCOP) on cortical EEG activity 40 days after unilateral quisqualic acid induced NB-lesion in 5 month-old awake and not moving control and apoE-deficient mice. Abbreviations: ­ = increase in the relative amplitude; ¯ = decrease in the relative amplitude; 0 = no change in the relative amplitude; D = drug effect; L = lesion effect.

Group	Delta 1-4 Hz	Theta 4-8 Hz	Alpha 8-12 Hz	Beta 12-20 Hz
Baseline
control mice	L¯	0	L­	0
apoE-deficient mice	L¯	0	L­	0
SCOP 0.05 mg/kg
control mice	0	0	0	0
apoE-deficient mice	0	0	0	0
SCOP 0.2 mg/kg
control mice	0	D¯	0	0
apoE-deficient mice	0	D¯	0	0

5.2.2 Ageing study

Ageing decreased the relative amount of alpha activity similarly in both strains. However, at 3 months of age, the relative amount of alpha activity and at 21 months of age, the relative amounts of alpha and beta activities were higher in apoE-deficient than in control mice (III, Table 3). Scopolamine 0.05 and 0.2 mg/kg increased the relative amount of delta activity and decreased the relative amount of theta activity more effectively in control than in apoE-deficient mice (III, Fig. 2A,B and 3A,B). Scopolamine 0.05 mg/kg more effectively decreased the relative proportion of theta activity in aged than in young mice (III, Fig. 2B). On the other hand, scopolamine 0.2 mg/kg increased the relative proportion of delta activity more effectively in aged than in young control mice, while it had almost no effect in apoE-deficient mice at any age (III, Fig. 3A). Scopolamine 0.2 mg/kg increased the relative proportion of alpha activity more effectively in apoE-deficient mice (III, Fig. 3C). A summary of the effects of ageing and scopolamine treatment on cortical EEG activity is presented in table 6.

Table 6. Cortical baseline EEG activity and the effects of scopolamine (SCOP) on cortical EEG activity in 3- to 21-month-old awake and not moving control and apoE-deficient mice. Abbreviations: ­-­­ ­ = increase in the relative amplitude; ¯ -¯¯¯ = decrease in the relative amplitude; 0 = no change in the relative amplitude; A = ageing effect; D = drug effect; GM = genetic manipulation effect.

Group	Delta1-4 Hz	Theta 4-8 Hz	Alpha 8-12 Hz	Beta 12-20 Hz
Baseline
control mice	0	0	A¯, GM1	GM2
apoE-deficient mice	0	0	A¯, GM1	GM2
SCOP 0.05 mg/kg
control mice	D­ ­	D¯ ¯	0	0
apoE-deficient mice	D­	D¯	0	0
SCOP 0.2 mg/kg
control mice	D­ ­ ­	D¯ ¯ ¯	D­	0
apoE-deficient mice	D­	D¯	D­ ­	0

GM^1,2 = at 3 and 21 months of age, the relative alpha¹ activity, and at 21 months of age, the relative beta² activity were higher in apoE-deficient mice

5.3 BEHAVIOR

There were no differences in escape latency or in the number of found platforms between the control and apoE-deficient mice during the initial acquisition stage or in reversal learning at the age of 5, 8 or 16 months (IV, Fig. 1-3). Furthermore, the probe trial accuracy, measuring the number of crossings over the old platform position, was similar in both strains at the age of 5 or 8 months (IV). The NB-lesion had no effect on the escape latency or on the number of found platforms during the initial acquisition after the short- (25 days) or long-term (106 days) recovery in control and apoE-deficient mice. The NB-lesion increased the escape latency and decreased the number of found platforms during reversal learning similarly in both strains after short-term recovery. However, NB-lesion had no effect on the escape latency or on the number of found platforms during reversal learning after long-term recovery (IV, Fig. 1 and 2). NB-lesion decreased the probe trial accuracy similarly in both strains after the short-term recovery but had no effect on probe trial accuracy after long-term recovery (IV). A summary of the effects of NB-lesion on WM behavior is presented in table 7.

5.4 BIOCHEMISTRY

The control and apoE-deficient mice had similar levels of cortical ChAT-activity at the age of 5, 8, 16 and 21 months and hippocampal ChAT-activity at the age of 21 months (III, Table 4; IV, Table 1). The NB-lesion decreased cortical ChAT-activity equally in both strains after recovery of 34, 46 or 115 days (III, Table 4; IV, Table 1). A summary of the effects of NB-lesion on cortical ChAT-activity is presented in table 7.

5.5 HISTOLOGY AND IMMUNOHISTOCHEMISTRY

In publications I and II, the histological analysis revealed that the tips of the infusion cannulas were correctly located in all of the rats with intrathalamic, intrahippocampal and i.c.v. implantations (I, Fig. 3).

In publication IV, the mice included in behavioral studies were decapitated 34 or 115 days after the NB-lesion at age of 5 or 8 months, respectively. The histological analysis revealed that lesions were correctly located in the NB area in all mice. The mice that were used for more accurate histology were killed 34 days after the NB-lesion at the age of 5 months. The gliosis around the needle tract confirmed the correct placement of the NB-lesion (IV, Fig. 4B). The decrease of acetylcholine-containing fibers in cortex of NB-lesioned mice was confirmed by AChE staining (IV, Fig. 4C,D). There were no differences in the number of ChAT-positive cells between the control and apoE-deficient mice and the NB-lesion decreased the number of ChAT-positive cells similarly in both strains (IV, Fig. 4E,F). A summary of the effects of NB-lesion on the number of ChAT-positive cells in NB is presented in table 7.

Table 7. The effects of unilateral or bilateral quisqualic acid induced NB-lesion on cortical ChAT-activity, number of ChAT-positive cells in NB, and WM initial acquisition or reversal learning in control and apoE-deficient mice after different recovery periods. Abbreviations: ¯ = decrease/impairment in ChAT-activity (%), number of ChAT-positive cells (%) or WM spatial navigation in NB-lesioned hemisphere/mice compared to sham-lesioned hemisphere/mice; 0 = no changes in WM behavior; * = unilateral NB-lesion. NB-lesion decreased ChAT-activity, ChAT-positive cells and WM reversal learning similarly in both strains. No strain differences or strain by lesion interaction effects were found in ChAT-activity, number of ChAT-positive cells or WM behavior.

	ChAT-activity			ChAT-positive	WM behavior
	in cortex (%)			cells in NB (%)	Initial		Reversal
Recovery (days)	34	46*	115	34	25	106	25	106
Control mice	21 %¯	27 %¯	14 %¯	17 %¯	0	0	¯	0
ApoE-deficient mice	28 %¯	38 %¯	15 %¯	20 %¯	0	0	¯	0

6. DISCUSSION

6.1 METHODOLOGICAL CONSIDERATIONS

6.1.1 Nature of transgenic and control mice

The transgenic and gene-targeted knockout mice have provided valuable models of human genetic disorders and increased our knowledge about the function of different genes in the CNS. However, there are a number of limitations and pitfalls that have to be taken into consideration when interpreting the results obtained from mutant mice. Perhaps the most important factors that may contribute to the observed phenotype are the nature of the transgenic and control lines. The differences in the transgene construct, embryonic stem cell line, host blastocyst strain, background strain and number of backcrossings may affect the phenotype as suggested by Anderson et al. (1998). For example, often the targeted mutation is made in embryonic stem cells derived from the mouse strain '129' which shows neurological anomalies and poor learning in many behavioral tasks (Gerlai 1996, Crawley et al. 1997). It is possible that the targeted locus remains flanked by the '129' mouse strain alleles, and that this piece of chromosome is transmitted to many subsequent generations of mice because of the genetic linkage. Thus, it can be argued that the particular phenotype may not be attributable to the transgene in the targeted locus but to flanking '129' mouse strain alleles (Gerlai 1996, Lipp and Wolfer 1998). In the present study, wild-type littermates were not available. Therefore, the control mice were obtained from a different colony than the apoE-deficient mice, and only matched with age and background strain. This methodological disadvantage may partly explain the lack of major behavioral, neuropathological and neurophysiological alterations observed in the apoE-deficient mice in this study. However, the results from the present study are in many respects very similar compared to the earlier studies conducted with the very same line of apoE-deficient mice (Piedrahita et al. 1992) which did have appropriate wild-type littermates (Anderson and Higgins 1997, Anderson et al. 1998). These results suggest that the behavioral phenotype of apoE-deficient mice of this strain is attributable to the nature of the transgenic line and the use of wild-type littermates vs. age-matched C57Bl/6J mice as controls may not affect the observed phenotype.

6.1.2 Quisqualic acid induced nucleus basalis lesion

The cholinergic neurons in NB are one of the most prominent systems implicated in AD-related cognitive deficits and EEG slowing. The rodent models of cholinergic hypofunction in NB-neocortex system have been widely used to examine the role of the NB in the regulation of cognitive processes and cortical electrical activity. The lesioning of NB electrolytically and with different excitotoxins (e.g. AMPA, ibotenic acid and quisqualic acid) (Winkler et al. 1998) is known to produce behavioral deficits and slowing of neocortical EEG in experimental animals (Riekkinen et al. 1990, Dekker et al. 1991, Wenk 1997). In this study, the NB of control and apoE-deficient mice was lesioned with quisqualic acid, a potent agonist for glutamate sites, which causes the influx of Ca2+ resulting in cell dysfunction and death. However, quisqualic acid induced NB-lesion is not selective for cholinergic neurons. More recently, the NB-lesion studies have been repeated using IgG-saporin, an immunotoxin that selectively destroys cholinergic neurons and induces marked depletion of cortical ChAT-activity. It has been shown that the selective loss of cholinergic neurons after an IgG-saporin induced NB-lesion is not associated with memory deficits and EEG slowing in rats (Wenk et al. 1994). These results have indicated that the memory impairments and EEG deficits are not well correlated with the cholinergic loss in basal forebrain and cortex. Instead, the behavioral and EEG deficits observed after NB-lesion are more likely due to destruction of non-cholinergic NB neurons, e.g. GABAergic neurons, and nearby structures involved in cognitive processes and control of cortical EEG, such as striatal and pallidal neurons (cortico-striato-pallidal loops), amygdala, and the NRT (Dunnett et al. 1991, Muir et al. 1993).

6.1.3 EEG recordings in rats and mice

In this study, a model of age-related EEG phenomenon observed in rats was used. In rats, the number of thalamocortically generated HVSs, which occur during states of low arousal and vigilance, has been shown to correlate with cholinergic deficits in aged and NB-lesioned rats (Buzsáki et al. 1988b, Riekkinen et al. 1991e, Riekkinen et al. 1992). Importantly, also in AD, the EEG slowing has been shown to correlate with loss of cholinergic markers (Soininen et al. 1992). The HVS model has been validated in our laboratory during recent years. The research has mainly focused on the cholinergic system. In addition, several reports relating to other neurotransmitter systems, such as serotonergic (Jäkälä et al. 1995, Jäkälä et al. 1996b) and noradrenergic systems (Riekkinen et al. 1991c, Jäkälä et al. 1992), which are also impaired in AD (Mann 1983, Palmer et al. 1987), have been published. Previously, it has been shown that systemically administered cholinergic drugs modulate HVS activity (Danober et al. 1993, Jäkälä et al. 1996a, Jäkälä et al. 1996b, Jäkälä et al. 1997). However, the site of action of systemically administered drugs to modulate cortical arousal is difficult to pinpoint. Therefore, in the present study, the effects of intrathalamic (NRT and VPM) infusions of cholinergic drugs on neocortical HVSs were investigated. It is possible that the NRT and VPM administered cholinergic drugs may diffuse to the adjacent thalamic nuclei, and that the activation of these nuclei may also contribute to the modulation of the HVS activity. In this model the effects of drug treatments on locomotor activity could also be measured. It was found that the intrathalamically, i.c.v and i.p administered cholinergic drugs not only affected HVS activity but also increased or decreased motor/behavioral activity as indicated by the respective changes in the recording times, suggesting that central cholinergic muscarinic receptors also regulate locomotor activity. However, it should be noted that the effects of the drugs on HVSs truly reflect a behavioral waking-immobility state in the animals, because all the movement related EEG epochs from the HVS recordings were excluded. Therefore, the effects of the drugs on neocortical HVSs during quiet waking-immobility behavior and on behavioral/motor activity can be differentiated.

Previously, in our laboratory, the mouse EEG recording system has been used to measure the effects of alpha2 agonist and antagonist on cortical arousal in alpha2C adrenoceptor overexpressing mice (Björklund et al. 1998). In this study, the effects of NB-lesion, age and drug treatments on cortical quantitative EEG activity were measured in control and apoE-deficient mice. In the NB-lesion EEG study, the number of mice was small, four apoE-deficient and five control mice were tested. This may lower the power of the results, and thus mask or cause misinterpretation of some of the findings. In the ageing study, the same mice were tested at 3, 12 and 21 months of age. Those animals, which had lost their EEG electrodes during the 18 months of ageing, were removed from the analyses. The proper attachment of EEG electrodes to skull of the mice was confirmed by measuring the baseline EEG activity on two consecutive days before the recordings at different ages were started. The mice that had poor connection of EEG electrodes and apparent EEG abnormalities were not included in the analyses. The relative amplitude of different frequency bands was used to reduce the variation between the mice due to the small group size (NB-lesion study) and to permit the comparisons of EEG recordings at different ages (ageing study). It is possible that some of the EEG activity changes observed in relative amplitude of certain frequency bands might hide the changes occurring in other frequency bands. However, in the present study, the EEG activity changes were not correlated with behavioral or physiological functions of mice but used to differentiate various experimental conditions (strain, age, NB-lesion and drug treatments). This enables the use of relative amplitude as a variable for changes in cortical EEG arousal after different manipulations.

6.1.4 Behavioral testing in water maze

The WM was originally developed for studying spatial learning in the rat (Morris 1984). Today, the WM is also widely used to study learning and memory functions in normal and mutant mice (Björklund et al., 1998; Ikonen et al., 1999; Rissanen et al., 1999; Wilson et al., 1999). The main advantage of the WM task when compared to some other tasks, e.g. radial-arm-maze, is the speed of training. No pretraining is required and the acquisition is rapid. This permits the use of a large number of animals, and different pharmacological and/or lesion treatments can be assessed in short periods of time (Hodges 1996). In the WM task, it is also possible to dissociate different aspects of learning and memory functions (e.g. reference vs. working memory) (Frick et al. 1995), and to assess motor (swimming speed) and visuospatial functions (Hodges 1996, Dudchenko et al. 1997). However, the WM task has been criticised for several reasons. First of all, the WM is an aversively motivated task, in which the immersion of the animal into water may cause endocrinological or other stress effects, which may interact with lesion or pharmacological manipulations in uncontrolled ways (Morris et al. 1982, Hodges 1996). It is also possible that the observed impairment in WM, especially in mice which are not as good swimmers as rats, is attributable to the use of noncognitive behavioral strategies such as passive floating and thigmotaxis (a term used to describe animal's inclination to maintain contact with the walls of the pool) (Lipp and Wolfer 1998). Moreover, the WM procedures vary between laboratories, which makes the comparison of results difficult. For example, the pool and platform sizes, the water temperature and training schedule (e.g. duration of training trial and inter-trial-interval) may affect the acquisition and performance in the WM task (Puoliväli et al. 2000b).

In this study, both reference and working memory components of the spatial navigation were tested. The mice were first trained to locate the platform, which was held in a constant position for five days (reference memory), after which the platform position was reversed for one day (working memory). The retention of spatial memory was tested with a probe trial after both initial acquisition and reversal learning. During the probe trial, the platform was removed and the number of crossings over the previous platform position was measured. The reversal of the platform to a new position requires the animal to adapt new rules to perform successfully in this phase of the task. The results from this study showing rapid learning curves during the initial acquisition phase and impaired performance of NB-lesioned mice during reversal learning are in good agreement with earlier studies using a similar kind of WM procedure in NB-lesioned mice (Sweeney et al. 1988, Sweeney et al. 1989).

6.2 CHOLINERGIC MODULATION OF CORTICAL AROUSAL IN RATS

Previous electrophysiological studies have shown that systemic administration of both muscarinic agonists and anticholinesterases can decrease HVSs (Danober et al. 1993, Riekkinen et al. 1993, Jäkälä et al. 1997, Riekkinen et al. 1997a). The results from this study show that i.c.v. administered oxotremorine (a muscarinic agonist) and McN-A-343 (a muscarinic M₁ selective agonist), also decreased HVSs, and that i.c.v. administered methoctramine (a muscarinic M₂ selective antagonist) increased HVSs and dose-dependently blocked the effect of both i.c.v. oxotremorine and i.p. THA (an anticholinesterase drug) to decrease HVSs. The previous and present results from systemic and i.c.v. administration of muscarinic acetylcholine active drugs suggest that central muscarinic acetylcholine receptors play an important role in the regulation of thalamocortical oscillations and their related neocortical HVSs. However, when the drugs are administered systemically or i.c.v., it is difficult to interpret the exact site of location.

To pinpoint the possible action site of systemically and i.c.v. administered drugs on HVS activity, the effects of intrathalamically administered cholinergic muscarinic drugs on thalamocortical oscillations were studied. The results show that intrathalamically (NRT and/or VPM) administered oxotremorine, McN-A-343 and methoctramine decreased HVS activity, demonstrating that thalamic muscarinic receptors are involved in the regulation of thalamocortical oscillations and neocortical arousal. The results from this study are supported by the anatomical findings showing that muscarinic M₂ receptors are abundant and that M₁ receptors do also exist in the thalamus (Wang et al. 1989, Wall et al. 1991, Wei et al. 1994). Also the earlier in vitro electrophysiological studies are in line with the results from this study. Namely, in thalamic slice preparations, the application of the muscarinic agonist acetyl-beta-methylcholine to guinea pig lateral geniculate relay nuclei neurons resulted in a slight hyperpolarization followed by a muscarinic slow depolarization, which inhibited the generation of neuronal burst discharges and promoted the generation of single spike activity in thalamic relay neurons (McCormick 1992a). Furthermore, application of acetylcholine to GABAergic interneurons in the lateral geniculate nucleus or NRT resulted in their inhibition and desynchronization of the EEG via an increase in membrane potassium conductance, this effect being mediated by the M₂ subclass of acetylcholine receptors (McCormick 1990, McCormick 1992b).

The synaptic site of action of muscarinic cholinergic drugs in modulating thalamocortically generated HVSs in thalamus may involve both pre- and post-synaptic muscarinic receptors. Muscarinic M₁ receptors are predominantly post-synaptic, whereas M₂ receptors are both post- and presynaptic in nature (Levey 1996). The presynaptic M₂ receptors are autoreceptors, and their activation results in an inhibition of acetylcholine release, whereas their blockade with antagonists increases the release of acetylcholine. The results from this and earlier in vitro studies suggest that administration of oxotremorine and McN-A-343 into the NRT decreases HVSs by activating post-synaptic muscarinic receptors. This in turn inhibits the generation of action potentials in GABAergic interneurons and prevents the thalamocortical relay neurons from hyperpolarizing to burst firing mode. Furthermore, the VPM administered oxotremorine may decrease the hyperpolarization of thalamocortical relay neurons and prevent the activation of Ca²⁺ -mediated spiking and the generation of oscillatory burst firing. The effects of methoctramine on HVS activity, instead, may be mediated via increased release of acetylcholine as a consequence of the blockade of the presynaptic muscarinic M₂ inhibitory autoreceptors in the NRT, leading to an increased activation of post-synaptic muscarinic and nicotinic receptors. Indeed, it has been previously shown that intrathalamic infusion of the nicotinic acetylcholine receptor agonist, nicotine, dose-dependently decreased HVS activity (Riekkinen et al. 1995).

In the present study, both i.c.v and intrathalamically administered oxotremorine and McN-A-343 decreased HVS activity. These results are in agreement with previous studies showing that systemic administration of muscarinic agonists suppressed HVSs (Riekkinen et al. 1993, Jäkälä et al. 1997). These results suggest that systemically or i.c.v administered muscarinic agonists may modulate neocortical HVS via the thalamus. Interestingly, however, the effects of i.c.v. and intrathalamic administration of methoctramine on HVS activity were different: the i.c.v. infusion of methoctramine, at the highest dose used (100 mg), increased whereas the intrathalamic methoctramine decreased HVS activity. It is possible that the i.c.v. administered methoctramine suppresses the activity of NB and brainstem cholinergic neurons innervating the thalamus. This releases the pacemaker neurons of NRT from the control of NB and brainstem ascending cholinergic system and thereby evokes the appearance of the HVSs in the neocortex. Indeed, it has been shown that both NB-lesion and treatment with low doses of muscarinic antagonist, scopolamine, decrease the activity of cholinergic system and increase the neocortical HVS activity in rats (Buzsáki et al. 1988b, Riekkinen et al. 1991e). On the other hand, the intrathalamic methoctramine may block the presynaptic muscarinic M2 receptors in NRT, and thus decrease the HVS activity by increasing the available acetylcholine in the synapse. These results suggest that methoctramine regulates HVS activity at least partly via thalamic muscarinic M₂ receptors. However, muscarinic M₂ receptors in other brain structures, such as NB, brainstem and cortex, may also be involved in the regulation of thalamocortical arousal.

The present findings may have some relevance for clinical disorders, such as AD that are associated with cholinergic loss, EEG slowing and cognitive decline. Drugs that inhibit acetylcholinesterase activity are currently being used to compensate for cholinergic loss and to facilitate cognitive functioning and neuropsychiatric symptoms in AD patients (Francis et al. 1999, Levy et al. 1999). Recently, it has also been shown that neocortical EEG slow-wave activity (delta and theta) in AD patients was significantly reduced after chronic acetylcholinesterase inhibitor, THA, treatment (Jelic et al. 1998). The present findings provide further evidence that muscarinic M₁ agonists and M₂ antagonists may facilitate the functioning of cortical and thalamocortical systems, and compensate for damaged cholinergic activating input in AD.

6.3 NEUROPHYSIOLOGICAL FINDINGS IN APOE-DEFICIENT MICE

The present study investigated the effects of quisqualic acid induced NB-lesion and ageing on cortical EEG activity and the EEG response to a muscarinic antagonist, scopolamine, in apoE-deficient and control mice. The effects of NB-lesion on cortical EEG activity in mice were previously not known. The main finding in the NB-lesion study was that the NB-lesion affected EEG activity similarly in control and apoE-deficient mice, indicating that the apoE-deficiency does not increase the vulnerability of NB neurons involved in the regulation of cortical EEG activity. In rats, it has been previously shown that the excitotoxic lesions of NB increase the neocortical slow-waves and HVSs (Buzsáki et al. 1988b, Riekkinen et al. 1991e). Similarly to the situation in NB-lesioned rats, the slowing of EEG in AD has been shown to correlate with the loss of ChAT-activity in frontal cortex reflecting reduced cholinergic input (Soininen et al. 1992). In the present study, the quisqualic acid induced NB-lesion did not induce the slowing of the EEG activity in either control or apoE-deficient mice. In fact, a slight decrease in the relative amount of delta activity in both strains was found in the NB-lesioned hemisphere. It is possible that the 27-38 % cortical ChAT-activity depletion observed in the present study in NB-lesioned hemisphere was not extensive enough to induce shifting to lower EEG frequencies. Indeed, the recent findings in rats showing that IgG-saporin induced NB-lesion producing about 60 % depletion in cortical ChAT-activity also failed to induce cortical EEG slowing, suggest that the selective destruction of NB cholinergic neurons may not be sufficient to slow the overall EEG activity (Wenk et al. 1994). Furthermore, in AD patients the cortical EEG slowing may also be partly related to decreased cerebral blood flow and reduced glucose metabolism (Passero et al. 1995, Bartenstein et al. 1997, Claus et al. 2000). However, it is likely that also the species difference partly accounts for the contradictory findings from this study when compared to the findings in rat studies. Moreover, it has to be stated that the group size in NB-lesion study was small (n = 4-5), which may also complicate the interpretation of the results.

In the ageing study, the effects of muscarinic antagonist, scopolamine, on cortical EEG activity were measured in control and apoE-deficient mice. The scopolamine treatment of mice was associated with a significant shift to lower EEG frequencies. This result suggests that the EEG recording system used in this study was sensitive to shifts in the EEG power spectrum to lower frequencies, a finding that was also to be expected from the NB-lesion study. The results from this study that show EEG slowing in mice after scopolamine treatment are in agreement with previous rats studies showing that scopolamine could induce an increase in the neocortical slow-waves and HVSs (Buzsáki et al. 1988b, Riekkinen et al. 1991e). The main finding in the ageing study was that the scopolamine induced EEG slowing was clearly blunted in apoE-deficient mice, and the difference between the strains became slightly clearer with ageing. Previously, it has been shown that the hypothermic response to a muscarinic agonist, oxotremorine, was also reduced in apoE-deficient mice (Gordon et al. 1995). At the same time, there were no differences in hypothalamic ChAT-activity, suggesting that the reduced thermal response of these mice might be due to the deficiencies in post-synaptic hypothalamic muscarinic receptors (Gordon et al. 1995). In the present study, no differences were found in cortical ChAT-activity between control and apoE-deficient mice, suggesting that the activity of presynaptic cholinergic neurons in cortex is normal.

It can be argued that the blunted EEG slowing in apoE-deficient mice after scopolamine treatment might be due to the reduced number or function of cortical muscarinic receptors. However, in a recent study it was reported that the receptor binding sites for muscarinic M₁ and M₂ receptors as well as for nicotinic receptors were relatively intact in frontal and parieto-temporal cortex and basal forebrain cholinergic nuclei (MS and VDB) of apoE-deficient mice (Krzywkowski et al. 1999). Therefore, it is possible that the reduced EEG slowing in apoE-deficient mice after scopolamine treatment might not be due to the altered number of cortical or basal forebrain muscarinic receptors. It is likely that the alterations in muscarinic receptors in other brain structures, such as brainstem cholinergic nuclei and thalamus, which are also involved in the regulation of cortical arousal, may result in blunted scopolamine response in apoE-deficient mice. Unfortunately, no muscarinic receptor binding studies on these brain structures have been done in apoE-deficient mice.

Interestingly, in AD patients the slowing of EEG activity after scopolamine treatment is also less prominent than in controls, probably reflecting the reduced cholinergic tone in AD (Neufeld et al. 1994). Indeed, in AD patients, decreases in the number of cortical M₁ (Rodriguez-Puertas et al. 1997) and M₂ (Quirion et al. 1989) receptors and alterations in signalling pathways after muscarinic receptor activation (Ferrari-DiLeo et al. 1995) as well as cortical ChAT-activity depletion have been documented (Bowen et al. 1976). Moreover, in AD, the apoE e4 allele is associated with altered cortical arousal and EEG response to cholinergic therapy. It has been reported that the AD patients carrying the apoE e4 allele have more pronounced EEG slow-wave activity (Lehtovirta et al. 1996a, Lehtovirta et al. 2000) and decreased responsiveness to the increase in cortical arousal induced by THA (Riekkinen et al. 1997b). These and the present results further indicate that the cholinergic system and its interaction with apoE have an important role in the modulation of cortical arousal.

6.4 BEHAVIORAL FINDINGS IN APOE-DEFICIENT MICE

One of the aims of the present study was to investigate the effects of apoE-deficiency on spatial navigation in WM during ageing. The results show that both young and aged apoE-deficient mice could learn the location of a hidden platform equally well as the control mice. Furthermore, the spatial retention during the probe trial was also similar in both strains. The results from this study that show no differences in spatial navigation in WM between control and apoE-deficient mice are in agreement with previous studies (Anderson and Higgins 1997, Anderson et al. 1998) carried out with apoE-deficient mice derived from the same transgenic line (Piedrahita et al. 1992). These results indicate that the spatial learning of apoE-deficient mice of this strain is not affected during ageing.

However, there are a number of studies showing severe cognitive deficits in WM in apoE-deficient mice of other strains (Masliah et al. 1997, Fisher et al. 1998, Krzywkowski et al. 1999). The apoE-deficient mice have been shown to be impaired especially in the working memory component of the WM task (Gordon et al. 1995, Gordon et al. 1996). Recently, Oitzl et al. (1997) also showed that apoE-deficient mice were not able to learn the WM task at all. In that study, the apoE-deficient mice developed neither spatial nor other strategies to locate the platform, but rather an unusual repetitive behavioral pattern of wall bumping (Oitzl et al. 1997). The learning and memory deficits observed in these studies cannot be explained by motor impairment of apoE-deficient mice since their swimming speed and general motor activity are indistinguishable from those of controls (Gordon et al. 1995, Krzywkowski et al. 1999). Furthermore, control and apoE-deficient mice performed equally well in a non-spatial version of the WM task in which the animals were tested with a visible platform (Krzywkowski et al. 1999). This result suggests that apoE-deficient mice do not suffer from visual impairment, which would have explained their poor performance in the spatial version of the WM task.

Instead, the behavioral alterations are usually considered to be attributable to biochemical, neurophysiological and neuropathological alterations observed in the apoE-deficient mice. In the study of Gordon et al. (1995), the apoE-deficient mice were cognitively impaired in WM and had markedly decreased cholinergic activity in hippocampus and cortex. Interestingly, the cholinergic and memory deficits observed in apoE-deficient mice could be ameliorated back to control levels by treatment with a M₁ selective agonist (Fisher et al. 1998). On the other hand, Krzywkowski et al. (1999) found that cognitively impaired apoE-deficient mice had markedly reduced LTP in hippocampus, but no changes in their cholinergic system. Moreover, apoE-deficient mice display synaptic and dendritic alterations in hippocampus and neocortex associated with cognitive impairment in WM (Masliah et al. 1997). In the same study, it was reported that apoE-deficient mice treated with apoE showed a significant improvement in WM learning, which was associated with restoration of neuronal structure in hippocampus and neocortex (Masliah et al. 1997). These results suggest that apoE-deficient mice could provide a model for development of therapeutic treatments based on cholinergic replacement therapy or for understanding the neurotrophic effect of apoE in human neurodegenerative diseases such as AD. However, the conflicting results between studies showing normal or severely impaired cognitive functioning in apoE-deficient mice complicate the interpretation of function of apoE in humans.

The inconsistent results between studies can be attributable to a variety of factors. First, the housing conditions and diet might vary between the laboratories, which may influence the results. Second, the handling of mice before the experiment and the experimental procedures of WM task itself were different, making the comparison of the results between the studies difficult. Finally, and perhaps most importantly, the differences in the nature of transgenic and control lines (Piedrahita et al. 1992, Plump et al. 1992) may contribute to the behavioral phenotype (Gerlai 1996, Gerlai 2000). It is possible that in this and earlier studies (Anderson and Higgins 1997, Anderson et al. 1998, Fagan et al. 1998) conducted with very same strain of apoE-deficient mice (Piedrahita et al. 1992), the lack of major biochemical, neurophysiological or neuropathological alterations resulting in normal cognitive functioning in WM, is due to the different genetic background of the apoE-deficient mice.

The present study also investigated whether quisqualic acid induced NB-lesion was associated with more severe WM performance deficits and impaired behavioral recovery in apoE-deficient mice. The results show that the initial acquisition after short- and long-term recovery was similar in sham- and NB-lesioned control and apoE-deficient mice. However, the NB-lesioned mice from both strains were similarly impaired in the reversal learning and subsequent spatial retention after short-term recovery. Furthermore, the impairment in the reversal learning that was observed after short-term recovery was similarly reversed in both strains after long-term recovery. These results suggest that apoE-deficiency does not increase the sensitivity to spatial navigation deficits induced by the NB-lesion and does not impair the behavioral recovery.

A variety of different kinds of NB-lesion models have been used to study the role of the basal forebrain cholinergic system in learning and memory functions in rats (Dekker et al. 1991, Wenk 1997). However, there are only a few studies where the effects of NB-lesion on spatial navigation in mice have been studied (Sweeney et al. 1988, Sweeney et al. 1989). The results from this study are consistent with previous studies in which the NB-lesioned mice were significantly impaired in reversal learning, but not in the acquisition stage of the WM task (Sweeney et al. 1988, Sweeney et al. 1989). However, in rats, impairments both in acquisition and reversal learning or only in reversal learning after NB-lesion have been reported (Riekkinen et al. 1991c). Furthermore, in this study, the NB-lesioned mice showed no impairment in reversal learning after long-term recovery. In contrast, the quisqualic acid NB-lesioned rats were similarly impaired in reversal learning after short- and long-term recovery (Riekkinen et al. 1991c). The inconsistent behavioral findings in WM can partly be explained by differences in WM paradigms and in the extent of NB-lesions (e.g. different NB-lesion methods and coordinates) as well as by strain and species differences.

The finding that apoE-deficient and control mice were similarly impaired (short-term recovery) and recovered (long-term recovery) after the NB-lesion in WM spatial navigation, indicate that the apoE-deficiency does not predispose NB neurons to quisqualic acid induced excitotoxic trauma and that apoE is not required in the functional modification of those neuronal circuits responsible for the behavioral recovery after NB-lesion. These results differ from the findings of Chen et al. (1997) who found that apoE-deficient mice had more severe behavioral deficits than control mice in the WM task and neuronal cell death in hippocampus after closed head injury. Thus, it can be argued that the recovery of brain structures other than NB, such as hippocampal formation, which are also involved in spatial learning and memory, is more seriously affected by apoE-deficiency. However, it is also possible that other apolipoproteins can compensate for the loss of apoE (Popko et al. 1993).

6.5 BIOCHEMICAL AND NEUROPATHOLOGICAL FINDINGS IN APOE-DEFICIENT MICE

In the present study, the effects of apoE-deficiency on cholinergic activity during ageing and on biochemical recovery after quisqualic acid induced NB-lesion were investigated. Both young and aged apoE-deficient and control mice had similar cortical and/or hippocampal ChAT-activity levels and an equal number of ChAT-positive neurons in NB. Furthermore, the NB-lesion decreased cortical ChAT-activity and the number of ChAT-positive neurons in NB similarly in both strains. Finally, the recovery of cortical ChAT-activity levels after the NB-lesion was not affected in apoE-deficient mice. These results suggest that apoE-deficient mice do not have impairments in their cholinergic activity during ageing and that the apoE-deficiency does not increase the sensitivity to cholinergic deficits induced by NB-lesion nor does it affect the biochemical recovery.

In agreement with this study, it has been shown that apoE-deficient mice derived from the same colony (Piedrahita et al. 1992) do not have any significant alterations in their brain cholinergic activity during ageing compared to controls (Anderson and Higgins 1997), and that the number and size of cholinergic neurons in MS during ageing and after fimbria-fornix lesion induced axotomy were similar in control and apoE-deficient mice (Fagan et al. 1998). However, the findings from this study are in contrast to a number of previous reports performed with different strains of apoE-deficient mice that show markedly decreased ChAT- and AChE-activities in cortex and hippocampus (Gordon et al. 1995, Chapman and Michaelson 1998, Fisher et al. 1998). Interestingly, these cholinergic deficits in the apoE-deficient mice could be reversed with M₁ selective agonist treatment (Fisher et al. 1998). ApoE-deficient mouse strains may also differ in their response to cholinergic drugs. Gordon et al. (1995) demonstrated a blunted hypothermic response to oxotremorine in apoE-deficient mice, whereas Anderson & Higgins (1997), with the same transgenic line as used in this study, found a similar hypothermic response in control and apoE-deficient mice after treatment with the anticholinesterase donepezil.

The importance of apoE in the regulation of lipid metabolism in the brain is particularly important for the function of the cholinergic system because it is largely dependent on production of acetylcholine from membrane lipid components. In AD, the patients with apoE e4 allele(s) have more severe neuronal loss and less extensive sprouting of NB neurons (Arendt et al. 1997) associated with a greater cortical and hippocampal ChAT-activity decline (Soininen et al. 1995) compared to patients lacking the apoE e4 allele. In one study, it was shown that AD patients with the apoE e4 allele have decreased hippocampal and frontal cortical apoE levels (Bertrand et al. 1995). It could be argued that the decreased level of apoE in AD patients with the apoE e4 allele compromises the neuronal reorganisation and plasticity in response to NB degeneration. However, the results from this study showing that apoE-deficiency does not influence the function of cholinergic neurons after NB-lesion, support the idea that some mechanism other than 'loss of function' (i.e. decreased levels of apoE) is responsible for the severe cholinergic deficits and impaired compensatory sprouting of NB cholinergic cells seen in AD patients who carry the apoE e4 allele. It is possible that the different function of apoE4 compared to other apoE isoforms is responsible for the reduced synaptic plasticity in AD patients with the apoE e4 allele. Indeed, both in vitro cell culture studies showing that apoE3 but not apoE4 increased neurite extension (Nathan et al. 1994, Holtzman et al. 1995) and in vivo transgenic model studies indicating that apoE3 protected more effectively from excitotoxin induced neuronal damage than apoE4 (Buttini et al. 1999), support this conclusion.

AD is also associated with decreased activity of other neurotransmitter systems such as the noradrenergic (Mann 1983, Marcyniuk et al. 1986, Reinikainen et al. 1990) and serotonergic systems (Palmer et al. 1987, Nazarali and Reynolds 1992, Reinikainen et al. 1990) as well as synaptic loss (Terry et al. 1991) and disruption of the cytoskeleton (McKee et al. 1991). AD patients with the apoE e4 allele(s) show less plastic dendritic changes and more severe degeneration of LC and raphe magnus nucleus neurons (Arendt et al. 1997). Similarly, it has been shown that the apoE-deficient mice have decreased noradrenergic activity in the brain regions situated distal to LC, such as hippocampus and parietal cortex (Chapman and Michaelson 1998, Puoliväli et al. 2000a), and a decreased number of serotonergic nerve terminals in their parietal cortex (Chapman and Michaelson 1998). Moreover, apoE-deficient mice have impaired recovery of cortical and hippocampal noradrenaline levels after lesioning of LC with a noradrenergic neurotoxin (Puoliväli et al. 2000a). Several reports have suggested that apoE-deficient mice also suffer from alterations in synaptic structure and have decreased synaptic regeneration after CNS injuries. Masliah et al. (1995a and 1995b) found that MAP-2 immunoreactive dendrites and synaptophysin-immunoreactive nerve terminals were decreased in hippocampus and neocortex of apoE-deficient mice and that apoE-deficient mice displayed poor reparative ability after lesioning of the perforant pathway, which results in a loss of synaptic input to the hippocampus. Similarly, apoE-deficient mice also had more pronounced hippocampal cell death and impaired ability to recover after closed head injury (Chen et al. 1997). However, Anderson et al. (1998) showed similar reactive synaptogenesis in apoE-deficient mice after entorhinal cortex lesion compared to control mice, although the apoE-deficient mice initially showed a greater synaptic loss.

In summary, apoE-deficient mice may provide a useful model for investigation of some features of the biochemical and neuropathological changes observed in AD. However, the discrepancy in the results between different studies suggest that factors other than apoE-deficiency, such as the nature of the transgenic line and housing conditions, may be responsible for the differences observed between control and apoE-deficient mice in other studies. More recently, transgenic mice with AD-like neuropathology (e.g. mice expressing extracellular b-amyloid deposits, neuritic plaques, and human apoE4) and cognitive deficits have also been developed (Hsiao et al. 1996, Borchelt et al. 1997, Buttini et al. 1999, Raber et al. 2000). These transgenic mice that mimic several features of AD, offer further models to examine the neuropathological events that lead to the deterioration of cognitive function in AD and will provide new ways of testing potential therapeutic treatments. 7. CONCLUSIONS

The intrathalamic (NRT and VPM) infusions of a muscarinic M₁ receptor agonist, McN-A-343, a muscarinic M₂ receptor antagonist, methoctramine, and a muscarinic receptor agonist, oxotremorine, increased neocortical arousal as indicated by the decrease in thalamocortically generated HVSs. These results suggest that intrathalamic infusions of cholinergic muscarinic drugs modulate neocortical arousal via muscarinic M₁ and M₂ receptors in thalamus. Furthermore, these results also indicate that i.c.v. and systemically administered cholinergic drugs may modulate cortical EEG activity via the thalamus.

The apoE-deficient mice do not have any apparent impairment in their cholinergic activity (ChAT-activity, the number of ChAT-positive neurons in NB) or in WM spatial navigation during ageing. Furthermore, apoE-deficiency does not increase the sensitivity to cholinergic, behavioral or EEG deficits induced by NB-lesion. However, the apoE-deficient mice have slightly altered cortical EEG activity during ageing and blunted EEG response to scopolamine treatment. These results suggest that apoE does not have to be present to preserve the viability of cholinergic neurons and that spatial navigation and behavioral recovery during ageing or after NB-lesion are not affected by apoE-deficiency. However, apoE-deficiency might alter the regulation of cortical arousal during ageing and after cholinergic drug (scopolamine) manipulation.

In summary, the present studies provide new information about the function of the cholinergic system and its interaction with apoE. This may have relevance in the development of new experimental models and therapeutic treatments for clinical disorders such as Alzheimer's disease.

REFERENCES

American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th ed. Washington DC: American Psychiatric Association, 1994.

Anderson R, Barnes JC, Bliss TV, Cain DP, Cambon K, Davies HA, Errington ML, Fellows LA, Gray RA, Hoh T, Stewart M, Large CH and Higgins GA. Behavioural, physiological and morphological analysis of a line of apolipoprotein E knockout mouse. Neuroscience 1998;85:93-110.

Anderson R and Higgins GA. Absence of central cholinergic deficits in ApoE knockout mice. Psychopharmacology (Berl) 1997;132:135-144.

Arendt T, Bruckner MK, Bigl V and Marcova L. Dendritic reorganisation in the basal forebrain under degenerative conditions and its defects in Alzheimer's disease. II. Ageing, Korsakoff's disease, Parkinson's disease, and Alzheimer's disease. J Comp Neurol 1995;351:189-222.

Arendt T, Bruckner MK, Gertz HJ and Marcova L. Cortical distribution of neurofibrillary tangles in Alzheimer's disease matches the pattern of neurons that retain their capacity of plastic remodelling in the adult brain. Neuroscience 1998;83:991-1002.

Arendt T, Schindler C, Bruckner MK, Eschrich K, Bigl V, Zedlick D and Marcova L. Plastic neuronal remodeling is impaired in patients with Alzheimer's disease carrying apolipoprotein epsilon 4 allele. J Neurosci 1997;17:516-529.

Armstrong DM, Sheffield R, Buzsáki G, Chen KS, Hersh LB, Nearing B and Gage FH. Morphologic alterations of choline acetyltransferase-positive neurons in the basal forebrain of aged behaviorally characterized Fisher 344 rats. Neurobiol Aging 1993;14:457-470.

Arneric SP, Sullivan JP and Williams M. Neuronal Nicotinic Acetylcholine Receptors. Novel Targets for Central Nervous System Therapeutics. In: Bloom FE and Kupfer DJ, eds. Psychopharmacology: The Fourth Generation of Progress. New York: Raven Press, Ltd., 1995, pp. 95-110.

Aubert I, Araujo DM, Cecyre D, Robitaille Y, Gauthier S and Quirion R. Comparative alterations of nicotinic and muscarinic binding sites in Alzheimer's and Parkinson's diseases. J Neurochem 1992;58:529-541.

Bartenstein P, Minoshima S, Hirsch C, Buch K, Willoch F, Mosch D, Schad D, Schwaiger M and Kurz A. Quantitative assessment of cerebral blood flow in patients with Alzheimer's disease by SPECT. J Nucl Med 1997;38:1095-1101.

Bartus RT, Dean RLd, Beer B and Lippa AS. The cholinergic hypothesis of geriatric memory dysfunction. Science 1982;217:408-414.

Beaulieu C and Somogyi P. Enrichment of cholinergic synaptic terminals on GABAergic neurons and coexistence of immunoreactive GABA and choline acetyltransferase in the same synaptic terminals in the striate cortex of the cat. J Comp Neurol 1991;304:666-680.

Beffert U, Danik M, Krzywkowski P, Ramassamy C, Berrada F and Poirier J. The neurobiology of apolipoproteins and their receptors in the CNS and Alzheimer's disease. Brain Res Brain Res Rev 1998;27:119-142.

Bernstein D, Olton DS, Ingram DK, Waller SB, Reynolds MA and London ED. Radial maze performance in young and aged mice: neurochemical correlates. Pharmacol Biochem Behav 1985;22:301-307.

Bertrand P, Poirier J, Oda T, Finch CE and Pasinetti GM. Association of apolipoprotein E genotype with brain levels of apolipoprotein E and apolipoprotein J (clusterin) in Alzheimer disease. Brain Res Mol Brain Res 1995;33:174-178.

Björklund M, Sirviö J, Puoliväli J, Sallinen J, Jäkälä P, Scheinin M, Kobilka BK and Riekkinen P, Jr. Alpha2C-adrenoceptor-overexpressing mice are impaired in executing nonspatial and spatial escape strategies. Mol Pharmacol 1998;54:569-576.

Blokland A. Acetylcholine: a neurotransmitter for learning and memory? Brain Res Brain Res Rev 1995;21:285-300.

Borchelt DR, Ratovitski T, van Lare J, Lee MK, Gonzales V, Jenkins NA, Copeland NG, Price DL and Sisodia SS. Accelerated amyloid deposition in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precursor proteins. Neuron 1997;19:939-945.

Bowen DM, Smith CB, White P and Davison AN. Neurotransmitter-related enzymes and indices of hypoxia in senile dementia and other abiotrophies. Brain 1976;99:459-496.

Boyles JK, Zoellner CD, Anderson LJ, Kosik LM, Pitas RE, Weisgraber KH, Hui DY, Mahley RW, Gebicke-Haerter PJ, Ignatius MJ and et al. A role for apolipoprotein E, apolipoprotein A-I, and low density lipoprotein receptors in cholesterol transport during regeneration and remyelination of the rat sciatic nerve. J Clin Invest 1989;83:1015-1031.

Breitner JC, Welsh KA, Helms MJ, Gaskell PC, Gau BA, Roses AD, Pericak-Vance MA and Saunders AM. Delayed onset of Alzheimer's disease with nonsteroidal anti-inflammatory and histamine H2 blocking drugs. Neurobiol Aging 1995;16:523-530.

Butcher LL. Cholinergic Neurons and Networks. In: Paxinos G, ed. The Rat Nervous System. San Diego: Academic Press, Inc., 1995, pp. 1003-1015.

Buttini M, Orth M, Bellosta S, Akeefe H, Pitas RE, Wyss-Coray T, Mucke L and Mahley RW. Expression of human apolipoprotein E3 or E4 in the brains of Apoe-/- mice: isoform-specific effects on neurodegeneration. J Neurosci 1999;19:4867-4880.

Buzsáki G, Bickford RG, Armstrong DM, Ponomareff G, Chen KS, Ruiz R, Thal LJ and Gage FH. Electric activity in the neocortex of freely moving young and aged rats. Neuroscience 1988a;26:735-744.

Buzsáki G, Bickford RG, Ponomareff G, Thal LJ, Mandel R and Gage FH. Nucleus basalis and thalamic control of neocortical activity in the freely moving rat. J Neurosci 1988b;8:4007-4026.

Chapman S and Michaelson DM. Specific neurochemical derangements of brain projecting neurons in apolipoprotein E-deficient mice. J Neurochem 1998;70:708-714.

Chen Y, Lomnitski L, Michaelson DM and Shohami E. Motor and cognitive deficits in apolipoprotein E-deficient mice after closed head injury. Neuroscience 1997;80:1255-1262.

Chui HC, Bondareff W, Zarow C and Slager U. Stability of neuronal number in the human nucleus basalis of Meynert with age. Neurobiol Aging 1984;5:83-88.

Claus JJ, Ongerboer De Visser BW, Bour LJ, Walstra GJ, Hijdra A, Verbeeten B, Jr., Van Royen EA, Kwa VI and van Gool WA. Determinants of quantitative spectral electroencephalography in early Alzheimer's disease: cognitive function, regional cerebral blood flow, and computed tomography. Dement Geriatr Cogn Disord 2000;11:81-89.

Coben LA, Danziger WL and Berg L. Frequency analysis of the resting awake EEG in mild senile dementia of Alzheimer type. Electroencephalogr Clin Neurophysiol 1983;55:372-380.

Corder EH, Saunders AM, Risch NJ, Strittmatter WJ, Schmechel DE, Gaskell PC, Jr., Rimmler JB, Locke PA, Conneally PM, Schmader KE and et al. Protective effect of apolipoprotein E type 2 allele for late onset Alzheimer disease. Nat Genet 1994;7:180-184.

Crawley JN, Belknap JK, Collins A, Crabbe JC, Frankel W, Henderson N, Hitzemann RJ, Maxson SC, Miner LL, Silva AJ, Wehner JM, Wynshaw-Boris A and Paylor R. Behavioral phenotypes of inbred mouse strains: implications and recommendations for molecular studies. Psychopharmacology (Berl) 1997;132:107-124.

Cummings BJ, Pike CJ, Shankle R and Cotman CW. Beta-amyloid deposition and other measures of neuropathology predict cognitive status in Alzheimer's disease. Neurobiol Aging 1996;17:921-933.

Cummings JL, Vinters HV, Cole GM and Khachaturian ZS. Alzheimer's disease: etiologies, pathophysiology, cognitive reserve, and treatment opportunities. Neurology 1998;51:S2-17.

Danober L, Depaulis A, Marescaux C and Vergnes M. Effects of cholinergic drugs on genetic absence seizures in rats. Eur J Pharmacol 1993;234:263-268.

Davies P. Neurotransmitter-related enzymes in senile dementia of the Alzheimer type. Brain Res 1979;171:319-327.

Davies P and Maloney AJ. Selective loss of central cholinergic neurons in Alzheimer's disease. Lancet 1976;2:1403.

Dekker AJ, Connor DJ and Thal LJ. The role of cholinergic projections from the nucleus basalis in memory. Neurosci Biobehav Rev 1991;15:299-317.

De Lacalle S, Cooper JD, Svendsen CN, Dunnett SB and Sofroniew MV. Reduced retrograde labelling with fluorescent tracer accompanies neuronal atrophy of basal forebrain cholinergic neurons in aged rats. Neuroscience 1996;75:19-27.

De Lacalle S, Iraizoz I and Ma Gonzalo L. Differential changes in cell size and number in topographic subdivisions of human basal nucleus in normal aging. Neuroscience 1991;43:445-456.

Détári L, Rasmusson DD and Semba K. The role of basal forebrain neurons in tonic and phasic activation of the cerebral cortex. Prog Neurobiol 1999;58:249-277.

Dornan WA, McCampbell AR, Tinkler GP, Hickman LJ, Bannon AW, Decker MW and Gunther KL. Comparison of site-specific injections into the basal forebrain on water maze and radial arm maze performance in the male rat after immunolesioning with 192 IgG saporin. Behav Brain Res 1996;82:93-101.

Dudchenko PA, Goodridge JP, Seiterle DA and Taube JS. Effects of repeated disorientation on the acquisition of spatial tasks in rats: dissociation between the appetitive radial arm maze and aversive water maze. J Exp Psychol Anim Behav Process 1997;23:194-210.

Dunnett SB, Everitt BJ and Robbins TW. The basal forebrain-cortical cholinergic system: interpreting the functional consequences of excitotoxic lesions. Trends Neurosci 1991;14:494-501.

Ebert U and Kirch W. Scopolamine model of dementia: electroencephalogram findings and cognitive performance. Eur J Clin Invest 1998;28:944-949.

Eckenstein FP, Baughman RW and Quinn J. An anatomical study of cholinergic innervation in rat cerebral cortex. Neuroscience 1988;25:457-474.

Ehlert FJ, Roeske WR and Yamamura HI. Molecular Biology, Pharmacology, and Brain Distribution of Subtypes of the Muscarinic Receptor. In: Bloom FE and Kupfer DJ, eds. Psychopharmacology: The Fourth Generation of Progress. New York: Raven Press, Ltd., 1995, pp. 111-124.

Everitt BJ and Robbins TW. Central cholinergic systems and cognition. Annu Rev Psychol 1997;48:649-684.

Fagan AM, Murphy BA, Patel SN, Kilbridge JF, Mobley WC, Bu G and Holtzman DM. Evidence for normal aging of the septo-hippocampal cholinergic system in apoE (-/-) mice but impaired clearance of axonal degeneration products following injury. Exp Neurol 1998;151:314-325.

Felder CC. Muscarinic acetylcholine receptors: signal transduction through multiple effectors. Faseb J 1995;9:619-625.

Ferrari-DiLeo G, Mash DC and Flynn DD. Attenuation of muscarinic receptor-G-protein interaction in Alzheimer disease. Mol Chem Neuropathol 1995;24:69-91.

Fisher A, Brandeis R, Chapman S, Pittel Z and Michaelson DM. M1 muscarinic agonist treatment reverses cognitive and cholinergic impairments of apolipoprotein E-deficient mice. J Neurochem 1998;70:1991-1997.

Flynn DD, Ferrari-DiLeo G, Mash DC and Levey AI. Differential regulation of molecular subtypes of muscarinic receptors in Alzheimer's disease. J Neurochem 1995;64:1888-1891.

Fonnum FA rapid radiochemical method for the determination of choline acetyltransferase. J Neurochem 1975;24:407-409.

Francis PT, Palmer AM, Snape M and Wilcock GK. The cholinergic hypothesis of Alzheimer's disease: a review of progress. J Neurol Neurosurg Psychiatry 1999;66:137-147.

Freeman SE and Dawson RM. Tacrine: a pharmacological review. Prog Neurobiol 1991;36:257-277.

Frick KM, Baxter MG, Markowska AL, Olton DS and Price DL. Age-related spatial reference and working memory deficits assessed in the water maze. Neurobiol Aging 1995;16:149-160.

Gerlai R. Gene-targeting studies of mammalian behavior: is it the mutation or the background genotype? Trends Neurosci 1996;19:177-181.

Gerlai R. Targeting genes and proteins in the analysis of learning and memory: caveats and future directions. Rev Neurosci 2000;11:15-26.

Geula C and Mesulam MM. Systematic regional variations in the loss of cortical cholinergic fibers in Alzheimer's disease. Cereb Cortex 1996;6:165-177.

Geula C and Mesulam MM. Cholinergic Systems in Alzheimer Disease. In: Terry RD, Katzman R, Bick KL and Sisodia SS, eds. Alzheimer Disease. Philadelphia: Lippincott Williams & Wilkins, 1999, pp. 269-292.

Giraldo E, Micheletti R, Montagna E, Giachetti A, Vigano MA, Ladinsky H and Melchiorre C. Binding and functional characterization of the cardioselective muscarinic antagonist methoctramine. J Pharmacol Exp Ther 1988;244:1016-1020.

Gomez-Isla T, Hollister R, West H, Mui S, Growdon JH, Petersen RC, Parisi JE and Hyman BT. Neuronal loss correlates with but exceeds neurofibrillary tangles in Alzheimer's disease. Ann Neurol 1997;41:17-24.

Gomez-Isla T, West HL, Rebeck GW, Harr SD, Growdon JH, Locascio JJ, Perls TT, Lipsitz LA and Hyman BT. Clinical and pathological correlates of apolipoprotein E epsilon 4 in Alzheimer's disease. Ann Neurol 1996;39:62-70.

Goodrum JF, Bouldin TW, Zhang SH, Maeda N and Popko B. Nerve regeneration and cholesterol reutilization occur in the absence of apolipoproteins E and A-I in mice. J Neurochem 1995;64:408-416.

Gordon I, Genis I, Grauer E, Sehayek E and Michaelson DM. Biochemical and cognitive studies of apolipoprotein-E-deficient mice. Mol Chem Neuropathol 1996;28:97-103.

Gordon I, Grauer E, Genis I, Sehayek E and Michaelson DM. Memory deficits and cholinergic impairments in apolipoprotein E-deficient mice. Neurosci Lett 1995;199:1-4.

Hamanaka H, Katoh-Fukui Y, Suzuki K, Kobayashi M, Suzuki R, Motegi Y, Nakahara Y, Takeshita A, Kawai M, Ishiguro K, Yokoyama M and Fujita SC. Altered cholesterol metabolism in human apolipoprotein E4 knock-in mice. Hum Mol Genet 2000;9:353-361.

Heinonen O, Lehtovirta M, Soininen H, Helisalmi S, Mannermaa A, Sorvari H, Kosunen O, Paljärvi L, Ryynänen M and Riekkinen PJ, Sr. Alzheimer pathology of patients carrying apolipoprotein E epsilon 4 allele. Neurobiol Aging 1995;16:505-513.

Hodges H. Maze procedures: the radial-arm and water maze compared. Brain Res Cogn Brain Res 1996;3:167-181.

Hohmann CF and Ebner FF. Development of cholinergic markers in mouse forebrain. I. Choline acetyltransferase enzyme activity and acetylcholinesterase histochemistry. Brain Res 1985;355:225-241.

Hohmann CF, Potter ED and Levey AI. Development of muscarinic receptor subtypes in the forebrain of the mouse. J Comp Neurol 1995;358:88-101.

Holcomb LA, Gordon MN, Jantzen P, Hsiao K, Duff K and Morgan D. Behavioral changes in transgenic mice expressing both amyloid precursor protein and presenilin-1 mutations: lack of association with amyloid deposits. Behav Genet 1999;29:177-185.

Holtzman DM, Pitas RE, Kilbridge J, Nathan B, Mahley RW, Bu G and Schwartz AL. Low density lipoprotein receptor-related protein mediates apolipoprotein E-dependent neurite outgrowth in a central nervous system-derived neuronal cell line. Proc Natl Acad Sci U S A 1995;92:9480-9484.

Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F and Cole G. Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 1996;274:99-102.

Hyman BT, Gomez-Isla T, Rebeck GW, Briggs M, Chung H, West HL, Greenberg S, Mui S, Nichols S, Wallace R and Growdon JH. Epidemiological, clinical, and neuropathological study of apolipoprotein E genotype in Alzheimer's disease. Ann N Y Acad Sci 1996;802:1-5.

Ignatius MJ, Gebicke-Harter PJ, Skene JH, Schilling JW, Weisgraber KH, Mahley RW and Shooter EM. Expression of apolipoprotein E during nerve degeneration and regeneration. Proc Natl Acad Sci U S A 1986;83:1125-1129.

Ikonen S, Schmidt BH and Riekkinen P, Jr. Characterization of learning and memory behaviors and the effects of metrifonate in the C57BL strain of mice. Eur J Pharmacol 1999;372:117-126.

Ingram DK and Corfman TP. An overview of neurobiological comparisons in mouse strains. Neurosci Biobehav Rev 1980;4:421-435.

Jelic V, Dierks T, Amberla K, Almkvist O, Winblad B and Nordberg A. Longitudinal changes in quantitative EEG during long-term tacrine treatment of patients with Alzheimer's disease. Neurosci Lett 1998;254:85-88.

Jones SV. Muscarinic receptor subtypes: modulation of ion channels. Life Sci 1993;52:457-464.

Jäkälä P, Björklund M, Koivisto E and Riekkinen P, Jr. The effects of cholinergic drugs on rat neocortical high-voltage spindles in ketanserin-treated rats. Eur J Pharmacol 1996a;316:181-193.

Jäkälä P, Björklund M and Riekkinen P, Jr. Suppression of neocortical high-voltage spindles by nicotinic acetylcholine and 5-HT2 receptor stimulation. Eur J Pharmacol 1996b;299:47-60.

Jäkälä P, Puoliväli J, Björklund M, Koivisto E and Riekkinen P, Jr. Activation of acetylcholine receptors and 5-HT2 receptors have additive effects in the suppression of neocortical high-voltage spindles in aged rats. Psychopharmacology (Berl) 1997;132:270-280.

Jäkälä P, Sirviö J, Koivisto E, Björklund M, Kaukua J and Riekkinen P, Jr. Modulation of rat neocortical high-voltage spindle activity by 5-HT1/5-HT2 receptor subtype specific drugs. Eur J Pharmacol 1995;282:39-55.

Jäkälä P, Viitamaa T, Sirviö J, Riekkinen P, Jr., Salonen J, Haapalinna A, Virtanen R and Riekkinen P. Continuous alpha 2-adrenoceptor blockade by atipamezole decreases neocortical high-voltage spindle activity in rats. Eur J Pharmacol 1992;221:351-357.

Kitt CA, Hohmann C, Coyle JT and Price DL. Cholinergic innervation of mouse forebrain structures. J Comp Neurol 1994;341:117-129.

Koliatsos VE, Martin LJ and Price DL. Efferent organization of the mammalian basal forebrain. In: Steriade M and Biesold D, eds. Brain cholinergic system. Oxford: Oxford University Press, 1990, pp. 120-152.

Krugers HJ, Mulder M, Korf J, Havekes L, de Kloet ER and Joels M. Altered synaptic plasticity in hippocampal CA1 area of apolipoprotein E deficient mice. Neuroreport 1997;8:2505-2510.

Krzywkowski P, Ghribi O, Gagne J, Chabot C, Kar S, Rochford J, Massicotte G and Poirier J. Cholinergic systems and long-term potentiation in memory-impaired apolipoprotein E-deficient mice. Neuroscience 1999;92:1273-1286.

Lambrecht G, Moser U, Grimm U, Pfaff O, Hermanni U, Hildebrandt C, Waelbroeck M, Christophe J and Mutschler E. New functionally selective muscarinic agonists. Life Sci 1993;52:481-488.

Leanza G, Muir J, Nilsson OG, Wiley RG, Dunnett SB and Björklund A. Selective immunolesioning of the basal forebrain cholinergic system disrupts short-term memory in rats. Eur J Neurosci 1996;8:1535-1544.

Leanza G, Nilsson OG, Wiley RG and Björklund A. Selective lesioning of the basal forebrain cholinergic system by intraventricular 192 IgG-saporin: behavioural, biochemical and stereological studies in the rat. Eur J Neurosci 1995;7:329-343.

Lebrun C, Durkin TP, Marighetto A and Jaffard R. A comparison of the working memory performances of young and aged mice combined with parallel measures of testing and drug-induced activations of septo-hippocampal and nbm-cortical cholinergic neurones. Neurobiol Aging 1990;11:515-521.

Lehtovirta M, Partanen J, Kononen M, Hiltunen J, Helisalmi S, Hartikainen P, Riekkinen P, Sr. and Soininen H. A longitudinal quantitative EEG study of Alzheimer's disease: relation to apolipoprotein E polymorphism. Dement Geriatr Cogn Disord 2000;11:29-35.

Lehtovirta M, Partanen J, Kononen M, Soininen H, Helisalmi S, Mannermaa A, Ryynänen M, Hartikainen P and Riekkinen P, Sr. Spectral analysis of EEG in Alzheimer's disease: relation to apolipoprotein E polymorphism. Neurobiol Aging 1996a;17:523-526.

Lehtovirta M, Soininen H, Helisalmi S, Mannermaa A, Helkala EL, Hartikainen P, Hänninen T, Ryynänen M and Riekkinen PJ. Clinical and neuropsychological characteristics in familial and sporadic Alzheimer's disease: relation to apolipoprotein E polymorphism. Neurology 1996b;46:413-419.

Lehtovirta M, Soininen H, Laakso MP, Partanen K, Helisalmi S, Mannermaa A, Ryynänen M, Kuikka J, Hartikainen P and Riekkinen PJ, Sr. SPECT and MRI analysis in Alzheimer's disease: relation to apolipoprotein E epsilon 4 allele. J Neurol Neurosurg Psychiatry 1996c;60:644-649.

Levey AI. Muscarinic acetylcholine receptor expression in memory circuits: implications for treatment of Alzheimer disease. Proc Natl Acad Sci U S A 1996;93:13541-13546.

Levey AI, Hallanger AE and Wainer BH. Choline acetyltransferase immunoreactivity in the rat thalamus. J Comp Neurol 1987;257:317-332.

Levy ML, Cummings JL and Kahn-Rose R. Neuropsychiatric symptoms and cholinergic therapy for Alzheimer's disease. Gerontology 1999;45:15-22.

Li M, Yasuda RP, Wall SJ, Wellstein A and Wolfe BB. Distribution of m2 muscarinic receptors in rat brain using antisera selective for m2 receptors. Mol Pharmacol 1991;40:28-35.

Lipp HP and Wolfer DP. Genetically modified mice and cognition. Curr Opin Neurobiol 1998;8:272-280.

Lysakowski A, Wainer BH, Rye DB, Bruce G and Hersh LB. Cholinergic innervation displays strikingly different laminar preferences in several cortical areas. Neurosci Lett 1986;64:102-108.

Mann DM. The locus coeruleus and its possible role in ageing and degenerative disease of the human central nervous system. Mech Ageing Dev 1983;23:73-94.

Manns ID, Alonso A and Jones BE. Discharge properties of juxtacellularly labeled and immunohistochemically identified cholinergic basal forebrain neurons recorded in association with the electroencephalogram in anesthetized rats. J Neurosci 2000;20:1505-1518.

Marcyniuk B, Mann DM and Yates PO. The topography of cell loss from locus caeruleus in Alzheimer's disease. J Neurol Sci 1986;76:335-345.

Masliah E, Mallory M, Alford N, Ge N and Mucke L. Abnormal synaptic regeneration in hAPP695 transgenic and APOE knockout mice. In: Iqbal K, Mortimer JA, Winblad B and Wisniewski HM, eds. Research Advances in Alzheimer's Disease and Related Disorders. New York: John Wiley & Sons Ltd., 1995a, pp. 405-414.

Masliah E, Mallory M, Ge N, Alford M, Veinbergs I and Roses AD. Neurodegeneration in the central nervous system of apoE-deficient mice. Exp Neurol 1995b;136:107-122.

Masliah E, Samuel W, Veinbergs I, Mallory M, Mante M and Saitoh T. Neurodegeneration and cognitive impairment in apoE-deficient mice is ameliorated by infusion of recombinant apoE. Brain Res 1997;751:307-314.

Matthews DA, Cotman C and Lynch G. An electron microscopic study of lesion-induced synaptogenesis in the dentate gyrus of the adult rat. II. Reappearance of morphologically normal synaptic contacts. Brain Res 1976;115:23-41.

McCormick DA. Cellular mechanisms of cholinergic control of neocortical and thalamic neuronal excitability. In: Steriade M and Biesold D, eds. Brain cholinergic system. Oxford: Oxford University Press, 1990, pp. 236-264.

McCormick DA. Cellular mechanisms underlying cholinergic and noradrenergic modulation of neuronal firing mode in the cat and guinea pig dorsal lateral geniculate nucleus. J Neurosci 1992a;12:278-289.

McCormick DA. Neurotransmitter actions in the thalamus and cerebral cortex and their role in neuromodulation of thalamocortical activity. Prog Neurobiol 1992b;39:337-388.

McCormick DA and Bal T. Sleep and arousal: thalamocortical mechanisms. Annu Rev Neurosci 1997;20:185-215.

McGaughly J, Dalley JW, Everitt BJ and Robbins TW. Behavioural and neurochemical evidence for role of cortical acetylcholine in visual attention. Br J Pharmacol 1999;128 Suppl:218P.

McKee AC, Kosik KS and Kowall NW. Neuritic pathology and dementia in Alzheimer's disease. Ann Neurol 1991;30:156-165.

Mesulam MM, Mufson EJ and Rogers J. Age-related shrinkage of cortically projecting cholinergic neurons: a selective effect. Ann Neurol 1987;22:31-36.

Mesulam MM, Mufson EJ, Wainer BH and Levey AI. Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Ch1-Ch6). Neuroscience 1983;10:1185-1201.

Meyer EM, St. Onge E and Crews FT. Effects of aging on rat cortical presynaptic cholinergic processes. Neurobiol Aging 1984;5:315-317.

Michel AD and Whiting RL. Methoctramine, a polymethylene tetraamine, differentiates three subtypes of muscarinic receptor in direct binding studies. Eur J Pharmacol 1988;145:61-66.

Micheletti R and Schiavone A. Functional determination of McN-A-343 affinity for M1 muscarinic receptors. J Pharmacol Exp Ther 1990;253:310-314.

Morris R. Developments of a water-maze procedure for studying spatial learning in the rat. J Neurosci Methods 1984;11:47-60.

Morris RG, Garrud P, Rawlins JN and Keefe O. Place navigation impaired in rats with hippocampal lesions. Nature 1982;297:681-683.

Muir JL, Dunnett SB, Robbins TW and Everitt BJ. Attentional functions of the forebrain cholinergic systems: effects of intraventricular hemicholinium, physostigmine, basal forebrain lesions and intracortical grafts on a multiple-choice serial reaction time task. Exp Brain Res 1992;89:611-622.

Muir JL, Page KJ, Sirinathsinghji DJ, Robbins TW and Everitt BJ. Excitotoxic lesions of basal forebrain cholinergic neurons: effects on learning, memory and attention. Behav Brain Res 1993;57:123-131.

Nagy Z, Esiri MM, Jobst KA, Johnston C, Litchfield S, Sim E and Smith AD. Influence of the apolipoprotein E genotype on amyloid deposition and neurofibrillary tangle formation in Alzheimer's disease. Neuroscience 1995;69:757-761.

Nathan BP, Bellosta S, Sanan DA, Weisgraber KH, Mahley RW and Pitas RE. Differential effects of apolipoproteins E3 and E4 on neuronal growth in vitro. Science 1994;264:850-852.

Nazarali AJ and Reynolds GP. Monoamine neurotransmitters and their metabolites in brain regions in Alzheimer's disease: a postmortem study. Cell Mol Neurobiol 1992;12:581-587.

Neufeld MY, Rabey MJ, Parmet Y, Sifris P, Treves TA and Korczyn AD. Effects of a single intravenous dose of scopolamine on the quantitative EEG in Alzheimer's disease patients and age-matched controls. Electroencephalogr Clin Neurophysiol 1994;91:407-412.

Ohm TG, Kirca M, Bohl J, Scharnagl H, Gross W and Marz W. Apolipoprotein E polymorphism influences not only cerebral senile plaque load but also Alzheimer-type neurofibrillary tangle formation. Neuroscience 1995;66:583-587.

Oitzl MS, Mulder M, Lucassen PJ, Havekes LM, Grootendorst J and de Kloet ER. Severe learning deficits in apolipoprotein E-knockout mice in a water maze task. Brain Res 1997;752:189-196.

Palmer AM, Francis PT, Benton JS, Sims NR, Mann DM, Neary D, Snowden JS and Bowen DM. Presynaptic serotonergic dysfunction in patients with Alzheimer's disease. J Neurochem 1987;48:8-15.

Passero S, Rocchi R, Vatti G, Burgalassi L and Battistini N. Quantitative EEG mapping, regional cerebral blood flow, and neuropsychological function in Alzheimer's disease. Dementia 1995;6:148-156.

Penttilä M, Partanen JV, Soininen H and Riekkinen PJ. Quantitative analysis of occipital EEG in different stages of Alzheimer's disease. Electroencephalogr Clin Neurophysiol 1985;60:1-6.

Perry E, Walker M, Grace J and Perry R. Acetylcholine in mind: a neurotransmitter correlate of consciousness? Trends Neurosci 1999;22:273-280.

Perry EK, Blessed G, Tomlinson BE, Perry RH, Crow TJ, Cross AJ, Dockray GJ, Dimaline R and Arregui A. Neurochemical activities in human temporal lobe related to aging and Alzheimer-type changes. Neurobiol Aging 1981;2:251-256.

Piedrahita JA, Zhang SH, Hagaman JR, Oliver PM and Maeda N. Generation of mice carrying a mutant apolipoprotein E gene inactivated by gene targeting in embryonic stem cells. Proc Natl Acad Sci U S A 1992;89:4471-4475.

Plump AS, Smith JD, Hayek T, Aalto-Setala K, Walsh A, Verstuyft JG, Rubin EM and Breslow JL. Severe hypercholesterolemia and atherosclerosis in apolipoprotein E-deficient mice created by homologous recombination in ES cells. Cell 1992;71:343-353.

Poirier J. Apolipoprotein E in animal models of CNS injury and in Alzheimer's disease. Trends Neurosci 1994;17:525-530.

Poirier J, Baccichet A, Dea D and Gauthier S. Cholesterol synthesis and lipoprotein reuptake during synaptic remodelling in hippocampus in adult rats. Neuroscience 1993;55:81-90.

Poirier J, Delisle MC, Quirion R, Aubert I, Farlow M, Lahiri D, Hui S, Bertrand P, Nalbantoglu J, Gilfix BM and et al. Apolipoprotein E4 allele as a predictor of cholinergic deficits and treatment outcome in Alzheimer disease. Proc Natl Acad Sci U S A 1995;92:12260-12264.

Poirier J, Hess M, May PC and Finch CE. Astrocytic apolipoprotein E mRNA and GFAP mRNA in hippocampus after entorhinal cortex lesioning. Brain Res Mol Brain Res 1991;11:97-106.

Popko B, Goodrum JF, Bouldin TW, Zhang SH and Maeda N. Nerve regeneration occurs in the absence of apolipoprotein E in mice. J Neurochem 1993;60:1155-1158.

Puoliväli J, Pradier L and Riekkinen P, Jr. Impaired recovery of noradrenaline levels in apolipoprotein E-deficient mice after N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine lesion. Neuroscience 2000a;95:353-358.

Puoliväli J, Tanila H, van Groen T and de Bruin J. Water maze testing in the mouse, role of training procedures, In: Behavioural Phenotyping of Mouse Mutants, ed. Office of the Associate Dean of Scientific Affairs University of Cologne, Germany, 2000b, p. 82.

Quirion R, Aubert I, Lapchak PA, Schaum RP, Teolis S, Gauthier S and Araujo DM. Muscarinic receptor subtypes in human neurodegenerative disorders: focus on Alzheimer's disease. Trends Pharmacol Sci 1989;Suppl:80-84.

Raber J, Wong D, Buttini M, Orth M, Bellosta S, Pitas RE, Mahley RW and Mucke L. Isoform-specific effects of human apolipoprotein E on brain function revealed in ApoE knockout mice: increased susceptibility of females. Proc Natl Acad Sci U S A 1998;95:10914-10919.

Raber J, Wong D, Yu GQ, Buttini M, Mahley RW, Pitas RE and Mucke L. Apolipoprotein E and cognitive performance. Nature 2000;404:352-354.

Reinikainen KJ, Soininen H and Riekkinen PJ. Neurotransmitter changes in Alzheimer's disease: implications to diagnostics and therapy. J Neurosci Res 1990;27:576-586.

Richardson RT and DeLong MR. Electrophysiological studies of the functions of the nucleus basalis in primates. Adv Exp Med Biol 1991;295:233-252.

Riekkinen P, Jr., Aaltonen M, Sirviö J and Riekkinen P. Tetrahydroaminoacridine alleviates medial septal lesion-induced and age-related spatial reference but not working memory deficits. Physiol Behav 1991a;49:1147-1152.

Riekkinen P, Buzsáki G, Riekkinen P, Jr., Soininen H and Partanen J. The cholinergic system and EEG slow waves. Electroencephalogr Clin Neurophysiol 1991b;78:89-96.

Riekkinen P, Jr., Kuitunen J and Riekkinen M. Effects of thalamic and nucleus basalis infusions of nicotine on cortical EEG. Neuroreport 1995;6:1625-1628.

Riekkinen P, Jr., Riekkinen M, Fisher A, Ekonsalo T and Sirviö J. Effects of muscarinic receptor agonists and anticholinesterase drugs on high voltage spindles and slow waves. Eur J Pharmacol 1993;240:1-7.

Riekkinen P, Jr., Riekkinen M, Sirviö J, Miettinen R and Riekkinen P. Comparison of the effects of acute and chronic ibotenic and quisqualic acid nucleus basalis lesioning. Brain Res Bull 1991c;27:199-206.

Riekkinen P, Jr., Riekkinen M, Sirviö J, Miettinen R and Riekkinen P. Loss of cholinergic neurons in the nucleus basalis induces neocortical electroencephalographic and passive avoidance deficits. Neuroscience 1992;47:823-831.

Riekkinen P, Jr., Schmidt B, Jäkälä P, Koivisto E and Björklund M. Chronic nimodipine and acute metrifonate treatment decreases age-related cortical high voltage spindles in rats. Psychopharmacology (Berl) 1997a;129:91-95.

Riekkinen P, Jr., Schmidt B, Stefanski R, Kuitunen J and Riekkinen M. Metrifonate improves spatial navigation and avoidance behavior in scopolamine-treated, medial septum-lesioned and aged rats. Eur J Pharmacol 1996;309:121-130.

Riekkinen P, Jr., Sirviö J, Hannila T, Miettinen R and Riekkinen P. Effects of quisqualic acid nucleus basalis lesioning on cortical EEG, passive avoidance and water maze performance. Brain Res Bull 1990;24:839-842.

Riekkinen P, Jr., Sirviö J, Jäkälä P, Riekkinen M, Lammintausta R and Riekkinen P. Effects of alpha 2-drugs and pilocarpine on the high-voltage spindle activity of young and aged control and DSP4-lesioned rats. Physiol Behav 1991d;50:955-959.

Riekkinen P, Jr., Sirviö J, Valjakka A, Riekkinen M, Lammintausta R and Riekkinen P. Effects of atipamezole and tetrahydroaminoacridine on nucleus basalis lesion-induced EEG changes. Brain Res Bull 1991e;27:231-235.

Riekkinen P, Jr., Soininen H, Partanen J, Pääkkönen A, Helisalmi S and Riekkinen P, Sr. The ability of THA treatment to increase cortical alpha waves is related to apolipoprotein E genotype of Alzheimer disease patients. Psychopharmacology (Berl) 1997b;129:285-288.

Rinne JO, Lönnberg P, Marjamäki P and Rinne UK. Brain muscarinic receptor subtypes are differently affected in Alzheimer's disease and Parkinson's disease. Brain Res 1989;483:402-406.

Rinne JO, Myllykylä T, Lönnberg P and Marjamäki P. A postmortem study of brain nicotinic receptors in Parkinson's and Alzheimer's disease. Brain Res 1991;547:167-170.

Rinne JO, Paljärvi L and Rinne UK. Neuronal size and density in the nucleus basalis of Meynert in Alzheimer's disease. J Neurol Sci 1987;79:67-76.

Rissanen A, Puoliväli J, van Groen T and Riekkinen P, Jr. In mice tonic estrogen replacement therapy improves non-spatial and spatial memory in a water maze task. Neuroreport 1999;10:1369-1372.

Robbins TW, Everitt BJ, Marston HM, Wilkinson J, Jones GH and Page KJ. Comparative effects of ibotenic acid- and quisqualic acid-induced lesions of the substantia innominata on attentional function in the rat: further implications for the role of the cholinergic neurons of the nucleus basalis in cognitive processes. Behav Brain Res 1989;35:221-240.

Rodriguez-Puertas R, Pascual J, Vilaro T and Pazos A. Autoradiographic distribution of M1, M2, M3, and M4 muscarinic receptor subtypes in Alzheimer's disease. Synapse 1997;26:341-350.

Sarter M and Bruno JP. Age-related changes in rodent cortical acetylcholine and cognition: main effects of age versus age as an intervening variable. Brain Res Brain Res Rev 1998;27:143-156.

Schmechel DE, Saunders AM, Strittmatter WJ, Crain BJ, Hulette CM, Joo SH, Pericak-Vance MA, Goldgaber D and Roses AD. Increased amyloid beta-peptide deposition in cerebral cortex as a consequence of apolipoprotein E genotype in late-onset Alzheimer disease. Proc Natl Acad Sci U S A 1993;90:9649-9653.

Schwegler H, Boldyreva M, Linke R, Wu J, Zilles K and Crusio WE. Genetic variation in the morphology of the septo-hippocampal cholinergic and GABAergic systems in mice: II. Morpho-behavioral correlations. Hippocampus 1996a;6:535-545.

Schwegler H, Boldyreva M, Pyrlik-Gohlmann M, Linke R, Wu J and Zilles K. Genetic variation in the morphology of the septo-hippocampal cholinergic and GABAergic system in mice. I. Cholinergic and GABAergic markers. Hippocampus 1996b;6:136-148.

Sherman KA and Friedman E. Pre- and post-synaptic cholinergic dysfunction in aged rodent brain regions: new findings and an interpretative review. Int J Dev Neurosci 1990;8:689-708.

Sirviö J, Hervonen A and Riekkinen PJ. Sodium dependent uptake of 3H-choline in the cerebral cortex of ageing male rats. Pharmacol Toxicol 1988;62:227-229.

Soares JC and Gershon S. THA--historical aspects, review of pharmacological properties and therapeutic effects. Dementia 1995;6:225-234.

Soininen H, Kosunen O, Helisalmi S, Mannermaa A, Paljärvi L, Talasniemi S, Ryynänen M and Riekkinen P, Sr. A severe loss of choline acetyltransferase in the frontal cortex of Alzheimer patients carrying apolipoprotein epsilon 4 allele. Neurosci Lett 1995;187:79-82.

Soininen H, Reinikainen KJ, Partanen J, Helkala EL, Paljärvi L and Riekkinen PJ. Slowing of electroencephalogram and choline acetyltransferase activity in post mortem frontal cortex in definite Alzheimer's disease. Neuroscience 1992;49:529-535.

Stanhope KJ, McLenachan AP and Dourish CT. Dissociation between cognitive and motor/motivational deficits in the delayed matching to position test: effects of scopolamine, 8-OH-DPAT and EAA antagonists. Psychopharmacology (Berl) 1995;122:268-280.

Steriade M and Buzsáki G. Parallel activation of thalamic and cortical neurons by brainstem and basal forebrain cholinergic systems. In: Steriade M and Biesold D, eds. Brain cholinergic system. Oxford: Oxford University Press, 1990, pp. 236-264.

Steriade M and Deschenes M. The thalamus as a neuronal oscillator. Brain Res 1984;320:1-63.

Steriade M, Gloor P, Llinas RR, Lopes de Silva FH and Mesulam MM. Report of IFCN Committee on Basic Mechanisms. Basic mechanisms of cerebral rhythmic activities. Electroencephalogr Clin Neurophysiol 1990;76:481-508.

Steriade M and Llnás RR. The functional states of the thalamus and the associated neuronal interplay. Physiol Rev 1988;68:649-742.

Strittmatter WJ and Roses AD. Apolipoprotein E and Alzheimer's disease. Annu Rev Neurosci 1996;19:53-77.

Sweeney JE, Hohmann CF, Moran TH and Coyle JT. A long-acting cholinesterase inhibitor reverses spatial memory deficits in mice. Pharmacol Biochem Behav 1988;31:141-147.

Sweeney JE, Puttfarcken PS and Coyle JT. Galanthamine, an acetylcholinesterase inhibitor: a time course of the effects on performance and neurochemical parameters in mice. Pharmacol Biochem Behav 1989;34:129-137.

Svensson AL, Warpman U, Hellstrom-Lindahl E, Bogdanovic N, Lannfelt L and Nordberg A. Nicotinic receptors, muscarinic receptors and choline acetyltransferase activity in the temporal cortex of Alzheimer patients with differing apolipoprotein E genotypes. Neurosci Lett 1997;232:37-40.

Tayebati SK, Piergentili A, Natale D and Amenta F. Evaluation of an agonist index: affinity ratio for compounds active on muscarinic cholinergic M2 receptors. J Auton Pharmacol 1999;19:77-84.

Terry RD, Masliah E, Salmon DP, Butters N, DeTeresa R, Hill R, Hansen LA and Katzman R. Physical basis of cognitive alterations in Alzheimer's disease: synapse loss is the major correlate of cognitive impairment. Ann Neurol 1991;30:572-580.

Torres EM, Perry TA, Blockland A, Wilkinson LS, Wiley RG, Lappi DA and Dunnet SB. Behavioural, histochemical and biochemical consequences of selective immunolesions in discrete regions of the basal forebrain cholinergic system. Neuroscience 1994;63:95-122.

Veinbergs I, Jung MW, Young SJ, Van Uden E, Groves PM and Masliah E. Altered long-term potentiation in the hippocampus of apolipoprotein E-deficient mice. Neurosci Lett 1998;249:71-74.

Waelbroeck M, Tastenoy M, Camus J and Christophe J. Binding of selective antagonists to four muscarinic receptors (M1 to M4) in rat forebrain. Mol Pharmacol 1990;38:267-273.

Wainer BH and Mesulam M-M. Ascending cholinergic pathways in the rat brain. In: Steriade M and Biesold D, eds. Brain cholinergic system. Oxford: Oxford University Press, 1990, pp. 63-119.

Wall SJ, Yasuda RP, Hory F, Flagg S, Martin BM, Ginns EI and Wolfe BB. Production of antisera selective for m1 muscarinic receptors using fusion proteins: distribution of m1 receptors in rat brain. Mol Pharmacol 1991;39:643-649.

Wang JX, Roeske WR, Hawkins KN, Gehlert DR and Yamamura HI. Quantitative autoradiography of M2 muscarinic receptors in the rat brain identified by using a selective radioligand [3H]AF-DX 116. Brain Res 1989;477:322-326.

Wei J, Walton EA, Milici A and Buccafusco JJ. m1-m5 muscarinic receptor distribution in rat CNS by RT-PCR and HPLC. J Neurochem 1994;63:815-821.

Wenk GL. The nucleus basalis magnocellularis cholinergic system: one hundred years of progress. Neurobiol Learn Mem 1997;67:85-95.

Wenk GL, Stoehr JD, Quintana G, Mobley S and Wiley RG. Behavioral, biochemical, histological, and electrophysiological effects of 192 IgG-saporin injections into the basal forebrain of rats. J Neurosci 1994;14:5986-5995.

Whitehouse PJ, Price DL, Struble RG, Clark AW, Coyle JT and Delon MR. Alzheimer's disease and senile dementia: loss of neurons in the basal forebrain. Science 1982;215:1237-1239.

Wilcock GK, Esiri MM, Bowen DM and Smith CC. Alzheimer's disease. Correlation of cortical choline acetyltransferase activity with the severity of dementia and histological abnormalities. J Neurol Sci 1982;57:407-417.

Wilson IA, Puoliväli J, Heikkinen T and Riekkinen P, Jr. Estrogen and NMDA receptor antagonism: effects upon reference and working memory. Eur J Pharmacol 1999;381:93-99.

Wilson PW, Schaefer EJ, Larson MG and Ordovas JM. Apolipoprotein E alleles and risk of coronary disease. A meta-analysis. Arterioscler Thromb Vasc Biol 1996;16:1250-1255.

Winkler J, Thal LJ, Gage FH and Fisher LJ. Cholinergic strategies for Alzheimer's disease. J Mol Med 1998;76:555-567.

Zhang SH, Reddick RL, Piedrahita JA and Maeda N. Spontaneous hypercholesterolemia and arterial lesions in mice lacking apolipoprotein E. Science 1992;258:468-471.